rdelaage
/
postgresql
mirror of https://git.postgresql.org/git/postgresql.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity
postgresql/src/backend/utils/adt/datetime.c

4844 lines
120 KiB
C
Raw Normal View History

Date/Time updates from Thomas...
1997-03-15 00:21:12 +01:00
/*-------------------------------------------------------------------------
*
Change my-function-name-- to my_function_name, and optimizer renames.
1999-02-14 00:22:53 +01:00
* datetime.c
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
* Support functions for date/time types.
Date/Time updates from Thomas...
1997-03-15 00:21:12 +01:00
*
Update copyrights for 2020 Backpatch-through: update all files in master, backpatch legal files through 9.4
2020-01-01 18:21:45 +01:00
* Portions Copyright (c) 1996-2020, PostgreSQL Global Development Group
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
* Portions Copyright (c) 1994, Regents of the University of California
Date/Time updates from Thomas...
1997-03-15 00:21:12 +01:00
*
*
* IDENTIFICATION
Remove cvs keywords from all files.
2010-09-20 22:08:53 +02:00
* src/backend/utils/adt/datetime.c
Date/Time updates from Thomas...
1997-03-15 00:21:12 +01:00
*
*-------------------------------------------------------------------------
*/
Remove a bunch of unused configure tests, in particular cases where * the result is not recorded anywhere * the result is not used anywhere * the result is only used in some places, whereas others have been getting away with it * the result is used improperly Also make command line options handling a little better (e.g., --disable-locale, while redundant, should really still *dis*able).
2000-07-13 00:59:15 +02:00
#include "postgres.h"
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
#include <ctype.h>
#include <limits.h>
Repair an embarrassingly large number of alphabetization mistakes in the datetime token tables. Even more embarrassing, the regression tests revealed some of the problems --- but evidently the bogus output wasn't questioned. Add code to postmaster startup to directly check the tables for correct ordering, in hopes of not being embarrassed like this again.
2003-01-16 01:26:49 +01:00
#include <math.h>
Remove a bunch of unused configure tests, in particular cases where * the result is not recorded anywhere * the result is not used anywhere * the result is only used in some places, whereas others have been getting away with it * the result is used improperly Also make command line options handling a little better (e.g., --disable-locale, while redundant, should really still *dis*able).
2000-07-13 00:59:15 +02:00
Split tuple struct defs from htup.h to htup_details.h This reduces unnecessary exposure of other headers through htup.h, which is very widely included by many files. I have chosen to move the function prototypes to the new file as well, because that means htup.h no longer needs to include tupdesc.h. In itself this doesn't have much effect in indirect inclusion of tupdesc.h throughout the tree, because it's also required by execnodes.h; but it's something to explore in the future, and it seemed best to do the htup.h change now while I'm busy with it.
2012-08-30 22:15:44 +02:00
#include "access/htup_details.h"
Clean up the rather historically encumbered interface to now() and current time: provide a GetCurrentTimestamp() function that returns current time in the form of a TimestampTz, instead of separate time_t and microseconds fields. This is what all the callers really want anyway, and it eliminates low-level dependencies on AbsoluteTime, which is a deprecated datatype that will have to disappear eventually.
2005-06-30 00:51:57 +02:00
#include "access/xact.h"
Remove hard-wired lists of timezone abbreviations in favor of providing configuration files that can be altered by a DBA. The australian_timezones GUC setting disappears, replaced by a timezone_abbreviations setting (set this to 'Australia' to get the effect of australian_timezones). The list of zone names defined by default has undergone a bit of cleanup, too. Documentation still needs some work --- in particular, should we fix Table B-4, or just get rid of it? Joachim Wieland, with some editorializing by moi.
2006-07-25 05:51:23 +02:00
#include "catalog/pg_type.h"
Move strtoint() to common Several places used similar code to convert a string to an int, so take the function that we already had and make it globally available. Reviewed-by: Michael Paquier <michael@paquier.xyz>
2018-03-13 15:21:09 +01:00
#include "common/string.h"
Remove hard-wired lists of timezone abbreviations in favor of providing configuration files that can be altered by a DBA. The australian_timezones GUC setting disappears, replaced by a timezone_abbreviations setting (set this to 'Australia' to get the effect of australian_timezones). The list of zone names defined by default has undergone a bit of cleanup, too. Documentation still needs some work --- in particular, should we fix Table B-4, or just get rid of it? Joachim Wieland, with some editorializing by moi.
2006-07-25 05:51:23 +02:00
#include "funcapi.h"
From: Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> Subject: [HACKERS] More date time functions Here are some additional patches mostly related to the date and time data types. It includes some type conversion routines to move between the different date types and some other date manipulation routines such as date_part(units,datetime). I noticed Edmund Mergl et al's neat trick for getting function overloading for builtin functions, so started to use that for the date and time stuff. Later, if someone figures out how to get function overloading directly for internal C code, then we can move to that technique. These patches include documentation updates (don't faint!) for the built-in man page. Doesn't yet include mention of timestamp, since I don't know much about it and since it may change a bit to become a _real_ ANSI timestamp which would include parser support for the declaration syntax (what do you think, Dan?). The patches were developed on the 970330 release, but have been rebuilt off of the 970402 release. The first patch below is to get libpq to compile, on my Linux box, but is not related to the rest of the patches and you can choose not to apply that one at this time. Thanks in advance, scrappy!
1997-04-02 20:36:24 +02:00
#include "miscadmin.h"
Add transform functions for various temporal typmod coercisions. This enables ALTER TABLE to skip table and index rebuilds in some cases. Noah Misch, with trivial changes by me.
2012-02-08 15:33:02 +01:00
#include "nodes/nodeFuncs.h"
Allow timezone names in SQL strings, '2006-05-24 21:11 Americas/New_York'::timestamptz Joachim Wieland
2006-06-08 00:32:31 +02:00
#include "utils/builtins.h"
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
#include "utils/date.h"
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
#include "utils/datetime.h"
Remove hard-wired lists of timezone abbreviations in favor of providing configuration files that can be altered by a DBA. The australian_timezones GUC setting disappears, replaced by a timezone_abbreviations setting (set this to 'Australia' to get the effect of australian_timezones). The list of zone names defined by default has undergone a bit of cleanup, too. Documentation still needs some work --- in particular, should we fix Table B-4, or just get rid of it? Joachim Wieland, with some editorializing by moi.
2006-07-25 05:51:23 +02:00
#include "utils/memutils.h"
#include "utils/tzparser.h"
Date/Time updates from Thomas...
1997-03-15 00:21:12 +01:00
Phase 2 pgindent run for v12. Switch to 2.1 version of pg_bsd_indent. This formats multiline function declarations "correctly", that is with additional lines of parameter declarations indented to match where the first line's left parenthesis is. Discussion: https://postgr.es/m/CAEepm=0P3FeTXRcU5B2W3jv3PgRVZ-kGUXLGfd42FFhUROO3ug@mail.gmail.com
2019-05-22 19:04:48 +02:00
static int DecodeNumber(int flen, char *field, bool haveTextMonth,
int fmask, int *tmask,
struct pg_tm *tm, fsec_t *fsec, bool *is2digits);
static int DecodeNumberField(int len, char *str,
int fmask, int *tmask,
struct pg_tm *tm, fsec_t *fsec, bool *is2digits);
static int DecodeTime(char *str, int fmask, int range,
int *tmask, struct pg_tm *tm, fsec_t *fsec);
Remove hard-wired lists of timezone abbreviations in favor of providing configuration files that can be altered by a DBA. The australian_timezones GUC setting disappears, replaced by a timezone_abbreviations setting (set this to 'Australia' to get the effect of australian_timezones). The list of zone names defined by default has undergone a bit of cleanup, too. Documentation still needs some work --- in particular, should we fix Table B-4, or just get rid of it? Joachim Wieland, with some editorializing by moi.
2006-07-25 05:51:23 +02:00
static const datetkn *datebsearch(const char *key, const datetkn *base, int nel);
Phase 2 pgindent run for v12. Switch to 2.1 version of pg_bsd_indent. This formats multiline function declarations "correctly", that is with additional lines of parameter declarations indented to match where the first line's left parenthesis is. Discussion: https://postgr.es/m/CAEepm=0P3FeTXRcU5B2W3jv3PgRVZ-kGUXLGfd42FFhUROO3ug@mail.gmail.com
2019-05-22 19:04:48 +02:00
static int DecodeDate(char *str, int fmask, int *tmask, bool *is2digits,
struct pg_tm *tm);
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
static char *AppendSeconds(char *cp, int sec, fsec_t fsec,
Phase 2 pgindent run for v12. Switch to 2.1 version of pg_bsd_indent. This formats multiline function declarations "correctly", that is with additional lines of parameter declarations indented to match where the first line's left parenthesis is. Discussion: https://postgr.es/m/CAEepm=0P3FeTXRcU5B2W3jv3PgRVZ-kGUXLGfd42FFhUROO3ug@mail.gmail.com
2019-05-22 19:04:48 +02:00
int precision, bool fillzeros);
Initial pgindent run with pg_bsd_indent version 2.0. The new indent version includes numerous fixes thanks to Piotr Stefaniak. The main changes visible in this commit are: * Nicer formatting of function-pointer declarations. * No longer unexpectedly removes spaces in expressions using casts, sizeof, or offsetof. * No longer wants to add a space in "struct structname *varname", as well as some similar cases for const- or volatile-qualified pointers. * Declarations using PG_USED_FOR_ASSERTS_ONLY are formatted more nicely. * Fixes bug where comments following declarations were sometimes placed with no space separating them from the code. * Fixes some odd decisions for comments following case labels. * Fixes some cases where comments following code were indented to less than the expected column 33. On the less good side, it now tends to put more whitespace around typedef names that are not listed in typedefs.list. This might encourage us to put more effort into typedef name collection; it's not really a bug in indent itself. There are more changes coming after this round, having to do with comment indentation and alignment of lines appearing within parentheses. I wanted to limit the size of the diffs to something that could be reviewed without one's eyes completely glazing over, so it seemed better to split up the changes as much as practical. Discussion: https://postgr.es/m/E1dAmxK-0006EE-1r@gemulon.postgresql.org Discussion: https://postgr.es/m/30527.1495162840@sss.pgh.pa.us
2017-06-21 20:39:04 +02:00
static void AdjustFractSeconds(double frac, struct pg_tm *tm, fsec_t *fsec,
Phase 2 pgindent run for v12. Switch to 2.1 version of pg_bsd_indent. This formats multiline function declarations "correctly", that is with additional lines of parameter declarations indented to match where the first line's left parenthesis is. Discussion: https://postgr.es/m/CAEepm=0P3FeTXRcU5B2W3jv3PgRVZ-kGUXLGfd42FFhUROO3ug@mail.gmail.com
2019-05-22 19:04:48 +02:00
int scale);
Initial pgindent run with pg_bsd_indent version 2.0. The new indent version includes numerous fixes thanks to Piotr Stefaniak. The main changes visible in this commit are: * Nicer formatting of function-pointer declarations. * No longer unexpectedly removes spaces in expressions using casts, sizeof, or offsetof. * No longer wants to add a space in "struct structname *varname", as well as some similar cases for const- or volatile-qualified pointers. * Declarations using PG_USED_FOR_ASSERTS_ONLY are formatted more nicely. * Fixes bug where comments following declarations were sometimes placed with no space separating them from the code. * Fixes some odd decisions for comments following case labels. * Fixes some cases where comments following code were indented to less than the expected column 33. On the less good side, it now tends to put more whitespace around typedef names that are not listed in typedefs.list. This might encourage us to put more effort into typedef name collection; it's not really a bug in indent itself. There are more changes coming after this round, having to do with comment indentation and alignment of lines appearing within parentheses. I wanted to limit the size of the diffs to something that could be reviewed without one's eyes completely glazing over, so it seemed better to split up the changes as much as practical. Discussion: https://postgr.es/m/E1dAmxK-0006EE-1r@gemulon.postgresql.org Discussion: https://postgr.es/m/30527.1495162840@sss.pgh.pa.us
2017-06-21 20:39:04 +02:00
static void AdjustFractDays(double frac, struct pg_tm *tm, fsec_t *fsec,
Phase 2 pgindent run for v12. Switch to 2.1 version of pg_bsd_indent. This formats multiline function declarations "correctly", that is with additional lines of parameter declarations indented to match where the first line's left parenthesis is. Discussion: https://postgr.es/m/CAEepm=0P3FeTXRcU5B2W3jv3PgRVZ-kGUXLGfd42FFhUROO3ug@mail.gmail.com
2019-05-22 19:04:48 +02:00
int scale);
static int DetermineTimeZoneOffsetInternal(struct pg_tm *tm, pg_tz *tzp,
pg_time_t *tp);
Don't require dynamic timezone abbreviations to match underlying time zone. Previously, we threw an error if a dynamic timezone abbreviation did not match any abbreviation recorded in the referenced IANA time zone entry. That seemed like a good consistency check at the time, but it turns out that a number of the abbreviations in the IANA database are things that Olson and crew made up out of whole cloth. Their current policy is to remove such names in favor of using simple numeric offsets. Perhaps unsurprisingly, a lot of these made-up abbreviations have varied in meaning over time, which meant that our commit b2cbced9e and later changes made them into dynamic abbreviations. So with newer IANA database versions that don't mention these abbreviations at all, we fail, as reported in bug #14307 from Neil Anderson. It's worse than just a few unused-in-the-wild abbreviations not working, because the pg_timezone_abbrevs view stops working altogether (since its underlying function tries to compute the whole view result in one call). We considered deleting these abbreviations from our abbreviations list, but the problem with that is that we can't stay ahead of possible future IANA changes. Instead, let's leave the abbreviations list alone, and treat any "orphaned" dynamic abbreviation as just meaning the referenced time zone. It will behave a bit differently than it used to, in that you can't any longer override the zone's standard vs. daylight rule by using the "wrong" abbreviation of a pair, but that's better than failing entirely. (Also, this solution can be interpreted as adding a small new feature, which is that any abbreviation a user wants can be defined as referencing a time zone name.) Back-patch to all supported branches, since this problem affects all of them when using tzdata 2016f or newer. Report: <20160902031551.15674.67337@wrigleys.postgresql.org> Discussion: <6189.1472820913@sss.pgh.pa.us>
2016-09-02 23:29:31 +02:00
static bool DetermineTimeZoneAbbrevOffsetInternal(pg_time_t t,
Phase 2 pgindent run for v12. Switch to 2.1 version of pg_bsd_indent. This formats multiline function declarations "correctly", that is with additional lines of parameter declarations indented to match where the first line's left parenthesis is. Discussion: https://postgr.es/m/CAEepm=0P3FeTXRcU5B2W3jv3PgRVZ-kGUXLGfd42FFhUROO3ug@mail.gmail.com
2019-05-22 19:04:48 +02:00
const char *abbr, pg_tz *tzp,
int *offset, int *isdst);
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
static pg_tz *FetchDynamicTimeZone(TimeZoneAbbrevTable *tbl, const datetkn *tp);
Support full POSIX-style time zone: EST+3, PST-3, etc. We probably support a superset of the spec, but I don't have the spec to confirm this. Update regression tests to include tests for this format.
2000-03-16 15:38:42 +01:00
Remove formatter's assumption that year would never exceed four digits. Enforce MAXTZLEN for all datestyles, not just some. Remove macro definitions that were redundant with datetime.h.
2001-07-10 03:41:47 +02:00
Adjust our timezone library to use pg_time_t (typedef'd as int64) in place of time_t, as per prior discussion. The behavior does not change on machines without a 64-bit-int type, but on machines with one, which is most, we are rid of the bizarre boundary behavior at the edges of the 32-bit-time_t range (1901 and 2038). The system will now treat times over the full supported timestamp range as being in your local time zone. It may seem a little bizarre to consider that times in 4000 BC are PST or EST, but this is surely at least as reasonable as propagating Gregorian calendar rules back that far. I did not modify the format of the zic timezone database files, which means that for the moment the system will not know about daylight-savings periods outside the range 1901-2038. Given the way the files are set up, it's not a simple decision like 'widen to 64 bits'; we have to actually think about the range of years that need to be supported. We should probably inquire what the plans of the upstream zic people are before making any decisions of our own.
2004-06-03 04:08:07 +02:00
const int day_tab[2][13] =
{
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
{31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31, 0},
Adjust our timezone library to use pg_time_t (typedef'd as int64) in place of time_t, as per prior discussion. The behavior does not change on machines without a 64-bit-int type, but on machines with one, which is most, we are rid of the bizarre boundary behavior at the edges of the 32-bit-time_t range (1901 and 2038). The system will now treat times over the full supported timestamp range as being in your local time zone. It may seem a little bizarre to consider that times in 4000 BC are PST or EST, but this is surely at least as reasonable as propagating Gregorian calendar rules back that far. I did not modify the format of the zic timezone database files, which means that for the moment the system will not know about daylight-savings periods outside the range 1901-2038. Given the way the files are set up, it's not a simple decision like 'widen to 64 bits'; we have to actually think about the range of years that need to be supported. We should probably inquire what the plans of the upstream zic people are before making any decisions of our own.
2004-06-03 04:08:07 +02:00
{31, 29, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31, 0}
};
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
Make various variables const (read-only). These changes should generally improve correctness/maintainability. A nice side benefit is that several kilobytes move from initialized data to text segment, allowing them to be shared across processes and probably reducing copy-on-write overhead while forking a new backend. Unfortunately this doesn't seem to help libpq in the same way (at least not when it's compiled with -fpic on x86_64), but we can hope the linker at least collects all nominally-const data together even if it's not actually part of the text segment. Also, make pg_encname_tbl[] static in encnames.c, since there seems no very good reason for any other code to use it; per a suggestion from Wim Lewis, who independently submitted a patch that was mostly a subset of this one. Oskari Saarenmaa, with some editorialization by me
2014-01-18 22:04:11 +01:00
const char *const months[] = {"Jan", "Feb", "Mar", "Apr", "May", "Jun",
pgindent run.
2002-09-04 22:31:48 +02:00
"Jul", "Aug", "Sep", "Oct", "Nov", "Dec", NULL};
Date/Time updates from Thomas...
1997-03-15 00:21:12 +01:00
Make various variables const (read-only). These changes should generally improve correctness/maintainability. A nice side benefit is that several kilobytes move from initialized data to text segment, allowing them to be shared across processes and probably reducing copy-on-write overhead while forking a new backend. Unfortunately this doesn't seem to help libpq in the same way (at least not when it's compiled with -fpic on x86_64), but we can hope the linker at least collects all nominally-const data together even if it's not actually part of the text segment. Also, make pg_encname_tbl[] static in encnames.c, since there seems no very good reason for any other code to use it; per a suggestion from Wim Lewis, who independently submitted a patch that was mostly a subset of this one. Oskari Saarenmaa, with some editorialization by me
2014-01-18 22:04:11 +01:00
const char *const days[] = {"Sunday", "Monday", "Tuesday", "Wednesday",
pgindent run.
2002-09-04 22:31:48 +02:00
"Thursday", "Friday", "Saturday", NULL};
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
/*****************************************************************************
* PRIVATE ROUTINES *
*****************************************************************************/
Date/Time updates from Thomas...
1997-03-15 00:21:12 +01:00
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
/*
Repair an embarrassingly large number of alphabetization mistakes in the datetime token tables. Even more embarrassing, the regression tests revealed some of the problems --- but evidently the bogus output wasn't questioned. Add code to postmaster startup to directly check the tables for correct ordering, in hopes of not being embarrassed like this again.
2003-01-16 01:26:49 +01:00
* datetktbl holds date/time keywords.
*
* Note that this table must be strictly alphabetically ordered to allow an
* O(ln(N)) search algorithm to be used.
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
*
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
* The token field must be NUL-terminated; we truncate entries to TOKMAXLEN
* characters to fit.
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
*
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
* The static table contains no TZ, DTZ, or DYNTZ entries; rather those
* are loaded from configuration files and stored in zoneabbrevtbl, whose
* abbrevs[] field has the same format as the static datetktbl.
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
*/
Remove hard-wired lists of timezone abbreviations in favor of providing configuration files that can be altered by a DBA. The australian_timezones GUC setting disappears, replaced by a timezone_abbreviations setting (set this to 'Australia' to get the effect of australian_timezones). The list of zone names defined by default has undergone a bit of cleanup, too. Documentation still needs some work --- in particular, should we fix Table B-4, or just get rid of it? Joachim Wieland, with some editorializing by moi.
2006-07-25 05:51:23 +02:00
static const datetkn datetktbl[] = {
Prevent potential overruns of fixed-size buffers. Coverity identified a number of places in which it couldn't prove that a string being copied into a fixed-size buffer would fit. We believe that most, perhaps all of these are in fact safe, or are copying data that is coming from a trusted source so that any overrun is not really a security issue. Nonetheless it seems prudent to forestall any risk by using strlcpy() and similar functions. Fixes by Peter Eisentraut and Jozef Mlich based on Coverity reports. In addition, fix a potential null-pointer-dereference crash in contrib/chkpass. The crypt(3) function is defined to return NULL on failure, but chkpass.c didn't check for that before using the result. The main practical case in which this could be an issue is if libc is configured to refuse to execute unapproved hashing algorithms (e.g., "FIPS mode"). This ideally should've been a separate commit, but since it touches code adjacent to one of the buffer overrun changes, I included it in this commit to avoid last-minute merge issues. This issue was reported by Honza Horak. Security: CVE-2014-0065 for buffer overruns, CVE-2014-0066 for crypt()
2014-02-17 17:20:21 +01:00
/* token, type, value */
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
{EARLY, RESERV, DTK_EARLY}, /* "-infinity" reserved for "early time" */
Fix to_char for 1 BC. Previously it returned 1 AD. Fix to_char(year) for BC dates. Previously it returned one less than the current year. Add documentation mentioning that there is no 0 AD.
2004-03-30 17:53:18 +02:00
{DA_D, ADBC, AD}, /* "ad" for years > 0 */
Phase 2 of pgindent updates. Change pg_bsd_indent to follow upstream rules for placement of comments to the right of code, and remove pgindent hack that caused comments following #endif to not obey the general rule. Commit e3860ffa4dd0dad0dd9eea4be9cc1412373a8c89 wasn't actually using the published version of pg_bsd_indent, but a hacked-up version that tried to minimize the amount of movement of comments to the right of code. The situation of interest is where such a comment has to be moved to the right of its default placement at column 33 because there's code there. BSD indent has always moved right in units of tab stops in such cases --- but in the previous incarnation, indent was working in 8-space tab stops, while now it knows we use 4-space tabs. So the net result is that in about half the cases, such comments are placed one tab stop left of before. This is better all around: it leaves more room on the line for comment text, and it means that in such cases the comment uniformly starts at the next 4-space tab stop after the code, rather than sometimes one and sometimes two tabs after. Also, ensure that comments following #endif are indented the same as comments following other preprocessor commands such as #else. That inconsistency turns out to have been self-inflicted damage from a poorly-thought-through post-indent "fixup" in pgindent. This patch is much less interesting than the first round of indent changes, but also bulkier, so I thought it best to separate the effects. Discussion: https://postgr.es/m/E1dAmxK-0006EE-1r@gemulon.postgresql.org Discussion: https://postgr.es/m/30527.1495162840@sss.pgh.pa.us
2017-06-21 21:18:54 +02:00
{"allballs", RESERV, DTK_ZULU}, /* 00:00:00 */
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
{"am", AMPM, AM},
Repair an embarrassingly large number of alphabetization mistakes in the datetime token tables. Even more embarrassing, the regression tests revealed some of the problems --- but evidently the bogus output wasn't questioned. Add code to postmaster startup to directly check the tables for correct ordering, in hopes of not being embarrassed like this again.
2003-01-16 01:26:49 +01:00
{"apr", MONTH, 4},
{"april", MONTH, 4},
Add guard code to protect from buffer overruns on long date/time input strings. Should go back in and look at doing this a bit more elegantly and (hopefully) cheaper. Probably not too bad anyway, but it seems a shame to scan the strings twice: once for length for this buffer overrun protection, and once to parse the line. Remove use of pow() in date/time handling; was already gone from everything *but* the time data types. Define macros for handling typmod manipulation for date/time types. Should be more robust than all of that brute-force inline code. Rename macros for masking and typmod manipulation to put TIMESTAMP_ or INTERVAL_ in front of the macro name, to reduce the possibility of name space collisions.
2002-08-04 08:44:47 +02:00
{"at", IGNORE_DTF, 0}, /* "at" (throwaway) */
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
{"aug", MONTH, 8},
{"august", MONTH, 8},
Fix to_char for 1 BC. Previously it returned 1 AD. Fix to_char(year) for BC dates. Previously it returned one less than the current year. Add documentation mentioning that there is no 0 AD.
2004-03-30 17:53:18 +02:00
{DB_C, ADBC, BC}, /* "bc" for years <= 0 */
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
{"d", UNITS, DTK_DAY}, /* "day of month" for ISO input */
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
{"dec", MONTH, 12},
{"december", MONTH, 12},
Move DTK_ISODOW DTK_DOW and DTK_DOY to be type UNITS rather than RESERV. RESERV is meant for tokens like "now" and having them in that category throws errors like these when used as an input date: stark=# SELECT 'doy'::timestamptz; ERROR: unexpected dtype 33 while parsing timestamptz "doy" LINE 1: SELECT 'doy'::timestamptz; ^ stark=# SELECT 'dow'::timestamptz; ERROR: unexpected dtype 32 while parsing timestamptz "dow" LINE 1: SELECT 'dow'::timestamptz; ^ Found by LLVM's Libfuzzer
2015-09-06 03:04:37 +02:00
{"dow", UNITS, DTK_DOW}, /* day of week */
{"doy", UNITS, DTK_DOY}, /* day of year */
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
{"dst", DTZMOD, SECS_PER_HOUR},
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
{EPOCH, RESERV, DTK_EPOCH}, /* "epoch" reserved for system epoch time */
{"feb", MONTH, 2},
{"february", MONTH, 2},
{"fri", DOW, 5},
{"friday", DOW, 5},
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
{"h", UNITS, DTK_HOUR}, /* "hour" */
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
{LATE, RESERV, DTK_LATE}, /* "infinity" reserved for "late time" */
Phase 2 of pgindent updates. Change pg_bsd_indent to follow upstream rules for placement of comments to the right of code, and remove pgindent hack that caused comments following #endif to not obey the general rule. Commit e3860ffa4dd0dad0dd9eea4be9cc1412373a8c89 wasn't actually using the published version of pg_bsd_indent, but a hacked-up version that tried to minimize the amount of movement of comments to the right of code. The situation of interest is where such a comment has to be moved to the right of its default placement at column 33 because there's code there. BSD indent has always moved right in units of tab stops in such cases --- but in the previous incarnation, indent was working in 8-space tab stops, while now it knows we use 4-space tabs. So the net result is that in about half the cases, such comments are placed one tab stop left of before. This is better all around: it leaves more room on the line for comment text, and it means that in such cases the comment uniformly starts at the next 4-space tab stop after the code, rather than sometimes one and sometimes two tabs after. Also, ensure that comments following #endif are indented the same as comments following other preprocessor commands such as #else. That inconsistency turns out to have been self-inflicted damage from a poorly-thought-through post-indent "fixup" in pgindent. This patch is much less interesting than the first round of indent changes, but also bulkier, so I thought it best to separate the effects. Discussion: https://postgr.es/m/E1dAmxK-0006EE-1r@gemulon.postgresql.org Discussion: https://postgr.es/m/30527.1495162840@sss.pgh.pa.us
2017-06-21 21:18:54 +02:00
{"isodow", UNITS, DTK_ISODOW}, /* ISO day of week, Sunday == 7 */
Add two new format fields for use with to_char(), to_date() and to_timestamp(): - ID for day-of-week - IDDD for day-of-year This makes it possible to convert ISO week dates to and from text fully represented in either week ('IYYY-IW-ID') or day-of-year ('IYYY-IDDD') format. I have also added an 'isoyear' field for use with extract / date_part. Brendan Jurd
2007-02-16 04:39:46 +01:00
{"isoyear", UNITS, DTK_ISOYEAR}, /* year in terms of the ISO week date */
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
{"j", UNITS, DTK_JULIAN},
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
{"jan", MONTH, 1},
{"january", MONTH, 1},
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
{"jd", UNITS, DTK_JULIAN},
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
{"jul", MONTH, 7},
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
{"julian", UNITS, DTK_JULIAN},
Add full name of the month of July to the lookup table. Thanks to Greg Sabino Mullane <greg@turnstep.com> for finding the problem.
2001-12-21 07:03:27 +01:00
{"july", MONTH, 7},
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
{"jun", MONTH, 6},
{"june", MONTH, 6},
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
{"m", UNITS, DTK_MONTH}, /* "month" for ISO input */
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
{"mar", MONTH, 3},
{"march", MONTH, 3},
{"may", MONTH, 5},
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
{"mm", UNITS, DTK_MINUTE}, /* "minute" for ISO input */
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
{"mon", DOW, 1},
{"monday", DOW, 1},
{"nov", MONTH, 11},
{"november", MONTH, 11},
{NOW, RESERV, DTK_NOW}, /* current transaction time */
{"oct", MONTH, 10},
{"october", MONTH, 10},
Add guard code to protect from buffer overruns on long date/time input strings. Should go back in and look at doing this a bit more elegantly and (hopefully) cheaper. Probably not too bad anyway, but it seems a shame to scan the strings twice: once for length for this buffer overrun protection, and once to parse the line. Remove use of pow() in date/time handling; was already gone from everything *but* the time data types. Define macros for handling typmod manipulation for date/time types. Should be more robust than all of that brute-force inline code. Rename macros for masking and typmod manipulation to put TIMESTAMP_ or INTERVAL_ in front of the macro name, to reduce the possibility of name space collisions.
2002-08-04 08:44:47 +02:00
{"on", IGNORE_DTF, 0}, /* "on" (throwaway) */
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
{"pm", AMPM, PM},
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
{"s", UNITS, DTK_SECOND}, /* "seconds" for ISO input */
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
{"sat", DOW, 6},
{"saturday", DOW, 6},
{"sep", MONTH, 9},
{"sept", MONTH, 9},
{"september", MONTH, 9},
{"sun", DOW, 0},
{"sunday", DOW, 0},
Add a large number of time zones to the lookup table. Fix a few apparently-wrong TZ vs DTZ declarations.
2002-02-25 17:17:04 +01:00
{"t", ISOTIME, DTK_TIME}, /* Filler for ISO time fields */
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
{"thu", DOW, 4},
{"thur", DOW, 4},
{"thurs", DOW, 4},
{"thursday", DOW, 4},
{TODAY, RESERV, DTK_TODAY}, /* midnight */
{TOMORROW, RESERV, DTK_TOMORROW}, /* tomorrow midnight */
{"tue", DOW, 2},
{"tues", DOW, 2},
{"tuesday", DOW, 2},
{"wed", DOW, 3},
{"wednesday", DOW, 3},
{"weds", DOW, 3},
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
{"y", UNITS, DTK_YEAR}, /* "year" for ISO input */
Fix up some problems in handling of zic-style time zone names in datetime input routines. Remove the former "DecodePosixTimezone" function in favor of letting the zic code handle POSIX-style zone specs (see tzparse()). In particular this means that "PST+3" now means the same as "-03", whereas it used to mean "-11" --- the zone abbreviation is effectively just a noise word in this syntax. Make sure that all named and POSIX-style zone names will be parsed as a single token. Fix long-standing bogosities in printing and input of fractional-hour timezone offsets (since the tzparse() code will accept these, we'd better make 'em work). Also correct an error in the original coding of the zic-zone-name patch: in "timestamp without time zone" input, zone names are supposed to be allowed but ignored, but the coding was such that the zone changed the interpretation anyway.
2006-10-17 23:03:21 +02:00
{YESTERDAY, RESERV, DTK_YESTERDAY} /* yesterday midnight */
Add GUC setting for Australian timezones. Uses new GUC boolean callback functions to clear date cache. Allow regression tests to pass when timezone set.
2001-06-18 18:14:44 +02:00
};
Const-ify a few more large static tables. Per research by Andres. Discussion: https://postgr.es/m/20181015200754.7y7zfuzsoux2c4ya@alap3.anarazel.de
2018-10-17 21:32:47 +02:00
static const int szdatetktbl = sizeof datetktbl / sizeof datetktbl[0];
Add GUC setting for Australian timezones. Uses new GUC boolean callback functions to clear date cache. Allow regression tests to pass when timezone set.
2001-06-18 18:14:44 +02:00
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
/*
* deltatktbl: same format as datetktbl, but holds keywords used to represent
* time units (eg, for intervals, and for EXTRACT).
*/
Make various variables const (read-only). These changes should generally improve correctness/maintainability. A nice side benefit is that several kilobytes move from initialized data to text segment, allowing them to be shared across processes and probably reducing copy-on-write overhead while forking a new backend. Unfortunately this doesn't seem to help libpq in the same way (at least not when it's compiled with -fpic on x86_64), but we can hope the linker at least collects all nominally-const data together even if it's not actually part of the text segment. Also, make pg_encname_tbl[] static in encnames.c, since there seems no very good reason for any other code to use it; per a suggestion from Wim Lewis, who independently submitted a patch that was mostly a subset of this one. Oskari Saarenmaa, with some editorialization by me
2014-01-18 22:04:11 +01:00
static const datetkn deltatktbl[] = {
Prevent potential overruns of fixed-size buffers. Coverity identified a number of places in which it couldn't prove that a string being copied into a fixed-size buffer would fit. We believe that most, perhaps all of these are in fact safe, or are copying data that is coming from a trusted source so that any overrun is not really a security issue. Nonetheless it seems prudent to forestall any risk by using strlcpy() and similar functions. Fixes by Peter Eisentraut and Jozef Mlich based on Coverity reports. In addition, fix a potential null-pointer-dereference crash in contrib/chkpass. The crypt(3) function is defined to return NULL on failure, but chkpass.c didn't check for that before using the result. The main practical case in which this could be an issue is if libc is configured to refuse to execute unapproved hashing algorithms (e.g., "FIPS mode"). This ideally should've been a separate commit, but since it touches code adjacent to one of the buffer overrun changes, I included it in this commit to avoid last-minute merge issues. This issue was reported by Honza Horak. Security: CVE-2014-0065 for buffer overruns, CVE-2014-0066 for crypt()
2014-02-17 17:20:21 +01:00
/* token, type, value */
Add guard code to protect from buffer overruns on long date/time input strings. Should go back in and look at doing this a bit more elegantly and (hopefully) cheaper. Probably not too bad anyway, but it seems a shame to scan the strings twice: once for length for this buffer overrun protection, and once to parse the line. Remove use of pow() in date/time handling; was already gone from everything *but* the time data types. Define macros for handling typmod manipulation for date/time types. Should be more robust than all of that brute-force inline code. Rename macros for masking and typmod manipulation to put TIMESTAMP_ or INTERVAL_ in front of the macro name, to reduce the possibility of name space collisions.
2002-08-04 08:44:47 +02:00
{"@", IGNORE_DTF, 0}, /* postgres relative prefix */
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
{DAGO, AGO, 0}, /* "ago" indicates negative time offset */
Accept an INTERVAL argument for SET TIME ZONE per SQL99. Modified the parser and the SET handlers to use full Node structures rather than simply a character string argument. Implement INTERVAL() YEAR TO MONTH (etc) syntax per SQL99. Does not yet accept the goofy string format that goes along with, but this should be fairly straight forward to fix now as a bug or later as a feature. Implement precision for the INTERVAL() type. Use the typmod mechanism for both of INTERVAL features. Fix the INTERVAL syntax in the parser: opt_interval was in the wrong place. INTERVAL is now a reserved word, otherwise we get reduce/reduce errors. Implement an explicit date_part() function for TIMETZ. Should fix coersion problem with INTERVAL reported by Peter E. Fix up some error messages for date/time types. Use all caps for type names within message. Fix recently introduced side-effect bug disabling 'epoch' as a recognized field for date_part() etc. Reported by Peter E. (??) Bump catalog version number. Rename "microseconds" current transaction time field from ...Msec to ...Usec. Duh! date/time regression tests updated for reference platform, but a few changes will be necessary for others.
2001-10-18 19:30:21 +02:00
{"c", UNITS, DTK_CENTURY}, /* "century" relative */
Phase 2 of pgindent updates. Change pg_bsd_indent to follow upstream rules for placement of comments to the right of code, and remove pgindent hack that caused comments following #endif to not obey the general rule. Commit e3860ffa4dd0dad0dd9eea4be9cc1412373a8c89 wasn't actually using the published version of pg_bsd_indent, but a hacked-up version that tried to minimize the amount of movement of comments to the right of code. The situation of interest is where such a comment has to be moved to the right of its default placement at column 33 because there's code there. BSD indent has always moved right in units of tab stops in such cases --- but in the previous incarnation, indent was working in 8-space tab stops, while now it knows we use 4-space tabs. So the net result is that in about half the cases, such comments are placed one tab stop left of before. This is better all around: it leaves more room on the line for comment text, and it means that in such cases the comment uniformly starts at the next 4-space tab stop after the code, rather than sometimes one and sometimes two tabs after. Also, ensure that comments following #endif are indented the same as comments following other preprocessor commands such as #else. That inconsistency turns out to have been self-inflicted damage from a poorly-thought-through post-indent "fixup" in pgindent. This patch is much less interesting than the first round of indent changes, but also bulkier, so I thought it best to separate the effects. Discussion: https://postgr.es/m/E1dAmxK-0006EE-1r@gemulon.postgresql.org Discussion: https://postgr.es/m/30527.1495162840@sss.pgh.pa.us
2017-06-21 21:18:54 +02:00
{"cent", UNITS, DTK_CENTURY}, /* "century" relative */
Accept an INTERVAL argument for SET TIME ZONE per SQL99. Modified the parser and the SET handlers to use full Node structures rather than simply a character string argument. Implement INTERVAL() YEAR TO MONTH (etc) syntax per SQL99. Does not yet accept the goofy string format that goes along with, but this should be fairly straight forward to fix now as a bug or later as a feature. Implement precision for the INTERVAL() type. Use the typmod mechanism for both of INTERVAL features. Fix the INTERVAL syntax in the parser: opt_interval was in the wrong place. INTERVAL is now a reserved word, otherwise we get reduce/reduce errors. Implement an explicit date_part() function for TIMETZ. Should fix coersion problem with INTERVAL reported by Peter E. Fix up some error messages for date/time types. Use all caps for type names within message. Fix recently introduced side-effect bug disabling 'epoch' as a recognized field for date_part() etc. Reported by Peter E. (??) Bump catalog version number. Rename "microseconds" current transaction time field from ...Msec to ...Usec. Duh! date/time regression tests updated for reference platform, but a few changes will be necessary for others.
2001-10-18 19:30:21 +02:00
{"centuries", UNITS, DTK_CENTURY}, /* "centuries" relative */
Phase 2 of pgindent updates. Change pg_bsd_indent to follow upstream rules for placement of comments to the right of code, and remove pgindent hack that caused comments following #endif to not obey the general rule. Commit e3860ffa4dd0dad0dd9eea4be9cc1412373a8c89 wasn't actually using the published version of pg_bsd_indent, but a hacked-up version that tried to minimize the amount of movement of comments to the right of code. The situation of interest is where such a comment has to be moved to the right of its default placement at column 33 because there's code there. BSD indent has always moved right in units of tab stops in such cases --- but in the previous incarnation, indent was working in 8-space tab stops, while now it knows we use 4-space tabs. So the net result is that in about half the cases, such comments are placed one tab stop left of before. This is better all around: it leaves more room on the line for comment text, and it means that in such cases the comment uniformly starts at the next 4-space tab stop after the code, rather than sometimes one and sometimes two tabs after. Also, ensure that comments following #endif are indented the same as comments following other preprocessor commands such as #else. That inconsistency turns out to have been self-inflicted damage from a poorly-thought-through post-indent "fixup" in pgindent. This patch is much less interesting than the first round of indent changes, but also bulkier, so I thought it best to separate the effects. Discussion: https://postgr.es/m/E1dAmxK-0006EE-1r@gemulon.postgresql.org Discussion: https://postgr.es/m/30527.1495162840@sss.pgh.pa.us
2017-06-21 21:18:54 +02:00
{DCENTURY, UNITS, DTK_CENTURY}, /* "century" relative */
pgindent run on all C files. Java run to follow. initdb/regression tests pass.
2001-10-25 07:50:21 +02:00
{"d", UNITS, DTK_DAY}, /* "day" relative */
{DDAY, UNITS, DTK_DAY}, /* "day" relative */
{"days", UNITS, DTK_DAY}, /* "days" relative */
{"dec", UNITS, DTK_DECADE}, /* "decade" relative */
Phase 2 of pgindent updates. Change pg_bsd_indent to follow upstream rules for placement of comments to the right of code, and remove pgindent hack that caused comments following #endif to not obey the general rule. Commit e3860ffa4dd0dad0dd9eea4be9cc1412373a8c89 wasn't actually using the published version of pg_bsd_indent, but a hacked-up version that tried to minimize the amount of movement of comments to the right of code. The situation of interest is where such a comment has to be moved to the right of its default placement at column 33 because there's code there. BSD indent has always moved right in units of tab stops in such cases --- but in the previous incarnation, indent was working in 8-space tab stops, while now it knows we use 4-space tabs. So the net result is that in about half the cases, such comments are placed one tab stop left of before. This is better all around: it leaves more room on the line for comment text, and it means that in such cases the comment uniformly starts at the next 4-space tab stop after the code, rather than sometimes one and sometimes two tabs after. Also, ensure that comments following #endif are indented the same as comments following other preprocessor commands such as #else. That inconsistency turns out to have been self-inflicted damage from a poorly-thought-through post-indent "fixup" in pgindent. This patch is much less interesting than the first round of indent changes, but also bulkier, so I thought it best to separate the effects. Discussion: https://postgr.es/m/E1dAmxK-0006EE-1r@gemulon.postgresql.org Discussion: https://postgr.es/m/30527.1495162840@sss.pgh.pa.us
2017-06-21 21:18:54 +02:00
{DDECADE, UNITS, DTK_DECADE}, /* "decade" relative */
{"decades", UNITS, DTK_DECADE}, /* "decades" relative */
Repair an embarrassingly large number of alphabetization mistakes in the datetime token tables. Even more embarrassing, the regression tests revealed some of the problems --- but evidently the bogus output wasn't questioned. Add code to postmaster startup to directly check the tables for correct ordering, in hopes of not being embarrassed like this again.
2003-01-16 01:26:49 +01:00
{"decs", UNITS, DTK_DECADE}, /* "decades" relative */
pgindent run on all C files. Java run to follow. initdb/regression tests pass.
2001-10-25 07:50:21 +02:00
{"h", UNITS, DTK_HOUR}, /* "hour" relative */
{DHOUR, UNITS, DTK_HOUR}, /* "hour" relative */
{"hours", UNITS, DTK_HOUR}, /* "hours" relative */
{"hr", UNITS, DTK_HOUR}, /* "hour" relative */
{"hrs", UNITS, DTK_HOUR}, /* "hours" relative */
{"m", UNITS, DTK_MINUTE}, /* "minute" relative */
Phase 2 of pgindent updates. Change pg_bsd_indent to follow upstream rules for placement of comments to the right of code, and remove pgindent hack that caused comments following #endif to not obey the general rule. Commit e3860ffa4dd0dad0dd9eea4be9cc1412373a8c89 wasn't actually using the published version of pg_bsd_indent, but a hacked-up version that tried to minimize the amount of movement of comments to the right of code. The situation of interest is where such a comment has to be moved to the right of its default placement at column 33 because there's code there. BSD indent has always moved right in units of tab stops in such cases --- but in the previous incarnation, indent was working in 8-space tab stops, while now it knows we use 4-space tabs. So the net result is that in about half the cases, such comments are placed one tab stop left of before. This is better all around: it leaves more room on the line for comment text, and it means that in such cases the comment uniformly starts at the next 4-space tab stop after the code, rather than sometimes one and sometimes two tabs after. Also, ensure that comments following #endif are indented the same as comments following other preprocessor commands such as #else. That inconsistency turns out to have been self-inflicted damage from a poorly-thought-through post-indent "fixup" in pgindent. This patch is much less interesting than the first round of indent changes, but also bulkier, so I thought it best to separate the effects. Discussion: https://postgr.es/m/E1dAmxK-0006EE-1r@gemulon.postgresql.org Discussion: https://postgr.es/m/30527.1495162840@sss.pgh.pa.us
2017-06-21 21:18:54 +02:00
{"microsecon", UNITS, DTK_MICROSEC}, /* "microsecond" relative */
{"mil", UNITS, DTK_MILLENNIUM}, /* "millennium" relative */
{"millennia", UNITS, DTK_MILLENNIUM}, /* "millennia" relative */
{DMILLENNIUM, UNITS, DTK_MILLENNIUM}, /* "millennium" relative */
{"millisecon", UNITS, DTK_MILLISEC}, /* relative */
pgindent run.
2002-09-04 22:31:48 +02:00
{"mils", UNITS, DTK_MILLENNIUM}, /* "millennia" relative */
{"min", UNITS, DTK_MINUTE}, /* "minute" relative */
Another pgindent run. Fixes enum indenting, and improves #endif spacing. Also adds space for one-line comments.
2001-10-28 07:26:15 +01:00
{"mins", UNITS, DTK_MINUTE}, /* "minutes" relative */
Phase 2 of pgindent updates. Change pg_bsd_indent to follow upstream rules for placement of comments to the right of code, and remove pgindent hack that caused comments following #endif to not obey the general rule. Commit e3860ffa4dd0dad0dd9eea4be9cc1412373a8c89 wasn't actually using the published version of pg_bsd_indent, but a hacked-up version that tried to minimize the amount of movement of comments to the right of code. The situation of interest is where such a comment has to be moved to the right of its default placement at column 33 because there's code there. BSD indent has always moved right in units of tab stops in such cases --- but in the previous incarnation, indent was working in 8-space tab stops, while now it knows we use 4-space tabs. So the net result is that in about half the cases, such comments are placed one tab stop left of before. This is better all around: it leaves more room on the line for comment text, and it means that in such cases the comment uniformly starts at the next 4-space tab stop after the code, rather than sometimes one and sometimes two tabs after. Also, ensure that comments following #endif are indented the same as comments following other preprocessor commands such as #else. That inconsistency turns out to have been self-inflicted damage from a poorly-thought-through post-indent "fixup" in pgindent. This patch is much less interesting than the first round of indent changes, but also bulkier, so I thought it best to separate the effects. Discussion: https://postgr.es/m/E1dAmxK-0006EE-1r@gemulon.postgresql.org Discussion: https://postgr.es/m/30527.1495162840@sss.pgh.pa.us
2017-06-21 21:18:54 +02:00
{DMINUTE, UNITS, DTK_MINUTE}, /* "minute" relative */
{"minutes", UNITS, DTK_MINUTE}, /* "minutes" relative */
pgindent run on all C files. Java run to follow. initdb/regression tests pass.
2001-10-25 07:50:21 +02:00
{"mon", UNITS, DTK_MONTH}, /* "months" relative */
{"mons", UNITS, DTK_MONTH}, /* "months" relative */
{DMONTH, UNITS, DTK_MONTH}, /* "month" relative */
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
{"months", UNITS, DTK_MONTH},
{"ms", UNITS, DTK_MILLISEC},
{"msec", UNITS, DTK_MILLISEC},
{DMILLISEC, UNITS, DTK_MILLISEC},
{"mseconds", UNITS, DTK_MILLISEC},
{"msecs", UNITS, DTK_MILLISEC},
Another pgindent run. Fixes enum indenting, and improves #endif spacing. Also adds space for one-line comments.
2001-10-28 07:26:15 +01:00
{"qtr", UNITS, DTK_QUARTER}, /* "quarter" relative */
Phase 2 of pgindent updates. Change pg_bsd_indent to follow upstream rules for placement of comments to the right of code, and remove pgindent hack that caused comments following #endif to not obey the general rule. Commit e3860ffa4dd0dad0dd9eea4be9cc1412373a8c89 wasn't actually using the published version of pg_bsd_indent, but a hacked-up version that tried to minimize the amount of movement of comments to the right of code. The situation of interest is where such a comment has to be moved to the right of its default placement at column 33 because there's code there. BSD indent has always moved right in units of tab stops in such cases --- but in the previous incarnation, indent was working in 8-space tab stops, while now it knows we use 4-space tabs. So the net result is that in about half the cases, such comments are placed one tab stop left of before. This is better all around: it leaves more room on the line for comment text, and it means that in such cases the comment uniformly starts at the next 4-space tab stop after the code, rather than sometimes one and sometimes two tabs after. Also, ensure that comments following #endif are indented the same as comments following other preprocessor commands such as #else. That inconsistency turns out to have been self-inflicted damage from a poorly-thought-through post-indent "fixup" in pgindent. This patch is much less interesting than the first round of indent changes, but also bulkier, so I thought it best to separate the effects. Discussion: https://postgr.es/m/E1dAmxK-0006EE-1r@gemulon.postgresql.org Discussion: https://postgr.es/m/30527.1495162840@sss.pgh.pa.us
2017-06-21 21:18:54 +02:00
{DQUARTER, UNITS, DTK_QUARTER}, /* "quarter" relative */
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
{"s", UNITS, DTK_SECOND},
{"sec", UNITS, DTK_SECOND},
{DSECOND, UNITS, DTK_SECOND},
{"seconds", UNITS, DTK_SECOND},
{"secs", UNITS, DTK_SECOND},
pgindent run.
2002-09-04 22:31:48 +02:00
{DTIMEZONE, UNITS, DTK_TZ}, /* "timezone" time offset */
pgindent run on all C files. Java run to follow. initdb/regression tests pass.
2001-10-25 07:50:21 +02:00
{"timezone_h", UNITS, DTK_TZ_HOUR}, /* timezone hour units */
Phase 2 of pgindent updates. Change pg_bsd_indent to follow upstream rules for placement of comments to the right of code, and remove pgindent hack that caused comments following #endif to not obey the general rule. Commit e3860ffa4dd0dad0dd9eea4be9cc1412373a8c89 wasn't actually using the published version of pg_bsd_indent, but a hacked-up version that tried to minimize the amount of movement of comments to the right of code. The situation of interest is where such a comment has to be moved to the right of its default placement at column 33 because there's code there. BSD indent has always moved right in units of tab stops in such cases --- but in the previous incarnation, indent was working in 8-space tab stops, while now it knows we use 4-space tabs. So the net result is that in about half the cases, such comments are placed one tab stop left of before. This is better all around: it leaves more room on the line for comment text, and it means that in such cases the comment uniformly starts at the next 4-space tab stop after the code, rather than sometimes one and sometimes two tabs after. Also, ensure that comments following #endif are indented the same as comments following other preprocessor commands such as #else. That inconsistency turns out to have been self-inflicted damage from a poorly-thought-through post-indent "fixup" in pgindent. This patch is much less interesting than the first round of indent changes, but also bulkier, so I thought it best to separate the effects. Discussion: https://postgr.es/m/E1dAmxK-0006EE-1r@gemulon.postgresql.org Discussion: https://postgr.es/m/30527.1495162840@sss.pgh.pa.us
2017-06-21 21:18:54 +02:00
{"timezone_m", UNITS, DTK_TZ_MINUTE}, /* timezone minutes units */
pgindent run.
2002-09-04 22:31:48 +02:00
{"us", UNITS, DTK_MICROSEC}, /* "microsecond" relative */
Phase 2 of pgindent updates. Change pg_bsd_indent to follow upstream rules for placement of comments to the right of code, and remove pgindent hack that caused comments following #endif to not obey the general rule. Commit e3860ffa4dd0dad0dd9eea4be9cc1412373a8c89 wasn't actually using the published version of pg_bsd_indent, but a hacked-up version that tried to minimize the amount of movement of comments to the right of code. The situation of interest is where such a comment has to be moved to the right of its default placement at column 33 because there's code there. BSD indent has always moved right in units of tab stops in such cases --- but in the previous incarnation, indent was working in 8-space tab stops, while now it knows we use 4-space tabs. So the net result is that in about half the cases, such comments are placed one tab stop left of before. This is better all around: it leaves more room on the line for comment text, and it means that in such cases the comment uniformly starts at the next 4-space tab stop after the code, rather than sometimes one and sometimes two tabs after. Also, ensure that comments following #endif are indented the same as comments following other preprocessor commands such as #else. That inconsistency turns out to have been self-inflicted damage from a poorly-thought-through post-indent "fixup" in pgindent. This patch is much less interesting than the first round of indent changes, but also bulkier, so I thought it best to separate the effects. Discussion: https://postgr.es/m/E1dAmxK-0006EE-1r@gemulon.postgresql.org Discussion: https://postgr.es/m/30527.1495162840@sss.pgh.pa.us
2017-06-21 21:18:54 +02:00
{"usec", UNITS, DTK_MICROSEC}, /* "microsecond" relative */
Accept an INTERVAL argument for SET TIME ZONE per SQL99. Modified the parser and the SET handlers to use full Node structures rather than simply a character string argument. Implement INTERVAL() YEAR TO MONTH (etc) syntax per SQL99. Does not yet accept the goofy string format that goes along with, but this should be fairly straight forward to fix now as a bug or later as a feature. Implement precision for the INTERVAL() type. Use the typmod mechanism for both of INTERVAL features. Fix the INTERVAL syntax in the parser: opt_interval was in the wrong place. INTERVAL is now a reserved word, otherwise we get reduce/reduce errors. Implement an explicit date_part() function for TIMETZ. Should fix coersion problem with INTERVAL reported by Peter E. Fix up some error messages for date/time types. Use all caps for type names within message. Fix recently introduced side-effect bug disabling 'epoch' as a recognized field for date_part() etc. Reported by Peter E. (??) Bump catalog version number. Rename "microseconds" current transaction time field from ...Msec to ...Usec. Duh! date/time regression tests updated for reference platform, but a few changes will be necessary for others.
2001-10-18 19:30:21 +02:00
{DMICROSEC, UNITS, DTK_MICROSEC}, /* "microsecond" relative */
{"useconds", UNITS, DTK_MICROSEC}, /* "microseconds" relative */
Phase 2 of pgindent updates. Change pg_bsd_indent to follow upstream rules for placement of comments to the right of code, and remove pgindent hack that caused comments following #endif to not obey the general rule. Commit e3860ffa4dd0dad0dd9eea4be9cc1412373a8c89 wasn't actually using the published version of pg_bsd_indent, but a hacked-up version that tried to minimize the amount of movement of comments to the right of code. The situation of interest is where such a comment has to be moved to the right of its default placement at column 33 because there's code there. BSD indent has always moved right in units of tab stops in such cases --- but in the previous incarnation, indent was working in 8-space tab stops, while now it knows we use 4-space tabs. So the net result is that in about half the cases, such comments are placed one tab stop left of before. This is better all around: it leaves more room on the line for comment text, and it means that in such cases the comment uniformly starts at the next 4-space tab stop after the code, rather than sometimes one and sometimes two tabs after. Also, ensure that comments following #endif are indented the same as comments following other preprocessor commands such as #else. That inconsistency turns out to have been self-inflicted damage from a poorly-thought-through post-indent "fixup" in pgindent. This patch is much less interesting than the first round of indent changes, but also bulkier, so I thought it best to separate the effects. Discussion: https://postgr.es/m/E1dAmxK-0006EE-1r@gemulon.postgresql.org Discussion: https://postgr.es/m/30527.1495162840@sss.pgh.pa.us
2017-06-21 21:18:54 +02:00
{"usecs", UNITS, DTK_MICROSEC}, /* "microseconds" relative */
Accept an INTERVAL argument for SET TIME ZONE per SQL99. Modified the parser and the SET handlers to use full Node structures rather than simply a character string argument. Implement INTERVAL() YEAR TO MONTH (etc) syntax per SQL99. Does not yet accept the goofy string format that goes along with, but this should be fairly straight forward to fix now as a bug or later as a feature. Implement precision for the INTERVAL() type. Use the typmod mechanism for both of INTERVAL features. Fix the INTERVAL syntax in the parser: opt_interval was in the wrong place. INTERVAL is now a reserved word, otherwise we get reduce/reduce errors. Implement an explicit date_part() function for TIMETZ. Should fix coersion problem with INTERVAL reported by Peter E. Fix up some error messages for date/time types. Use all caps for type names within message. Fix recently introduced side-effect bug disabling 'epoch' as a recognized field for date_part() etc. Reported by Peter E. (??) Bump catalog version number. Rename "microseconds" current transaction time field from ...Msec to ...Usec. Duh! date/time regression tests updated for reference platform, but a few changes will be necessary for others.
2001-10-18 19:30:21 +02:00
{"w", UNITS, DTK_WEEK}, /* "week" relative */
{DWEEK, UNITS, DTK_WEEK}, /* "week" relative */
{"weeks", UNITS, DTK_WEEK}, /* "weeks" relative */
{"y", UNITS, DTK_YEAR}, /* "year" relative */
{DYEAR, UNITS, DTK_YEAR}, /* "year" relative */
{"years", UNITS, DTK_YEAR}, /* "years" relative */
{"yr", UNITS, DTK_YEAR}, /* "year" relative */
Fix up some problems in handling of zic-style time zone names in datetime input routines. Remove the former "DecodePosixTimezone" function in favor of letting the zic code handle POSIX-style zone specs (see tzparse()). In particular this means that "PST+3" now means the same as "-03", whereas it used to mean "-11" --- the zone abbreviation is effectively just a noise word in this syntax. Make sure that all named and POSIX-style zone names will be parsed as a single token. Fix long-standing bogosities in printing and input of fractional-hour timezone offsets (since the tzparse() code will accept these, we'd better make 'em work). Also correct an error in the original coding of the zic-zone-name patch: in "timestamp without time zone" input, zone names are supposed to be allowed but ignored, but the coding was such that the zone changed the interpretation anyway.
2006-10-17 23:03:21 +02:00
{"yrs", UNITS, DTK_YEAR} /* "years" relative */
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
};
Const-ify a few more large static tables. Per research by Andres. Discussion: https://postgr.es/m/20181015200754.7y7zfuzsoux2c4ya@alap3.anarazel.de
2018-10-17 21:32:47 +02:00
static const int szdeltatktbl = sizeof deltatktbl / sizeof deltatktbl[0];
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
static TimeZoneAbbrevTable *zoneabbrevtbl = NULL;
/* Caches of recent lookup results in the above tables */
Remove hard-wired lists of timezone abbreviations in favor of providing configuration files that can be altered by a DBA. The australian_timezones GUC setting disappears, replaced by a timezone_abbreviations setting (set this to 'Australia' to get the effect of australian_timezones). The list of zone names defined by default has undergone a bit of cleanup, too. Documentation still needs some work --- in particular, should we fix Table B-4, or just get rid of it? Joachim Wieland, with some editorializing by moi.
2006-07-25 05:51:23 +02:00
static const datetkn *datecache[MAXDATEFIELDS] = {NULL};
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
Remove hard-wired lists of timezone abbreviations in favor of providing configuration files that can be altered by a DBA. The australian_timezones GUC setting disappears, replaced by a timezone_abbreviations setting (set this to 'Australia' to get the effect of australian_timezones). The list of zone names defined by default has undergone a bit of cleanup, too. Documentation still needs some work --- in particular, should we fix Table B-4, or just get rid of it? Joachim Wieland, with some editorializing by moi.
2006-07-25 05:51:23 +02:00
static const datetkn *deltacache[MAXDATEFIELDS] = {NULL};
Date/Time updates from Thomas...
1997-03-15 00:21:12 +01:00
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
static const datetkn *abbrevcache[MAXDATEFIELDS] = {NULL};
Date/Time updates from Thomas...
1997-03-15 00:21:12 +01:00
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
/*
* Calendar time to Julian date conversions.
* Julian date is commonly used in astronomical applications,
* since it is numerically accurate and computationally simple.
* The algorithms here will accurately convert between Julian day
* and calendar date for all non-negative Julian days
The following patches eliminate the overflows in the j2date() and date2j() functions which limited the maximum date for a timestamp to AD 1465001. The new limit is AD 5874897. The files affected are: doc/src/sgml/datatype.sgml: Documentation change due to patch. Included is a notice about the reduced range when using an eight-byte integer for timestamps. src/backend/utils/adt/datetime.c: Replacement functions for j2date() and date2j() functions. src/include/utils/datetime.h: Corrected a bug with the limit on the earliest possible date, Nov 23,-4713 has a Julian day count of -1. The earliest possible date should be Nov 24, -4713 with a day count of 0. src/test/regress/expected/horology-no-DST-before-1970.out: src/test/regress/expected/horology-solaris-1947.out: src/test/regress/expected/horology.out: Copies of expected output for regression testing. Note: Only horology.out has been physically tested. I do not have access to a Solaris box and I don't know how to provoke the "pre-1970" test. src/test/regress/sql/horology.sql: Added some test cases to check extended range. John Cochran
2003-02-19 04:48:11 +01:00
* (i.e. from Nov 24, -4713 on).
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
*
The following patches eliminate the overflows in the j2date() and date2j() functions which limited the maximum date for a timestamp to AD 1465001. The new limit is AD 5874897. The files affected are: doc/src/sgml/datatype.sgml: Documentation change due to patch. Included is a notice about the reduced range when using an eight-byte integer for timestamps. src/backend/utils/adt/datetime.c: Replacement functions for j2date() and date2j() functions. src/include/utils/datetime.h: Corrected a bug with the limit on the earliest possible date, Nov 23,-4713 has a Julian day count of -1. The earliest possible date should be Nov 24, -4713 with a day count of 0. src/test/regress/expected/horology-no-DST-before-1970.out: src/test/regress/expected/horology-solaris-1947.out: src/test/regress/expected/horology.out: Copies of expected output for regression testing. Note: Only horology.out has been physically tested. I do not have access to a Solaris box and I don't know how to provoke the "pre-1970" test. src/test/regress/sql/horology.sql: Added some test cases to check extended range. John Cochran
2003-02-19 04:48:11 +01:00
* Rewritten to eliminate overflow problems. This now allows the
* routines to work correctly for all Julian day counts from
pgindent run.
2003-08-04 02:43:34 +02:00
* 0 to 2147483647 (Nov 24, -4713 to Jun 3, 5874898) assuming
The following patches eliminate the overflows in the j2date() and date2j() functions which limited the maximum date for a timestamp to AD 1465001. The new limit is AD 5874897. The files affected are: doc/src/sgml/datatype.sgml: Documentation change due to patch. Included is a notice about the reduced range when using an eight-byte integer for timestamps. src/backend/utils/adt/datetime.c: Replacement functions for j2date() and date2j() functions. src/include/utils/datetime.h: Corrected a bug with the limit on the earliest possible date, Nov 23,-4713 has a Julian day count of -1. The earliest possible date should be Nov 24, -4713 with a day count of 0. src/test/regress/expected/horology-no-DST-before-1970.out: src/test/regress/expected/horology-solaris-1947.out: src/test/regress/expected/horology.out: Copies of expected output for regression testing. Note: Only horology.out has been physically tested. I do not have access to a Solaris box and I don't know how to provoke the "pre-1970" test. src/test/regress/sql/horology.sql: Added some test cases to check extended range. John Cochran
2003-02-19 04:48:11 +01:00
* a 32-bit integer. Longer types should also work to the limits
* of their precision.
Be more careful about out-of-range dates and timestamps. Tighten the semantics of boundary-case timestamptz so that we allow timestamps >= '4714-11-24 00:00+00 BC' and < 'ENDYEAR-01-01 00:00+00 AD' exactly, no more and no less, but it is allowed to enter timestamps within that range using non-GMT timezone offsets (which could make the nominal date 4714-11-23 BC or ENDYEAR-01-01 AD). This eliminates dump/reload failure conditions for timestamps near the endpoints. To do this, separate checking of the inputs for date2j() from the final range check, and allow the Julian date code to handle a range slightly wider than the nominal range of the datatypes. Also add a bunch of checks to detect out-of-range dates and timestamps that formerly could be returned by operations such as date-plus-integer. All C-level functions that return date, timestamp, or timestamptz should now be proof against returning a value that doesn't pass IS_VALID_DATE() or IS_VALID_TIMESTAMP(). Vitaly Burovoy, reviewed by Anastasia Lubennikova, and substantially whacked around by me
2016-03-17 00:09:04 +01:00
*
* Actually, date2j() will work sanely, in the sense of producing
* valid negative Julian dates, significantly before Nov 24, -4713.
* We rely on it to do so back to Nov 1, -4713; see IS_VALID_JULIAN()
* and associated commentary in timestamp.h.
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
*/
Date/Time updates from Thomas...
1997-03-15 00:21:12 +01:00
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
int
date2j(int y, int m, int d)
Date/Time updates from Thomas...
1997-03-15 00:21:12 +01:00
{
The following patches eliminate the overflows in the j2date() and date2j() functions which limited the maximum date for a timestamp to AD 1465001. The new limit is AD 5874897. The files affected are: doc/src/sgml/datatype.sgml: Documentation change due to patch. Included is a notice about the reduced range when using an eight-byte integer for timestamps. src/backend/utils/adt/datetime.c: Replacement functions for j2date() and date2j() functions. src/include/utils/datetime.h: Corrected a bug with the limit on the earliest possible date, Nov 23,-4713 has a Julian day count of -1. The earliest possible date should be Nov 24, -4713 with a day count of 0. src/test/regress/expected/horology-no-DST-before-1970.out: src/test/regress/expected/horology-solaris-1947.out: src/test/regress/expected/horology.out: Copies of expected output for regression testing. Note: Only horology.out has been physically tested. I do not have access to a Solaris box and I don't know how to provoke the "pre-1970" test. src/test/regress/sql/horology.sql: Added some test cases to check extended range. John Cochran
2003-02-19 04:48:11 +01:00
int julian;
int century;
pgindent run.
2003-08-04 02:43:34 +02:00
if (m > 2)
{
The following patches eliminate the overflows in the j2date() and date2j() functions which limited the maximum date for a timestamp to AD 1465001. The new limit is AD 5874897. The files affected are: doc/src/sgml/datatype.sgml: Documentation change due to patch. Included is a notice about the reduced range when using an eight-byte integer for timestamps. src/backend/utils/adt/datetime.c: Replacement functions for j2date() and date2j() functions. src/include/utils/datetime.h: Corrected a bug with the limit on the earliest possible date, Nov 23,-4713 has a Julian day count of -1. The earliest possible date should be Nov 24, -4713 with a day count of 0. src/test/regress/expected/horology-no-DST-before-1970.out: src/test/regress/expected/horology-solaris-1947.out: src/test/regress/expected/horology.out: Copies of expected output for regression testing. Note: Only horology.out has been physically tested. I do not have access to a Solaris box and I don't know how to provoke the "pre-1970" test. src/test/regress/sql/horology.sql: Added some test cases to check extended range. John Cochran
2003-02-19 04:48:11 +01:00
m += 1;
y += 4800;
pgindent run.
2003-08-04 02:43:34 +02:00
}
else
{
The following patches eliminate the overflows in the j2date() and date2j() functions which limited the maximum date for a timestamp to AD 1465001. The new limit is AD 5874897. The files affected are: doc/src/sgml/datatype.sgml: Documentation change due to patch. Included is a notice about the reduced range when using an eight-byte integer for timestamps. src/backend/utils/adt/datetime.c: Replacement functions for j2date() and date2j() functions. src/include/utils/datetime.h: Corrected a bug with the limit on the earliest possible date, Nov 23,-4713 has a Julian day count of -1. The earliest possible date should be Nov 24, -4713 with a day count of 0. src/test/regress/expected/horology-no-DST-before-1970.out: src/test/regress/expected/horology-solaris-1947.out: src/test/regress/expected/horology.out: Copies of expected output for regression testing. Note: Only horology.out has been physically tested. I do not have access to a Solaris box and I don't know how to provoke the "pre-1970" test. src/test/regress/sql/horology.sql: Added some test cases to check extended range. John Cochran
2003-02-19 04:48:11 +01:00
m += 13;
y += 4799;
}
pgindent run.
2003-08-04 02:43:34 +02:00
century = y / 100;
julian = y * 365 - 32167;
julian += y / 4 - century + century / 4;
julian += 7834 * m / 256 + d;
Date/Time updates from Thomas...
1997-03-15 00:21:12 +01:00
The following patches eliminate the overflows in the j2date() and date2j() functions which limited the maximum date for a timestamp to AD 1465001. The new limit is AD 5874897. The files affected are: doc/src/sgml/datatype.sgml: Documentation change due to patch. Included is a notice about the reduced range when using an eight-byte integer for timestamps. src/backend/utils/adt/datetime.c: Replacement functions for j2date() and date2j() functions. src/include/utils/datetime.h: Corrected a bug with the limit on the earliest possible date, Nov 23,-4713 has a Julian day count of -1. The earliest possible date should be Nov 24, -4713 with a day count of 0. src/test/regress/expected/horology-no-DST-before-1970.out: src/test/regress/expected/horology-solaris-1947.out: src/test/regress/expected/horology.out: Copies of expected output for regression testing. Note: Only horology.out has been physically tested. I do not have access to a Solaris box and I don't know how to provoke the "pre-1970" test. src/test/regress/sql/horology.sql: Added some test cases to check extended range. John Cochran
2003-02-19 04:48:11 +01:00
return julian;
Initial pgindent run with pg_bsd_indent version 2.0. The new indent version includes numerous fixes thanks to Piotr Stefaniak. The main changes visible in this commit are: * Nicer formatting of function-pointer declarations. * No longer unexpectedly removes spaces in expressions using casts, sizeof, or offsetof. * No longer wants to add a space in "struct structname *varname", as well as some similar cases for const- or volatile-qualified pointers. * Declarations using PG_USED_FOR_ASSERTS_ONLY are formatted more nicely. * Fixes bug where comments following declarations were sometimes placed with no space separating them from the code. * Fixes some odd decisions for comments following case labels. * Fixes some cases where comments following code were indented to less than the expected column 33. On the less good side, it now tends to put more whitespace around typedef names that are not listed in typedefs.list. This might encourage us to put more effort into typedef name collection; it's not really a bug in indent itself. There are more changes coming after this round, having to do with comment indentation and alignment of lines appearing within parentheses. I wanted to limit the size of the diffs to something that could be reviewed without one's eyes completely glazing over, so it seemed better to split up the changes as much as practical. Discussion: https://postgr.es/m/E1dAmxK-0006EE-1r@gemulon.postgresql.org Discussion: https://postgr.es/m/30527.1495162840@sss.pgh.pa.us
2017-06-21 20:39:04 +02:00
} /* date2j() */
Date/Time updates from Thomas...
1997-03-15 00:21:12 +01:00
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
void
j2date(int jd, int *year, int *month, int *day)
Date/Time updates from Thomas...
1997-03-15 00:21:12 +01:00
{
pgindent run.
2003-08-04 02:43:34 +02:00
unsigned int julian;
unsigned int quad;
unsigned int extra;
The following patches eliminate the overflows in the j2date() and date2j() functions which limited the maximum date for a timestamp to AD 1465001. The new limit is AD 5874897. The files affected are: doc/src/sgml/datatype.sgml: Documentation change due to patch. Included is a notice about the reduced range when using an eight-byte integer for timestamps. src/backend/utils/adt/datetime.c: Replacement functions for j2date() and date2j() functions. src/include/utils/datetime.h: Corrected a bug with the limit on the earliest possible date, Nov 23,-4713 has a Julian day count of -1. The earliest possible date should be Nov 24, -4713 with a day count of 0. src/test/regress/expected/horology-no-DST-before-1970.out: src/test/regress/expected/horology-solaris-1947.out: src/test/regress/expected/horology.out: Copies of expected output for regression testing. Note: Only horology.out has been physically tested. I do not have access to a Solaris box and I don't know how to provoke the "pre-1970" test. src/test/regress/sql/horology.sql: Added some test cases to check extended range. John Cochran
2003-02-19 04:48:11 +01:00
int y;
julian = jd;
julian += 32044;
pgindent run.
2003-08-04 02:43:34 +02:00
quad = julian / 146097;
extra = (julian - quad * 146097) * 4 + 3;
julian += 60 + quad * 3 + extra / 146097;
quad = julian / 1461;
julian -= quad * 1461;
The following patches eliminate the overflows in the j2date() and date2j() functions which limited the maximum date for a timestamp to AD 1465001. The new limit is AD 5874897. The files affected are: doc/src/sgml/datatype.sgml: Documentation change due to patch. Included is a notice about the reduced range when using an eight-byte integer for timestamps. src/backend/utils/adt/datetime.c: Replacement functions for j2date() and date2j() functions. src/include/utils/datetime.h: Corrected a bug with the limit on the earliest possible date, Nov 23,-4713 has a Julian day count of -1. The earliest possible date should be Nov 24, -4713 with a day count of 0. src/test/regress/expected/horology-no-DST-before-1970.out: src/test/regress/expected/horology-solaris-1947.out: src/test/regress/expected/horology.out: Copies of expected output for regression testing. Note: Only horology.out has been physically tested. I do not have access to a Solaris box and I don't know how to provoke the "pre-1970" test. src/test/regress/sql/horology.sql: Added some test cases to check extended range. John Cochran
2003-02-19 04:48:11 +01:00
y = julian * 4 / 1461;
julian = ((y != 0) ? ((julian + 305) % 365) : ((julian + 306) % 366))
+ 123;
pgindent run.
2003-08-04 02:43:34 +02:00
y += quad * 4;
The following patches eliminate the overflows in the j2date() and date2j() functions which limited the maximum date for a timestamp to AD 1465001. The new limit is AD 5874897. The files affected are: doc/src/sgml/datatype.sgml: Documentation change due to patch. Included is a notice about the reduced range when using an eight-byte integer for timestamps. src/backend/utils/adt/datetime.c: Replacement functions for j2date() and date2j() functions. src/include/utils/datetime.h: Corrected a bug with the limit on the earliest possible date, Nov 23,-4713 has a Julian day count of -1. The earliest possible date should be Nov 24, -4713 with a day count of 0. src/test/regress/expected/horology-no-DST-before-1970.out: src/test/regress/expected/horology-solaris-1947.out: src/test/regress/expected/horology.out: Copies of expected output for regression testing. Note: Only horology.out has been physically tested. I do not have access to a Solaris box and I don't know how to provoke the "pre-1970" test. src/test/regress/sql/horology.sql: Added some test cases to check extended range. John Cochran
2003-02-19 04:48:11 +01:00
*year = y - 4800;
quad = julian * 2141 / 65536;
pgindent run.
2003-08-04 02:43:34 +02:00
*day = julian - 7834 * quad / 256;
Use macros for time-based constants, rather than constants.
2011-03-12 15:31:18 +01:00
*month = (quad + 10) % MONTHS_PER_YEAR + 1;
Initial pgindent run with pg_bsd_indent version 2.0. The new indent version includes numerous fixes thanks to Piotr Stefaniak. The main changes visible in this commit are: * Nicer formatting of function-pointer declarations. * No longer unexpectedly removes spaces in expressions using casts, sizeof, or offsetof. * No longer wants to add a space in "struct structname *varname", as well as some similar cases for const- or volatile-qualified pointers. * Declarations using PG_USED_FOR_ASSERTS_ONLY are formatted more nicely. * Fixes bug where comments following declarations were sometimes placed with no space separating them from the code. * Fixes some odd decisions for comments following case labels. * Fixes some cases where comments following code were indented to less than the expected column 33. On the less good side, it now tends to put more whitespace around typedef names that are not listed in typedefs.list. This might encourage us to put more effort into typedef name collection; it's not really a bug in indent itself. There are more changes coming after this round, having to do with comment indentation and alignment of lines appearing within parentheses. I wanted to limit the size of the diffs to something that could be reviewed without one's eyes completely glazing over, so it seemed better to split up the changes as much as practical. Discussion: https://postgr.es/m/E1dAmxK-0006EE-1r@gemulon.postgresql.org Discussion: https://postgr.es/m/30527.1495162840@sss.pgh.pa.us
2017-06-21 20:39:04 +02:00
} /* j2date() */
Date/Time updates from Thomas...
1997-03-15 00:21:12 +01:00
Avoid repeated computation of the constants date2j(1970, 1, 1) and date2j(2000, 1, 1). Should make for some marginal speed improvement in date/time operations.
2003-04-04 06:50:44 +02:00
/*
* j2day - convert Julian date to day-of-week (0..6 == Sun..Sat)
*
* Note: various places use the locution j2day(date - 1) to produce a
pgindent run for 9.4 This includes removing tabs after periods in C comments, which was applied to back branches, so this change should not effect backpatching.
2014-05-06 18:12:18 +02:00
* result according to the convention 0..6 = Mon..Sun. This is a bit of
Avoid repeated computation of the constants date2j(1970, 1, 1) and date2j(2000, 1, 1). Should make for some marginal speed improvement in date/time operations.
2003-04-04 06:50:44 +02:00
* a crock, but will work as long as the computation here is just a modulo.
*/
Date/Time updates from Thomas...
1997-03-15 00:21:12 +01:00
int
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
j2day(int date)
Date/Time updates from Thomas...
1997-03-15 00:21:12 +01:00
{
Fix j2day() to behave sanely for negative Julian dates. Somebody had apparently once figured that casting to unsigned int would produce the right output for negative inputs, but that would only be true if 2^32 were a multiple of 7, which of course it ain't. We need to use a signed division and then correct the sign of the remainder. AFAICT, the only case where this would arise currently is when doing ISO-week calculations for dates in 4714BC, where we'd compute a negative Julian date representing 4714-01-04BC and then do some arithmetic with it. Since we don't even really document support for such dates, this is not of much consequence. But we may as well get it right. Per report from Vitaly Burovoy.
2016-03-17 01:57:45 +01:00
date += 1;
date %= 7;
/* Cope if division truncates towards zero, as it probably does */
if (date < 0)
date += 7;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
Fix j2day() to behave sanely for negative Julian dates. Somebody had apparently once figured that casting to unsigned int would produce the right output for negative inputs, but that would only be true if 2^32 were a multiple of 7, which of course it ain't. We need to use a signed division and then correct the sign of the remainder. AFAICT, the only case where this would arise currently is when doing ISO-week calculations for dates in 4714BC, where we'd compute a negative Julian date representing 4714-01-04BC and then do some arithmetic with it. Since we don't even really document support for such dates, this is not of much consequence. But we may as well get it right. Per report from Vitaly Burovoy.
2016-03-17 01:57:45 +01:00
return date;
Initial pgindent run with pg_bsd_indent version 2.0. The new indent version includes numerous fixes thanks to Piotr Stefaniak. The main changes visible in this commit are: * Nicer formatting of function-pointer declarations. * No longer unexpectedly removes spaces in expressions using casts, sizeof, or offsetof. * No longer wants to add a space in "struct structname *varname", as well as some similar cases for const- or volatile-qualified pointers. * Declarations using PG_USED_FOR_ASSERTS_ONLY are formatted more nicely. * Fixes bug where comments following declarations were sometimes placed with no space separating them from the code. * Fixes some odd decisions for comments following case labels. * Fixes some cases where comments following code were indented to less than the expected column 33. On the less good side, it now tends to put more whitespace around typedef names that are not listed in typedefs.list. This might encourage us to put more effort into typedef name collection; it's not really a bug in indent itself. There are more changes coming after this round, having to do with comment indentation and alignment of lines appearing within parentheses. I wanted to limit the size of the diffs to something that could be reviewed without one's eyes completely glazing over, so it seemed better to split up the changes as much as practical. Discussion: https://postgr.es/m/E1dAmxK-0006EE-1r@gemulon.postgresql.org Discussion: https://postgr.es/m/30527.1495162840@sss.pgh.pa.us
2017-06-21 20:39:04 +02:00
} /* j2day() */
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
Clean up the rather historically encumbered interface to now() and current time: provide a GetCurrentTimestamp() function that returns current time in the form of a TimestampTz, instead of separate time_t and microseconds fields. This is what all the callers really want anyway, and it eliminates low-level dependencies on AbsoluteTime, which is a deprecated datatype that will have to disappear eventually.
2005-06-30 00:51:57 +02:00
/*
* GetCurrentDateTime()
*
* Get the transaction start time ("now()") broken down as a struct pg_tm.
*/
void
Initial pgindent run with pg_bsd_indent version 2.0. The new indent version includes numerous fixes thanks to Piotr Stefaniak. The main changes visible in this commit are: * Nicer formatting of function-pointer declarations. * No longer unexpectedly removes spaces in expressions using casts, sizeof, or offsetof. * No longer wants to add a space in "struct structname *varname", as well as some similar cases for const- or volatile-qualified pointers. * Declarations using PG_USED_FOR_ASSERTS_ONLY are formatted more nicely. * Fixes bug where comments following declarations were sometimes placed with no space separating them from the code. * Fixes some odd decisions for comments following case labels. * Fixes some cases where comments following code were indented to less than the expected column 33. On the less good side, it now tends to put more whitespace around typedef names that are not listed in typedefs.list. This might encourage us to put more effort into typedef name collection; it's not really a bug in indent itself. There are more changes coming after this round, having to do with comment indentation and alignment of lines appearing within parentheses. I wanted to limit the size of the diffs to something that could be reviewed without one's eyes completely glazing over, so it seemed better to split up the changes as much as practical. Discussion: https://postgr.es/m/E1dAmxK-0006EE-1r@gemulon.postgresql.org Discussion: https://postgr.es/m/30527.1495162840@sss.pgh.pa.us
2017-06-21 20:39:04 +02:00
GetCurrentDateTime(struct pg_tm *tm)
Clean up the rather historically encumbered interface to now() and current time: provide a GetCurrentTimestamp() function that returns current time in the form of a TimestampTz, instead of separate time_t and microseconds fields. This is what all the callers really want anyway, and it eliminates low-level dependencies on AbsoluteTime, which is a deprecated datatype that will have to disappear eventually.
2005-06-30 00:51:57 +02:00
{
int tz;
fsec_t fsec;
timestamp2tm(GetCurrentTransactionStartTimestamp(), &tz, tm, &fsec,
NULL, NULL);
/* Note: don't pass NULL tzp to timestamp2tm; affects behavior */
}
/*
* GetCurrentTimeUsec()
*
* Get the transaction start time ("now()") broken down as a struct pg_tm,
* including fractional seconds and timezone offset.
*/
void
Initial pgindent run with pg_bsd_indent version 2.0. The new indent version includes numerous fixes thanks to Piotr Stefaniak. The main changes visible in this commit are: * Nicer formatting of function-pointer declarations. * No longer unexpectedly removes spaces in expressions using casts, sizeof, or offsetof. * No longer wants to add a space in "struct structname *varname", as well as some similar cases for const- or volatile-qualified pointers. * Declarations using PG_USED_FOR_ASSERTS_ONLY are formatted more nicely. * Fixes bug where comments following declarations were sometimes placed with no space separating them from the code. * Fixes some odd decisions for comments following case labels. * Fixes some cases where comments following code were indented to less than the expected column 33. On the less good side, it now tends to put more whitespace around typedef names that are not listed in typedefs.list. This might encourage us to put more effort into typedef name collection; it's not really a bug in indent itself. There are more changes coming after this round, having to do with comment indentation and alignment of lines appearing within parentheses. I wanted to limit the size of the diffs to something that could be reviewed without one's eyes completely glazing over, so it seemed better to split up the changes as much as practical. Discussion: https://postgr.es/m/E1dAmxK-0006EE-1r@gemulon.postgresql.org Discussion: https://postgr.es/m/30527.1495162840@sss.pgh.pa.us
2017-06-21 20:39:04 +02:00
GetCurrentTimeUsec(struct pg_tm *tm, fsec_t *fsec, int *tzp)
Clean up the rather historically encumbered interface to now() and current time: provide a GetCurrentTimestamp() function that returns current time in the form of a TimestampTz, instead of separate time_t and microseconds fields. This is what all the callers really want anyway, and it eliminates low-level dependencies on AbsoluteTime, which is a deprecated datatype that will have to disappear eventually.
2005-06-30 00:51:57 +02:00
{
int tz;
timestamp2tm(GetCurrentTransactionStartTimestamp(), &tz, tm, fsec,
NULL, NULL);
/* Note: don't pass NULL tzp to timestamp2tm; affects behavior */
if (tzp != NULL)
*tzp = tz;
}
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
/*
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
* Append seconds and fractional seconds (if any) at *cp.
*
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
* precision is the max number of fraction digits, fillzeros says to
* pad to two integral-seconds digits.
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
*
* Returns a pointer to the new end of string. No NUL terminator is put
* there; callers are responsible for NUL terminating str themselves.
*
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
* Note that any sign is stripped from the input seconds values.
*/
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
static char *
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
AppendSeconds(char *cp, int sec, fsec_t fsec, int precision, bool fillzeros)
{
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
Assert(precision >= 0);
if (fillzeros)
Optimizations for integer to decimal output. Using a lookup table of digit pairs reduces the number of divisions needed, and calculating the length upfront saves some work; these ideas are taken from the code previously committed for floats. David Fetter, reviewed by Kyotaro Horiguchi, Tels, and me. Discussion: https://postgr.es/m/20190924052620.GP31596%40fetter.org
2020-02-01 22:57:14 +01:00
cp = pg_ultostr_zeropad(cp, Abs(sec), 2);
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
else
Optimizations for integer to decimal output. Using a lookup table of digit pairs reduces the number of divisions needed, and calculating the length upfront saves some work; these ideas are taken from the code previously committed for floats. David Fetter, reviewed by Kyotaro Horiguchi, Tels, and me. Discussion: https://postgr.es/m/20190924052620.GP31596%40fetter.org
2020-02-01 22:57:14 +01:00
cp = pg_ultostr(cp, Abs(sec));
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
Remove now-dead code for !HAVE_INT64_TIMESTAMP. This is a basically mechanical removal of #ifdef HAVE_INT64_TIMESTAMP tests and the negative-case controlled code. Discussion: https://postgr.es/m/26788.1487455319@sss.pgh.pa.us
2017-02-23 20:04:43 +01:00
/* fsec_t is just an int32 */
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
if (fsec != 0)
{
int32 value = Abs(fsec);
char *end = &cp[precision + 1];
bool gotnonzero = false;
*cp++ = '.';
/*
* Append the fractional seconds part. Note that we don't want any
* trailing zeros here, so since we're building the number in reverse
* we'll skip appending zeros until we've output a non-zero digit.
*/
while (precision--)
{
int32 oldval = value;
int32 remainder;
value /= 10;
remainder = oldval - value * 10;
/* check if we got a non-zero */
if (remainder)
gotnonzero = true;
if (gotnonzero)
cp[precision] = '0' + remainder;
else
end = &cp[precision];
}
/*
* If we still have a non-zero value then precision must have not been
* enough to print the number. We punt the problem to pg_ltostr(),
* which will generate a correct answer in the minimum valid width.
*/
if (value)
Optimizations for integer to decimal output. Using a lookup table of digit pairs reduces the number of divisions needed, and calculating the length upfront saves some work; these ideas are taken from the code previously committed for floats. David Fetter, reviewed by Kyotaro Horiguchi, Tels, and me. Discussion: https://postgr.es/m/20190924052620.GP31596%40fetter.org
2020-02-01 22:57:14 +01:00
return pg_ultostr(cp, Abs(fsec));
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
return end;
}
else
return cp;
Accept an INTERVAL argument for SET TIME ZONE per SQL99. Modified the parser and the SET handlers to use full Node structures rather than simply a character string argument. Implement INTERVAL() YEAR TO MONTH (etc) syntax per SQL99. Does not yet accept the goofy string format that goes along with, but this should be fairly straight forward to fix now as a bug or later as a feature. Implement precision for the INTERVAL() type. Use the typmod mechanism for both of INTERVAL features. Fix the INTERVAL syntax in the parser: opt_interval was in the wrong place. INTERVAL is now a reserved word, otherwise we get reduce/reduce errors. Implement an explicit date_part() function for TIMETZ. Should fix coersion problem with INTERVAL reported by Peter E. Fix up some error messages for date/time types. Use all caps for type names within message. Fix recently introduced side-effect bug disabling 'epoch' as a recognized field for date_part() etc. Reported by Peter E. (??) Bump catalog version number. Rename "microseconds" current transaction time field from ...Msec to ...Usec. Duh! date/time regression tests updated for reference platform, but a few changes will be necessary for others.
2001-10-18 19:30:21 +02:00
}
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
/*
* Variant of above that's specialized to timestamp case.
*
* Returns a pointer to the new end of string. No NUL terminator is put
* there; callers are responsible for NUL terminating str themselves.
*/
static char *
Initial pgindent run with pg_bsd_indent version 2.0. The new indent version includes numerous fixes thanks to Piotr Stefaniak. The main changes visible in this commit are: * Nicer formatting of function-pointer declarations. * No longer unexpectedly removes spaces in expressions using casts, sizeof, or offsetof. * No longer wants to add a space in "struct structname *varname", as well as some similar cases for const- or volatile-qualified pointers. * Declarations using PG_USED_FOR_ASSERTS_ONLY are formatted more nicely. * Fixes bug where comments following declarations were sometimes placed with no space separating them from the code. * Fixes some odd decisions for comments following case labels. * Fixes some cases where comments following code were indented to less than the expected column 33. On the less good side, it now tends to put more whitespace around typedef names that are not listed in typedefs.list. This might encourage us to put more effort into typedef name collection; it's not really a bug in indent itself. There are more changes coming after this round, having to do with comment indentation and alignment of lines appearing within parentheses. I wanted to limit the size of the diffs to something that could be reviewed without one's eyes completely glazing over, so it seemed better to split up the changes as much as practical. Discussion: https://postgr.es/m/E1dAmxK-0006EE-1r@gemulon.postgresql.org Discussion: https://postgr.es/m/30527.1495162840@sss.pgh.pa.us
2017-06-21 20:39:04 +02:00
AppendTimestampSeconds(char *cp, struct pg_tm *tm, fsec_t fsec)
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
{
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
return AppendSeconds(cp, tm->tm_sec, fsec, MAX_TIMESTAMP_PRECISION, true);
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
}
/*
* Multiply frac by scale (to produce seconds) and add to *tm & *fsec.
* We assume the input frac is less than 1 so overflow is not an issue.
*/
static void
Initial pgindent run with pg_bsd_indent version 2.0. The new indent version includes numerous fixes thanks to Piotr Stefaniak. The main changes visible in this commit are: * Nicer formatting of function-pointer declarations. * No longer unexpectedly removes spaces in expressions using casts, sizeof, or offsetof. * No longer wants to add a space in "struct structname *varname", as well as some similar cases for const- or volatile-qualified pointers. * Declarations using PG_USED_FOR_ASSERTS_ONLY are formatted more nicely. * Fixes bug where comments following declarations were sometimes placed with no space separating them from the code. * Fixes some odd decisions for comments following case labels. * Fixes some cases where comments following code were indented to less than the expected column 33. On the less good side, it now tends to put more whitespace around typedef names that are not listed in typedefs.list. This might encourage us to put more effort into typedef name collection; it's not really a bug in indent itself. There are more changes coming after this round, having to do with comment indentation and alignment of lines appearing within parentheses. I wanted to limit the size of the diffs to something that could be reviewed without one's eyes completely glazing over, so it seemed better to split up the changes as much as practical. Discussion: https://postgr.es/m/E1dAmxK-0006EE-1r@gemulon.postgresql.org Discussion: https://postgr.es/m/30527.1495162840@sss.pgh.pa.us
2017-06-21 20:39:04 +02:00
AdjustFractSeconds(double frac, struct pg_tm *tm, fsec_t *fsec, int scale)
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
{
8.4 pgindent run, with new combined Linux/FreeBSD/MinGW typedef list provided by Andrew.
2009-06-11 16:49:15 +02:00
int sec;
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
if (frac == 0)
return;
8.4 pgindent run, with new combined Linux/FreeBSD/MinGW typedef list provided by Andrew.
2009-06-11 16:49:15 +02:00
frac *= scale;
sec = (int) frac;
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
tm->tm_sec += sec;
8.4 pgindent run, with new combined Linux/FreeBSD/MinGW typedef list provided by Andrew.
2009-06-11 16:49:15 +02:00
frac -= sec;
*fsec += rint(frac * 1000000);
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
}
/* As above, but initial scale produces days */
static void
Initial pgindent run with pg_bsd_indent version 2.0. The new indent version includes numerous fixes thanks to Piotr Stefaniak. The main changes visible in this commit are: * Nicer formatting of function-pointer declarations. * No longer unexpectedly removes spaces in expressions using casts, sizeof, or offsetof. * No longer wants to add a space in "struct structname *varname", as well as some similar cases for const- or volatile-qualified pointers. * Declarations using PG_USED_FOR_ASSERTS_ONLY are formatted more nicely. * Fixes bug where comments following declarations were sometimes placed with no space separating them from the code. * Fixes some odd decisions for comments following case labels. * Fixes some cases where comments following code were indented to less than the expected column 33. On the less good side, it now tends to put more whitespace around typedef names that are not listed in typedefs.list. This might encourage us to put more effort into typedef name collection; it's not really a bug in indent itself. There are more changes coming after this round, having to do with comment indentation and alignment of lines appearing within parentheses. I wanted to limit the size of the diffs to something that could be reviewed without one's eyes completely glazing over, so it seemed better to split up the changes as much as practical. Discussion: https://postgr.es/m/E1dAmxK-0006EE-1r@gemulon.postgresql.org Discussion: https://postgr.es/m/30527.1495162840@sss.pgh.pa.us
2017-06-21 20:39:04 +02:00
AdjustFractDays(double frac, struct pg_tm *tm, fsec_t *fsec, int scale)
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
{
8.4 pgindent run, with new combined Linux/FreeBSD/MinGW typedef list provided by Andrew.
2009-06-11 16:49:15 +02:00
int extra_days;
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
if (frac == 0)
return;
8.4 pgindent run, with new combined Linux/FreeBSD/MinGW typedef list provided by Andrew.
2009-06-11 16:49:15 +02:00
frac *= scale;
extra_days = (int) frac;
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
tm->tm_mday += extra_days;
8.4 pgindent run, with new combined Linux/FreeBSD/MinGW typedef list provided by Andrew.
2009-06-11 16:49:15 +02:00
frac -= extra_days;
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
AdjustFractSeconds(frac, tm, fsec, SECS_PER_DAY);
}
/* Fetch a fractional-second value with suitable error checking */
static int
ParseFractionalSecond(char *cp, fsec_t *fsec)
{
double frac;
/* Caller should always pass the start of the fraction part */
Assert(*cp == '.');
errno = 0;
frac = strtod(cp, &cp);
/* check for parse failure */
if (*cp != '\0' || errno != 0)
return DTERR_BAD_FORMAT;
*fsec = rint(frac * 1000000);
return 0;
}
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
/* ParseDateTime()
Improve documentation of ParseDateTime(). Reorder tests to prevent writing one more value into return arrays than will fit. This is potentially a stack smash, though I do not think it is a problem in current uses of the routine, since a failure return causes elog anyway.
2003-08-05 20:30:21 +02:00
* Break string into tokens based on a date/time context.
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
* Returns 0 if successful, DTERR code if bogus input detected.
Improve documentation of ParseDateTime(). Reorder tests to prevent writing one more value into return arrays than will fit. This is potentially a stack smash, though I do not think it is a problem in current uses of the routine, since a failure return causes elog anyway.
2003-08-05 20:30:21 +02:00
*
* timestr - the input string
Adjust datetime parsing to be more robust. We now pass the length of the working buffer into ParseDateTime() and reject too-long input there, rather than checking the length of the input string before calling ParseDateTime(). The old method was bogus because ParseDateTime() can use a variable amount of working space, depending on the content of the input string (e.g. how many fields need to be NUL terminated). This fixes a minor stack overrun -- I don't _think_ it's exploitable, although I won't claim to be an expert. Along the way, fix a bug reported by Mark Dilger: the working buffer allocated by interval_in() was too short, which resulted in rejecting some perfectly valid interval input values. I added a regression test for this fix.
2005-05-26 04:04:14 +02:00
* workbuf - workspace for field string storage. This must be
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* larger than the largest legal input for this datetime type --
* some additional space will be needed to NUL terminate fields.
Adjust datetime parsing to be more robust. We now pass the length of the working buffer into ParseDateTime() and reject too-long input there, rather than checking the length of the input string before calling ParseDateTime(). The old method was bogus because ParseDateTime() can use a variable amount of working space, depending on the content of the input string (e.g. how many fields need to be NUL terminated). This fixes a minor stack overrun -- I don't _think_ it's exploitable, although I won't claim to be an expert. Along the way, fix a bug reported by Mark Dilger: the working buffer allocated by interval_in() was too short, which resulted in rejecting some perfectly valid interval input values. I added a regression test for this fix.
2005-05-26 04:04:14 +02:00
* buflen - the size of workbuf
Improve documentation of ParseDateTime(). Reorder tests to prevent writing one more value into return arrays than will fit. This is potentially a stack smash, though I do not think it is a problem in current uses of the routine, since a failure return causes elog anyway.
2003-08-05 20:30:21 +02:00
* field[] - pointers to field strings are returned in this array
* ftype[] - field type indicators are returned in this array
* maxfields - dimensions of the above two arrays
* *numfields - set to the actual number of fields detected
*
Adjust datetime parsing to be more robust. We now pass the length of the working buffer into ParseDateTime() and reject too-long input there, rather than checking the length of the input string before calling ParseDateTime(). The old method was bogus because ParseDateTime() can use a variable amount of working space, depending on the content of the input string (e.g. how many fields need to be NUL terminated). This fixes a minor stack overrun -- I don't _think_ it's exploitable, although I won't claim to be an expert. Along the way, fix a bug reported by Mark Dilger: the working buffer allocated by interval_in() was too short, which resulted in rejecting some perfectly valid interval input values. I added a regression test for this fix.
2005-05-26 04:04:14 +02:00
* The fields extracted from the input are stored as separate,
* null-terminated strings in the workspace at workbuf. Any text is
* converted to lower case.
Improve documentation of ParseDateTime(). Reorder tests to prevent writing one more value into return arrays than will fit. This is potentially a stack smash, though I do not think it is a problem in current uses of the routine, since a failure return causes elog anyway.
2003-08-05 20:30:21 +02:00
*
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
* Several field types are assigned:
pgindent run.
2002-09-04 22:31:48 +02:00
* DTK_NUMBER - digits and (possibly) a decimal point
* DTK_DATE - digits and two delimiters, or digits and text
* DTK_TIME - digits, colon delimiters, and possibly a decimal point
Fix up some problems in handling of zic-style time zone names in datetime input routines. Remove the former "DecodePosixTimezone" function in favor of letting the zic code handle POSIX-style zone specs (see tzparse()). In particular this means that "PST+3" now means the same as "-03", whereas it used to mean "-11" --- the zone abbreviation is effectively just a noise word in this syntax. Make sure that all named and POSIX-style zone names will be parsed as a single token. Fix long-standing bogosities in printing and input of fractional-hour timezone offsets (since the tzparse() code will accept these, we'd better make 'em work). Also correct an error in the original coding of the zic-zone-name patch: in "timestamp without time zone" input, zone names are supposed to be allowed but ignored, but the coding was such that the zone changed the interpretation anyway.
2006-10-17 23:03:21 +02:00
* DTK_STRING - text (no digits or punctuation)
pgindent run.
2002-09-04 22:31:48 +02:00
* DTK_SPECIAL - leading "+" or "-" followed by text
Clean up a couple of weird corner cases in interval parsing: make -yyyy-mm be interpreted as expected (the sign should affect months too), and get rid of hard-wired assumption that unmarked signed values must be hours (if integers) or seconds (if floats). The former was just a bug in my previous patch, while the latter may have made sense at one time but seems illogical now that we support determination of the units from typmod information. Ron Mayer and myself.
2008-09-17 00:31:21 +02:00
* DTK_TZ - leading "+" or "-" followed by digits (also eats ':', '.', '-')
Improve documentation of ParseDateTime(). Reorder tests to prevent writing one more value into return arrays than will fit. This is potentially a stack smash, though I do not think it is a problem in current uses of the routine, since a failure return causes elog anyway.
2003-08-05 20:30:21 +02:00
*
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
* Note that some field types can hold unexpected items:
pgindent run.
2002-09-04 22:31:48 +02:00
* DTK_NUMBER can hold date fields (yy.ddd)
* DTK_STRING can hold months (January) and time zones (PST)
Fix up some problems in handling of zic-style time zone names in datetime input routines. Remove the former "DecodePosixTimezone" function in favor of letting the zic code handle POSIX-style zone specs (see tzparse()). In particular this means that "PST+3" now means the same as "-03", whereas it used to mean "-11" --- the zone abbreviation is effectively just a noise word in this syntax. Make sure that all named and POSIX-style zone names will be parsed as a single token. Fix long-standing bogosities in printing and input of fractional-hour timezone offsets (since the tzparse() code will accept these, we'd better make 'em work). Also correct an error in the original coding of the zic-zone-name patch: in "timestamp without time zone" input, zone names are supposed to be allowed but ignored, but the coding was such that the zone changed the interpretation anyway.
2006-10-17 23:03:21 +02:00
* DTK_DATE can hold time zone names (America/New_York, GMT-8)
Add conversion from datetime to time data type. Rename date+time conversion to datetime to ensure less than 16 characters in routine name (required to fit in pg_proc table).
1997-10-25 07:16:09 +02:00
*/
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
int
Adjust datetime parsing to be more robust. We now pass the length of the working buffer into ParseDateTime() and reject too-long input there, rather than checking the length of the input string before calling ParseDateTime(). The old method was bogus because ParseDateTime() can use a variable amount of working space, depending on the content of the input string (e.g. how many fields need to be NUL terminated). This fixes a minor stack overrun -- I don't _think_ it's exploitable, although I won't claim to be an expert. Along the way, fix a bug reported by Mark Dilger: the working buffer allocated by interval_in() was too short, which resulted in rejecting some perfectly valid interval input values. I added a regression test for this fix.
2005-05-26 04:04:14 +02:00
ParseDateTime(const char *timestr, char *workbuf, size_t buflen,
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
char **field, int *ftype, int maxfields, int *numfields)
Date/Time updates from Thomas...
1997-03-15 00:21:12 +01:00
{
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
int nf = 0;
Improve documentation of ParseDateTime(). Reorder tests to prevent writing one more value into return arrays than will fit. This is potentially a stack smash, though I do not think it is a problem in current uses of the routine, since a failure return causes elog anyway.
2003-08-05 20:30:21 +02:00
const char *cp = timestr;
Adjust datetime parsing to be more robust. We now pass the length of the working buffer into ParseDateTime() and reject too-long input there, rather than checking the length of the input string before calling ParseDateTime(). The old method was bogus because ParseDateTime() can use a variable amount of working space, depending on the content of the input string (e.g. how many fields need to be NUL terminated). This fixes a minor stack overrun -- I don't _think_ it's exploitable, although I won't claim to be an expert. Along the way, fix a bug reported by Mark Dilger: the working buffer allocated by interval_in() was too short, which resulted in rejecting some perfectly valid interval input values. I added a regression test for this fix.
2005-05-26 04:04:14 +02:00
char *bufp = workbuf;
const char *bufend = workbuf + buflen;
/*
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* Set the character pointed-to by "bufptr" to "newchar", and increment
* "bufptr". "end" gives the end of the buffer -- we return an error if
* there is no space left to append a character to the buffer. Note that
* "bufptr" is evaluated twice.
Adjust datetime parsing to be more robust. We now pass the length of the working buffer into ParseDateTime() and reject too-long input there, rather than checking the length of the input string before calling ParseDateTime(). The old method was bogus because ParseDateTime() can use a variable amount of working space, depending on the content of the input string (e.g. how many fields need to be NUL terminated). This fixes a minor stack overrun -- I don't _think_ it's exploitable, although I won't claim to be an expert. Along the way, fix a bug reported by Mark Dilger: the working buffer allocated by interval_in() was too short, which resulted in rejecting some perfectly valid interval input values. I added a regression test for this fix.
2005-05-26 04:04:14 +02:00
*/
#define APPEND_CHAR(bufptr, end, newchar) \
do \
{ \
if (((bufptr) + 1) >= (end)) \
return DTERR_BAD_FORMAT; \
*(bufptr)++ = newchar; \
} while (0)
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
/* outer loop through fields */
while (*cp != '\0')
{
Improve documentation of ParseDateTime(). Reorder tests to prevent writing one more value into return arrays than will fit. This is potentially a stack smash, though I do not think it is a problem in current uses of the routine, since a failure return causes elog anyway.
2003-08-05 20:30:21 +02:00
/* Ignore spaces between fields */
if (isspace((unsigned char) *cp))
{
cp++;
continue;
}
/* Record start of current field */
if (nf >= maxfields)
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
return DTERR_BAD_FORMAT;
Adjust datetime parsing to be more robust. We now pass the length of the working buffer into ParseDateTime() and reject too-long input there, rather than checking the length of the input string before calling ParseDateTime(). The old method was bogus because ParseDateTime() can use a variable amount of working space, depending on the content of the input string (e.g. how many fields need to be NUL terminated). This fixes a minor stack overrun -- I don't _think_ it's exploitable, although I won't claim to be an expert. Along the way, fix a bug reported by Mark Dilger: the working buffer allocated by interval_in() was too short, which resulted in rejecting some perfectly valid interval input values. I added a regression test for this fix.
2005-05-26 04:04:14 +02:00
field[nf] = bufp;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
/* leading digit? then date or time */
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
if (isdigit((unsigned char) *cp))
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
{
Adjust datetime parsing to be more robust. We now pass the length of the working buffer into ParseDateTime() and reject too-long input there, rather than checking the length of the input string before calling ParseDateTime(). The old method was bogus because ParseDateTime() can use a variable amount of working space, depending on the content of the input string (e.g. how many fields need to be NUL terminated). This fixes a minor stack overrun -- I don't _think_ it's exploitable, although I won't claim to be an expert. Along the way, fix a bug reported by Mark Dilger: the working buffer allocated by interval_in() was too short, which resulted in rejecting some perfectly valid interval input values. I added a regression test for this fix.
2005-05-26 04:04:14 +02:00
APPEND_CHAR(bufp, bufend, *cp++);
Ensure that all uses of <ctype.h> functions are applied to unsigned-char values, whether the local char type is signed or not. This is necessary for portability. Per discussion on pghackers around 9/16/00.
2000-12-03 21:45:40 +01:00
while (isdigit((unsigned char) *cp))
Adjust datetime parsing to be more robust. We now pass the length of the working buffer into ParseDateTime() and reject too-long input there, rather than checking the length of the input string before calling ParseDateTime(). The old method was bogus because ParseDateTime() can use a variable amount of working space, depending on the content of the input string (e.g. how many fields need to be NUL terminated). This fixes a minor stack overrun -- I don't _think_ it's exploitable, although I won't claim to be an expert. Along the way, fix a bug reported by Mark Dilger: the working buffer allocated by interval_in() was too short, which resulted in rejecting some perfectly valid interval input values. I added a regression test for this fix.
2005-05-26 04:04:14 +02:00
APPEND_CHAR(bufp, bufend, *cp++);
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
/* time field? */
if (*cp == ':')
{
ftype[nf] = DTK_TIME;
Adjust datetime parsing to be more robust. We now pass the length of the working buffer into ParseDateTime() and reject too-long input there, rather than checking the length of the input string before calling ParseDateTime(). The old method was bogus because ParseDateTime() can use a variable amount of working space, depending on the content of the input string (e.g. how many fields need to be NUL terminated). This fixes a minor stack overrun -- I don't _think_ it's exploitable, although I won't claim to be an expert. Along the way, fix a bug reported by Mark Dilger: the working buffer allocated by interval_in() was too short, which resulted in rejecting some perfectly valid interval input values. I added a regression test for this fix.
2005-05-26 04:04:14 +02:00
APPEND_CHAR(bufp, bufend, *cp++);
Ensure that all uses of <ctype.h> functions are applied to unsigned-char values, whether the local char type is signed or not. This is necessary for portability. Per discussion on pghackers around 9/16/00.
2000-12-03 21:45:40 +01:00
while (isdigit((unsigned char) *cp) ||
(*cp == ':') || (*cp == '.'))
Adjust datetime parsing to be more robust. We now pass the length of the working buffer into ParseDateTime() and reject too-long input there, rather than checking the length of the input string before calling ParseDateTime(). The old method was bogus because ParseDateTime() can use a variable amount of working space, depending on the content of the input string (e.g. how many fields need to be NUL terminated). This fixes a minor stack overrun -- I don't _think_ it's exploitable, although I won't claim to be an expert. Along the way, fix a bug reported by Mark Dilger: the working buffer allocated by interval_in() was too short, which resulted in rejecting some perfectly valid interval input values. I added a regression test for this fix.
2005-05-26 04:04:14 +02:00
APPEND_CHAR(bufp, bufend, *cp++);
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
}
/* date field? allow embedded text month */
Remove unnecessary parentheses in datetime/timestamp code.
2005-05-23 19:13:14 +02:00
else if (*cp == '-' || *cp == '/' || *cp == '.')
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
{
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
/* save delimiting character to use later */
Another pgindent run with updated typedefs.
2003-08-08 23:42:59 +02:00
char delim = *cp;
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
Adjust datetime parsing to be more robust. We now pass the length of the working buffer into ParseDateTime() and reject too-long input there, rather than checking the length of the input string before calling ParseDateTime(). The old method was bogus because ParseDateTime() can use a variable amount of working space, depending on the content of the input string (e.g. how many fields need to be NUL terminated). This fixes a minor stack overrun -- I don't _think_ it's exploitable, although I won't claim to be an expert. Along the way, fix a bug reported by Mark Dilger: the working buffer allocated by interval_in() was too short, which resulted in rejecting some perfectly valid interval input values. I added a regression test for this fix.
2005-05-26 04:04:14 +02:00
APPEND_CHAR(bufp, bufend, *cp++);
Measure the current transaction time to milliseconds. Define a new function, GetCurrentTransactionStartTimeUsec() to get the time to this precision. Allow now() and timestamp 'now' to use this higher precision result so we now have fractional seconds in this "constant". Add timestamp without time zone type. Move previous timestamp type to timestamp with time zone. Accept another ISO variant for date/time values: yyyy-mm-ddThh:mm:ss (note the "T" separating the day from hours information). Remove 'current' from date/time types; convert to 'now' in input. Separate time and timetz regression tests. Separate timestamp and timestamptz regression test.
2001-09-28 10:09:14 +02:00
/* second field is all digits? then no embedded text month */
if (isdigit((unsigned char) *cp))
{
Improve documentation of ParseDateTime(). Reorder tests to prevent writing one more value into return arrays than will fit. This is potentially a stack smash, though I do not think it is a problem in current uses of the routine, since a failure return causes elog anyway.
2003-08-05 20:30:21 +02:00
ftype[nf] = ((delim == '.') ? DTK_NUMBER : DTK_DATE);
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
while (isdigit((unsigned char) *cp))
Adjust datetime parsing to be more robust. We now pass the length of the working buffer into ParseDateTime() and reject too-long input there, rather than checking the length of the input string before calling ParseDateTime(). The old method was bogus because ParseDateTime() can use a variable amount of working space, depending on the content of the input string (e.g. how many fields need to be NUL terminated). This fixes a minor stack overrun -- I don't _think_ it's exploitable, although I won't claim to be an expert. Along the way, fix a bug reported by Mark Dilger: the working buffer allocated by interval_in() was too short, which resulted in rejecting some perfectly valid interval input values. I added a regression test for this fix.
2005-05-26 04:04:14 +02:00
APPEND_CHAR(bufp, bufend, *cp++);
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
pgindent run.
2002-09-04 22:31:48 +02:00
/*
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* insist that the delimiters match to get a three-field
* date.
pgindent run.
2002-09-04 22:31:48 +02:00
*/
Improve documentation of ParseDateTime(). Reorder tests to prevent writing one more value into return arrays than will fit. This is potentially a stack smash, though I do not think it is a problem in current uses of the routine, since a failure return causes elog anyway.
2003-08-05 20:30:21 +02:00
if (*cp == delim)
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
{
ftype[nf] = DTK_DATE;
Adjust datetime parsing to be more robust. We now pass the length of the working buffer into ParseDateTime() and reject too-long input there, rather than checking the length of the input string before calling ParseDateTime(). The old method was bogus because ParseDateTime() can use a variable amount of working space, depending on the content of the input string (e.g. how many fields need to be NUL terminated). This fixes a minor stack overrun -- I don't _think_ it's exploitable, although I won't claim to be an expert. Along the way, fix a bug reported by Mark Dilger: the working buffer allocated by interval_in() was too short, which resulted in rejecting some perfectly valid interval input values. I added a regression test for this fix.
2005-05-26 04:04:14 +02:00
APPEND_CHAR(bufp, bufend, *cp++);
Remove unnecessary parentheses in datetime/timestamp code.
2005-05-23 19:13:14 +02:00
while (isdigit((unsigned char) *cp) || *cp == delim)
Adjust datetime parsing to be more robust. We now pass the length of the working buffer into ParseDateTime() and reject too-long input there, rather than checking the length of the input string before calling ParseDateTime(). The old method was bogus because ParseDateTime() can use a variable amount of working space, depending on the content of the input string (e.g. how many fields need to be NUL terminated). This fixes a minor stack overrun -- I don't _think_ it's exploitable, although I won't claim to be an expert. Along the way, fix a bug reported by Mark Dilger: the working buffer allocated by interval_in() was too short, which resulted in rejecting some perfectly valid interval input values. I added a regression test for this fix.
2005-05-26 04:04:14 +02:00
APPEND_CHAR(bufp, bufend, *cp++);
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
}
Measure the current transaction time to milliseconds. Define a new function, GetCurrentTransactionStartTimeUsec() to get the time to this precision. Allow now() and timestamp 'now' to use this higher precision result so we now have fractional seconds in this "constant". Add timestamp without time zone type. Move previous timestamp type to timestamp with time zone. Accept another ISO variant for date/time values: yyyy-mm-ddThh:mm:ss (note the "T" separating the day from hours information). Remove 'current' from date/time types; convert to 'now' in input. Separate time and timetz regression tests. Separate timestamp and timestamptz regression test.
2001-09-28 10:09:14 +02:00
}
else
{
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
ftype[nf] = DTK_DATE;
Remove unnecessary parentheses in datetime/timestamp code.
2005-05-23 19:13:14 +02:00
while (isalnum((unsigned char) *cp) || *cp == delim)
Adjust datetime parsing to be more robust. We now pass the length of the working buffer into ParseDateTime() and reject too-long input there, rather than checking the length of the input string before calling ParseDateTime(). The old method was bogus because ParseDateTime() can use a variable amount of working space, depending on the content of the input string (e.g. how many fields need to be NUL terminated). This fixes a minor stack overrun -- I don't _think_ it's exploitable, although I won't claim to be an expert. Along the way, fix a bug reported by Mark Dilger: the working buffer allocated by interval_in() was too short, which resulted in rejecting some perfectly valid interval input values. I added a regression test for this fix.
2005-05-26 04:04:14 +02:00
APPEND_CHAR(bufp, bufend, pg_tolower((unsigned char) *cp++));
Measure the current transaction time to milliseconds. Define a new function, GetCurrentTransactionStartTimeUsec() to get the time to this precision. Allow now() and timestamp 'now' to use this higher precision result so we now have fractional seconds in this "constant". Add timestamp without time zone type. Move previous timestamp type to timestamp with time zone. Accept another ISO variant for date/time values: yyyy-mm-ddThh:mm:ss (note the "T" separating the day from hours information). Remove 'current' from date/time types; convert to 'now' in input. Separate time and timetz regression tests. Separate timestamp and timestamptz regression test.
2001-09-28 10:09:14 +02:00
}
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
}
pgindent run.
2002-09-04 22:31:48 +02:00
/*
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* otherwise, number only and will determine year, month, day, or
* concatenated fields later...
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
*/
else
ftype[nf] = DTK_NUMBER;
}
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
/* Leading decimal point? Then fractional seconds... */
else if (*cp == '.')
{
Adjust datetime parsing to be more robust. We now pass the length of the working buffer into ParseDateTime() and reject too-long input there, rather than checking the length of the input string before calling ParseDateTime(). The old method was bogus because ParseDateTime() can use a variable amount of working space, depending on the content of the input string (e.g. how many fields need to be NUL terminated). This fixes a minor stack overrun -- I don't _think_ it's exploitable, although I won't claim to be an expert. Along the way, fix a bug reported by Mark Dilger: the working buffer allocated by interval_in() was too short, which resulted in rejecting some perfectly valid interval input values. I added a regression test for this fix.
2005-05-26 04:04:14 +02:00
APPEND_CHAR(bufp, bufend, *cp++);
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
while (isdigit((unsigned char) *cp))
Adjust datetime parsing to be more robust. We now pass the length of the working buffer into ParseDateTime() and reject too-long input there, rather than checking the length of the input string before calling ParseDateTime(). The old method was bogus because ParseDateTime() can use a variable amount of working space, depending on the content of the input string (e.g. how many fields need to be NUL terminated). This fixes a minor stack overrun -- I don't _think_ it's exploitable, although I won't claim to be an expert. Along the way, fix a bug reported by Mark Dilger: the working buffer allocated by interval_in() was too short, which resulted in rejecting some perfectly valid interval input values. I added a regression test for this fix.
2005-05-26 04:04:14 +02:00
APPEND_CHAR(bufp, bufend, *cp++);
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
ftype[nf] = DTK_NUMBER;
}
pgindent run.
2002-09-04 22:31:48 +02:00
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
/*
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* text? then date string, month, day of week, special, or timezone
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
*/
Ensure that all uses of <ctype.h> functions are applied to unsigned-char values, whether the local char type is signed or not. This is necessary for portability. Per discussion on pghackers around 9/16/00.
2000-12-03 21:45:40 +01:00
else if (isalpha((unsigned char) *cp))
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
{
pgindent run for 8.3.
2007-11-15 22:14:46 +01:00
bool is_date;
Fix up some problems in handling of zic-style time zone names in datetime input routines. Remove the former "DecodePosixTimezone" function in favor of letting the zic code handle POSIX-style zone specs (see tzparse()). In particular this means that "PST+3" now means the same as "-03", whereas it used to mean "-11" --- the zone abbreviation is effectively just a noise word in this syntax. Make sure that all named and POSIX-style zone names will be parsed as a single token. Fix long-standing bogosities in printing and input of fractional-hour timezone offsets (since the tzparse() code will accept these, we'd better make 'em work). Also correct an error in the original coding of the zic-zone-name patch: in "timestamp without time zone" input, zone names are supposed to be allowed but ignored, but the coding was such that the zone changed the interpretation anyway.
2006-10-17 23:03:21 +02:00
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
ftype[nf] = DTK_STRING;
Adjust datetime parsing to be more robust. We now pass the length of the working buffer into ParseDateTime() and reject too-long input there, rather than checking the length of the input string before calling ParseDateTime(). The old method was bogus because ParseDateTime() can use a variable amount of working space, depending on the content of the input string (e.g. how many fields need to be NUL terminated). This fixes a minor stack overrun -- I don't _think_ it's exploitable, although I won't claim to be an expert. Along the way, fix a bug reported by Mark Dilger: the working buffer allocated by interval_in() was too short, which resulted in rejecting some perfectly valid interval input values. I added a regression test for this fix.
2005-05-26 04:04:14 +02:00
APPEND_CHAR(bufp, bufend, pg_tolower((unsigned char) *cp++));
Ensure that all uses of <ctype.h> functions are applied to unsigned-char values, whether the local char type is signed or not. This is necessary for portability. Per discussion on pghackers around 9/16/00.
2000-12-03 21:45:40 +01:00
while (isalpha((unsigned char) *cp))
Adjust datetime parsing to be more robust. We now pass the length of the working buffer into ParseDateTime() and reject too-long input there, rather than checking the length of the input string before calling ParseDateTime(). The old method was bogus because ParseDateTime() can use a variable amount of working space, depending on the content of the input string (e.g. how many fields need to be NUL terminated). This fixes a minor stack overrun -- I don't _think_ it's exploitable, although I won't claim to be an expert. Along the way, fix a bug reported by Mark Dilger: the working buffer allocated by interval_in() was too short, which resulted in rejecting some perfectly valid interval input values. I added a regression test for this fix.
2005-05-26 04:04:14 +02:00
APPEND_CHAR(bufp, bufend, pg_tolower((unsigned char) *cp++));
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
Ye-old pgindent run. Same 4-space tabs.
2000-04-12 19:17:23 +02:00
/*
Fix up some problems in handling of zic-style time zone names in datetime input routines. Remove the former "DecodePosixTimezone" function in favor of letting the zic code handle POSIX-style zone specs (see tzparse()). In particular this means that "PST+3" now means the same as "-03", whereas it used to mean "-11" --- the zone abbreviation is effectively just a noise word in this syntax. Make sure that all named and POSIX-style zone names will be parsed as a single token. Fix long-standing bogosities in printing and input of fractional-hour timezone offsets (since the tzparse() code will accept these, we'd better make 'em work). Also correct an error in the original coding of the zic-zone-name patch: in "timestamp without time zone" input, zone names are supposed to be allowed but ignored, but the coding was such that the zone changed the interpretation anyway.
2006-10-17 23:03:21 +02:00
* Dates can have embedded '-', '/', or '.' separators. It could
pgindent run for 8.3.
2007-11-15 22:14:46 +01:00
* also be a timezone name containing embedded '/', '+', '-', '_',
* or ':' (but '_' or ':' can't be the first punctuation). If the
* next character is a digit or '+', we need to check whether what
* we have so far is a recognized non-timezone keyword --- if so,
* don't believe that this is the start of a timezone.
Support full POSIX-style time zone: EST+3, PST-3, etc. We probably support a superset of the spec, but I don't have the spec to confirm this. Update regression tests to include tests for this format.
2000-03-16 15:38:42 +01:00
*/
Fix up some problems in handling of zic-style time zone names in datetime input routines. Remove the former "DecodePosixTimezone" function in favor of letting the zic code handle POSIX-style zone specs (see tzparse()). In particular this means that "PST+3" now means the same as "-03", whereas it used to mean "-11" --- the zone abbreviation is effectively just a noise word in this syntax. Make sure that all named and POSIX-style zone names will be parsed as a single token. Fix long-standing bogosities in printing and input of fractional-hour timezone offsets (since the tzparse() code will accept these, we'd better make 'em work). Also correct an error in the original coding of the zic-zone-name patch: in "timestamp without time zone" input, zone names are supposed to be allowed but ignored, but the coding was such that the zone changed the interpretation anyway.
2006-10-17 23:03:21 +02:00
is_date = false;
Remove unnecessary parentheses in datetime/timestamp code.
2005-05-23 19:13:14 +02:00
if (*cp == '-' || *cp == '/' || *cp == '.')
Fix up some problems in handling of zic-style time zone names in datetime input routines. Remove the former "DecodePosixTimezone" function in favor of letting the zic code handle POSIX-style zone specs (see tzparse()). In particular this means that "PST+3" now means the same as "-03", whereas it used to mean "-11" --- the zone abbreviation is effectively just a noise word in this syntax. Make sure that all named and POSIX-style zone names will be parsed as a single token. Fix long-standing bogosities in printing and input of fractional-hour timezone offsets (since the tzparse() code will accept these, we'd better make 'em work). Also correct an error in the original coding of the zic-zone-name patch: in "timestamp without time zone" input, zone names are supposed to be allowed but ignored, but the coding was such that the zone changed the interpretation anyway.
2006-10-17 23:03:21 +02:00
is_date = true;
else if (*cp == '+' || isdigit((unsigned char) *cp))
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
{
Fix up some problems in handling of zic-style time zone names in datetime input routines. Remove the former "DecodePosixTimezone" function in favor of letting the zic code handle POSIX-style zone specs (see tzparse()). In particular this means that "PST+3" now means the same as "-03", whereas it used to mean "-11" --- the zone abbreviation is effectively just a noise word in this syntax. Make sure that all named and POSIX-style zone names will be parsed as a single token. Fix long-standing bogosities in printing and input of fractional-hour timezone offsets (since the tzparse() code will accept these, we'd better make 'em work). Also correct an error in the original coding of the zic-zone-name patch: in "timestamp without time zone" input, zone names are supposed to be allowed but ignored, but the coding was such that the zone changed the interpretation anyway.
2006-10-17 23:03:21 +02:00
*bufp = '\0'; /* null-terminate current field value */
/* we need search only the core token table, not TZ names */
if (datebsearch(field[nf], datetktbl, szdatetktbl) == NULL)
is_date = true;
}
if (is_date)
{
ftype[nf] = DTK_DATE;
do
Allow timezone names in SQL strings, '2006-05-24 21:11 Americas/New_York'::timestamptz Joachim Wieland
2006-06-08 00:32:31 +02:00
{
Fix up some problems in handling of zic-style time zone names in datetime input routines. Remove the former "DecodePosixTimezone" function in favor of letting the zic code handle POSIX-style zone specs (see tzparse()). In particular this means that "PST+3" now means the same as "-03", whereas it used to mean "-11" --- the zone abbreviation is effectively just a noise word in this syntax. Make sure that all named and POSIX-style zone names will be parsed as a single token. Fix long-standing bogosities in printing and input of fractional-hour timezone offsets (since the tzparse() code will accept these, we'd better make 'em work). Also correct an error in the original coding of the zic-zone-name patch: in "timestamp without time zone" input, zone names are supposed to be allowed but ignored, but the coding was such that the zone changed the interpretation anyway.
2006-10-17 23:03:21 +02:00
APPEND_CHAR(bufp, bufend, pg_tolower((unsigned char) *cp++));
} while (*cp == '+' || *cp == '-' ||
*cp == '/' || *cp == '_' ||
*cp == '.' || *cp == ':' ||
isalnum((unsigned char) *cp));
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
}
}
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
/* sign? then special or numeric timezone */
Remove unnecessary parentheses in datetime/timestamp code.
2005-05-23 19:13:14 +02:00
else if (*cp == '+' || *cp == '-')
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
{
Adjust datetime parsing to be more robust. We now pass the length of the working buffer into ParseDateTime() and reject too-long input there, rather than checking the length of the input string before calling ParseDateTime(). The old method was bogus because ParseDateTime() can use a variable amount of working space, depending on the content of the input string (e.g. how many fields need to be NUL terminated). This fixes a minor stack overrun -- I don't _think_ it's exploitable, although I won't claim to be an expert. Along the way, fix a bug reported by Mark Dilger: the working buffer allocated by interval_in() was too short, which resulted in rejecting some perfectly valid interval input values. I added a regression test for this fix.
2005-05-26 04:04:14 +02:00
APPEND_CHAR(bufp, bufend, *cp++);
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
/* soak up leading whitespace */
Ensure that all uses of <ctype.h> functions are applied to unsigned-char values, whether the local char type is signed or not. This is necessary for portability. Per discussion on pghackers around 9/16/00.
2000-12-03 21:45:40 +01:00
while (isspace((unsigned char) *cp))
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
cp++;
/* numeric timezone? */
Clean up a couple of weird corner cases in interval parsing: make -yyyy-mm be interpreted as expected (the sign should affect months too), and get rid of hard-wired assumption that unmarked signed values must be hours (if integers) or seconds (if floats). The former was just a bug in my previous patch, while the latter may have made sense at one time but seems illogical now that we support determination of the units from typmod information. Ron Mayer and myself.
2008-09-17 00:31:21 +02:00
/* note that "DTK_TZ" could also be a signed float or yyyy-mm */
Ensure that all uses of <ctype.h> functions are applied to unsigned-char values, whether the local char type is signed or not. This is necessary for portability. Per discussion on pghackers around 9/16/00.
2000-12-03 21:45:40 +01:00
if (isdigit((unsigned char) *cp))
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
{
ftype[nf] = DTK_TZ;
Adjust datetime parsing to be more robust. We now pass the length of the working buffer into ParseDateTime() and reject too-long input there, rather than checking the length of the input string before calling ParseDateTime(). The old method was bogus because ParseDateTime() can use a variable amount of working space, depending on the content of the input string (e.g. how many fields need to be NUL terminated). This fixes a minor stack overrun -- I don't _think_ it's exploitable, although I won't claim to be an expert. Along the way, fix a bug reported by Mark Dilger: the working buffer allocated by interval_in() was too short, which resulted in rejecting some perfectly valid interval input values. I added a regression test for this fix.
2005-05-26 04:04:14 +02:00
APPEND_CHAR(bufp, bufend, *cp++);
Ensure that all uses of <ctype.h> functions are applied to unsigned-char values, whether the local char type is signed or not. This is necessary for portability. Per discussion on pghackers around 9/16/00.
2000-12-03 21:45:40 +01:00
while (isdigit((unsigned char) *cp) ||
Clean up a couple of weird corner cases in interval parsing: make -yyyy-mm be interpreted as expected (the sign should affect months too), and get rid of hard-wired assumption that unmarked signed values must be hours (if integers) or seconds (if floats). The former was just a bug in my previous patch, while the latter may have made sense at one time but seems illogical now that we support determination of the units from typmod information. Ron Mayer and myself.
2008-09-17 00:31:21 +02:00
*cp == ':' || *cp == '.' || *cp == '-')
Adjust datetime parsing to be more robust. We now pass the length of the working buffer into ParseDateTime() and reject too-long input there, rather than checking the length of the input string before calling ParseDateTime(). The old method was bogus because ParseDateTime() can use a variable amount of working space, depending on the content of the input string (e.g. how many fields need to be NUL terminated). This fixes a minor stack overrun -- I don't _think_ it's exploitable, although I won't claim to be an expert. Along the way, fix a bug reported by Mark Dilger: the working buffer allocated by interval_in() was too short, which resulted in rejecting some perfectly valid interval input values. I added a regression test for this fix.
2005-05-26 04:04:14 +02:00
APPEND_CHAR(bufp, bufend, *cp++);
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
}
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
/* special? */
Ensure that all uses of <ctype.h> functions are applied to unsigned-char values, whether the local char type is signed or not. This is necessary for portability. Per discussion on pghackers around 9/16/00.
2000-12-03 21:45:40 +01:00
else if (isalpha((unsigned char) *cp))
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
{
ftype[nf] = DTK_SPECIAL;
Adjust datetime parsing to be more robust. We now pass the length of the working buffer into ParseDateTime() and reject too-long input there, rather than checking the length of the input string before calling ParseDateTime(). The old method was bogus because ParseDateTime() can use a variable amount of working space, depending on the content of the input string (e.g. how many fields need to be NUL terminated). This fixes a minor stack overrun -- I don't _think_ it's exploitable, although I won't claim to be an expert. Along the way, fix a bug reported by Mark Dilger: the working buffer allocated by interval_in() was too short, which resulted in rejecting some perfectly valid interval input values. I added a regression test for this fix.
2005-05-26 04:04:14 +02:00
APPEND_CHAR(bufp, bufend, pg_tolower((unsigned char) *cp++));
Ensure that all uses of <ctype.h> functions are applied to unsigned-char values, whether the local char type is signed or not. This is necessary for portability. Per discussion on pghackers around 9/16/00.
2000-12-03 21:45:40 +01:00
while (isalpha((unsigned char) *cp))
Adjust datetime parsing to be more robust. We now pass the length of the working buffer into ParseDateTime() and reject too-long input there, rather than checking the length of the input string before calling ParseDateTime(). The old method was bogus because ParseDateTime() can use a variable amount of working space, depending on the content of the input string (e.g. how many fields need to be NUL terminated). This fixes a minor stack overrun -- I don't _think_ it's exploitable, although I won't claim to be an expert. Along the way, fix a bug reported by Mark Dilger: the working buffer allocated by interval_in() was too short, which resulted in rejecting some perfectly valid interval input values. I added a regression test for this fix.
2005-05-26 04:04:14 +02:00
APPEND_CHAR(bufp, bufend, pg_tolower((unsigned char) *cp++));
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
}
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
/* otherwise something wrong... */
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
else
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
return DTERR_BAD_FORMAT;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
}
Improve documentation of ParseDateTime(). Reorder tests to prevent writing one more value into return arrays than will fit. This is potentially a stack smash, though I do not think it is a problem in current uses of the routine, since a failure return causes elog anyway.
2003-08-05 20:30:21 +02:00
/* ignore other punctuation but use as delimiter */
Ensure that all uses of <ctype.h> functions are applied to unsigned-char values, whether the local char type is signed or not. This is necessary for portability. Per discussion on pghackers around 9/16/00.
2000-12-03 21:45:40 +01:00
else if (ispunct((unsigned char) *cp))
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
{
cp++;
continue;
}
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
/* otherwise, something is not right... */
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
else
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
return DTERR_BAD_FORMAT;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
/* force in a delimiter after each field */
Adjust datetime parsing to be more robust. We now pass the length of the working buffer into ParseDateTime() and reject too-long input there, rather than checking the length of the input string before calling ParseDateTime(). The old method was bogus because ParseDateTime() can use a variable amount of working space, depending on the content of the input string (e.g. how many fields need to be NUL terminated). This fixes a minor stack overrun -- I don't _think_ it's exploitable, although I won't claim to be an expert. Along the way, fix a bug reported by Mark Dilger: the working buffer allocated by interval_in() was too short, which resulted in rejecting some perfectly valid interval input values. I added a regression test for this fix.
2005-05-26 04:04:14 +02:00
*bufp++ = '\0';
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
nf++;
}
*numfields = nf;
Date/Time updates from Thomas...
1997-03-15 00:21:12 +01:00
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
return 0;
Improve documentation of ParseDateTime(). Reorder tests to prevent writing one more value into return arrays than will fit. This is potentially a stack smash, though I do not think it is a problem in current uses of the routine, since a failure return causes elog anyway.
2003-08-05 20:30:21 +02:00
}
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
/* DecodeDateTime()
* Interpret previously parsed fields for general date and time.
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
* Return 0 if full date, 1 if only time, and negative DTERR code if problems.
* (Currently, all callers treat 1 as an error return too.)
*
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
* External format(s):
* "<weekday> <month>-<day>-<year> <hour>:<minute>:<second>"
* "Fri Feb-7-1997 15:23:27"
* "Feb-7-1997 15:23:27"
* "2-7-1997 15:23:27"
* "1997-2-7 15:23:27"
* "1997.038 15:23:27" (day of year 1-366)
* Also supports input in compact time:
* "970207 152327"
* "97038 152327"
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
* "20011225T040506.789-07"
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
*
* Use the system-provided functions to get the current time zone
Replace time_t with pg_time_t (same values, but always int64) in on-disk data structures and backend internal APIs. This solves problems we've seen recently with inconsistent layout of pg_control between machines that have 32-bit time_t and those that have already migrated to 64-bit time_t. Also, we can get out from under the problem that Windows' Unix-API emulation is not consistent about the width of time_t. There are a few remaining places where local time_t variables are used to hold the current or recent result of time(NULL). I didn't bother changing these since they do not affect any cross-module APIs and surely all platforms will have 64-bit time_t before overflow becomes an actual risk. time_t should be avoided for anything visible to extension modules, however.
2008-02-17 03:09:32 +01:00
* if not specified in the input string.
*
* If the date is outside the range of pg_time_t (in practice that could only
* happen if pg_time_t is just 32 bits), then assume UTC time zone - thomas
* 1997-05-27
Add conversion from datetime to time data type. Rename date+time conversion to datetime to ensure less than 16 characters in routine name (required to fit in pg_proc table).
1997-10-25 07:16:09 +02:00
*/
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
int
DecodeDateTime(char **field, int *ftype, int nf,
Initial pgindent run with pg_bsd_indent version 2.0. The new indent version includes numerous fixes thanks to Piotr Stefaniak. The main changes visible in this commit are: * Nicer formatting of function-pointer declarations. * No longer unexpectedly removes spaces in expressions using casts, sizeof, or offsetof. * No longer wants to add a space in "struct structname *varname", as well as some similar cases for const- or volatile-qualified pointers. * Declarations using PG_USED_FOR_ASSERTS_ONLY are formatted more nicely. * Fixes bug where comments following declarations were sometimes placed with no space separating them from the code. * Fixes some odd decisions for comments following case labels. * Fixes some cases where comments following code were indented to less than the expected column 33. On the less good side, it now tends to put more whitespace around typedef names that are not listed in typedefs.list. This might encourage us to put more effort into typedef name collection; it's not really a bug in indent itself. There are more changes coming after this round, having to do with comment indentation and alignment of lines appearing within parentheses. I wanted to limit the size of the diffs to something that could be reviewed without one's eyes completely glazing over, so it seemed better to split up the changes as much as practical. Discussion: https://postgr.es/m/E1dAmxK-0006EE-1r@gemulon.postgresql.org Discussion: https://postgr.es/m/30527.1495162840@sss.pgh.pa.us
2017-06-21 20:39:04 +02:00
int *dtype, struct pg_tm *tm, fsec_t *fsec, int *tzp)
Date/Time updates from Thomas...
1997-03-15 00:21:12 +01:00
{
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
int fmask = 0,
tmask,
type;
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
int ptype = 0; /* "prefix type" for ISO y2001m02d04 format */
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
int i;
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
int val;
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
int dterr;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
int mer = HR24;
Change TRUE/FALSE to true/false The lower case spellings are C and C++ standard and are used in most parts of the PostgreSQL sources. The upper case spellings are only used in some files/modules. So standardize on the standard spellings. The APIs for ICU, Perl, and Windows define their own TRUE and FALSE, so those are left as is when using those APIs. In code comments, we use the lower-case spelling for the C concepts and keep the upper-case spelling for the SQL concepts. Reviewed-by: Michael Paquier <michael.paquier@gmail.com>
2017-08-16 06:22:32 +02:00
bool haveTextMonth = false;
bool isjulian = false;
bool is2digits = false;
bool bc = false;
Fix up some problems in handling of zic-style time zone names in datetime input routines. Remove the former "DecodePosixTimezone" function in favor of letting the zic code handle POSIX-style zone specs (see tzparse()). In particular this means that "PST+3" now means the same as "-03", whereas it used to mean "-11" --- the zone abbreviation is effectively just a noise word in this syntax. Make sure that all named and POSIX-style zone names will be parsed as a single token. Fix long-standing bogosities in printing and input of fractional-hour timezone offsets (since the tzparse() code will accept these, we'd better make 'em work). Also correct an error in the original coding of the zic-zone-name patch: in "timestamp without time zone" input, zone names are supposed to be allowed but ignored, but the coding was such that the zone changed the interpretation anyway.
2006-10-17 23:03:21 +02:00
pg_tz *namedTz = NULL;
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
pg_tz *abbrevTz = NULL;
pg_tz *valtz;
char *abbrev = NULL;
Fix parsing of time string followed by yesterday/today/tomorrow. Previously, 'yesterday 04:00:00'::timestamp didn't do the same thing as '04:00:00 yesterday'::timestamp, and the return value from the latter was midnight rather than the specified time. Dean Rasheed, with some stylistic changes
2011-08-30 17:34:29 +02:00
struct pg_tm cur_tm;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
/*
pgindent run on all C files. Java run to follow. initdb/regression tests pass.
2001-10-25 07:50:21 +02:00
* We'll insist on at least all of the date fields, but initialize the
* remaining fields in case they are not set later...
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
*/
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
*dtype = DTK_DATE;
tm->tm_hour = 0;
tm->tm_min = 0;
tm->tm_sec = 0;
*fsec = 0;
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
/* don't know daylight savings time status apriori */
tm->tm_isdst = -1;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
if (tzp != NULL)
*tzp = 0;
for (i = 0; i < nf; i++)
{
switch (ftype[i])
{
case DTK_DATE:
pgindent run for 9.4 This includes removing tabs after periods in C comments, which was applied to back branches, so this change should not effect backpatching.
2014-05-06 18:12:18 +02:00
Adjust C comments that would be wrap-able.
2013-10-02 01:44:53 +02:00
/*
pgindent run for 9.4 This includes removing tabs after periods in C comments, which was applied to back branches, so this change should not effect backpatching.
2014-05-06 18:12:18 +02:00
* Integral julian day with attached time zone? All other
* forms with JD will be separated into distinct fields, so we
* handle just this case here.
Adjust C comments that would be wrap-able.
2013-10-02 01:44:53 +02:00
*/
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
if (ptype == DTK_JULIAN)
Measure the current transaction time to milliseconds. Define a new function, GetCurrentTransactionStartTimeUsec() to get the time to this precision. Allow now() and timestamp 'now' to use this higher precision result so we now have fractional seconds in this "constant". Add timestamp without time zone type. Move previous timestamp type to timestamp with time zone. Accept another ISO variant for date/time values: yyyy-mm-ddThh:mm:ss (note the "T" separating the day from hours information). Remove 'current' from date/time types; convert to 'now' in input. Separate time and timetz regression tests. Separate timestamp and timestamptz regression test.
2001-09-28 10:09:14 +02:00
{
pgindent run on all C files. Java run to follow. initdb/regression tests pass.
2001-10-25 07:50:21 +02:00
char *cp;
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
int val;
Ye-old pgindent run. Same 4-space tabs.
2000-04-12 19:17:23 +02:00
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
if (tzp == NULL)
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
return DTERR_BAD_FORMAT;
Measure the current transaction time to milliseconds. Define a new function, GetCurrentTransactionStartTimeUsec() to get the time to this precision. Allow now() and timestamp 'now' to use this higher precision result so we now have fractional seconds in this "constant". Add timestamp without time zone type. Move previous timestamp type to timestamp with time zone. Accept another ISO variant for date/time values: yyyy-mm-ddThh:mm:ss (note the "T" separating the day from hours information). Remove 'current' from date/time types; convert to 'now' in input. Separate time and timetz regression tests. Separate timestamp and timestamptz regression test.
2001-09-28 10:09:14 +02:00
Remove comment on errno=0 lines, but add mention to port/strtol.c function.
2005-12-02 03:49:11 +01:00
errno = 0;
Rename strtoi() to strtoint(). NetBSD has seen fit to invent a libc function named strtoi(), which conflicts with the long-established static functions of the same name in datetime.c and ecpg's interval.c. While muttering darkly about intrusions on application namespace, we'll rename our functions to avoid the conflict. Back-patch to all supported branches, since this would affect attempts to build any of them on recent NetBSD. Thomas Munro
2016-04-23 22:53:15 +02:00
val = strtoint(field[i], &cp, 10);
Re-allow input of Julian dates prior to 0001-01-01 AD. This was unintentionally broken in 8.4 while tightening up checking of ordinary non-Julian date inputs to forbid references to "year zero". Per bug #5672 from Benjamin Gigot.
2010-09-23 05:48:07 +02:00
if (errno == ERANGE || val < 0)
Check for overflow in strtol() while parsing datetime inputs. Michael Fuhr.
2005-12-01 18:56:34 +01:00
return DTERR_FIELD_OVERFLOW;
Measure the current transaction time to milliseconds. Define a new function, GetCurrentTransactionStartTimeUsec() to get the time to this precision. Allow now() and timestamp 'now' to use this higher precision result so we now have fractional seconds in this "constant". Add timestamp without time zone type. Move previous timestamp type to timestamp with time zone. Accept another ISO variant for date/time values: yyyy-mm-ddThh:mm:ss (note the "T" separating the day from hours information). Remove 'current' from date/time types; convert to 'now' in input. Separate time and timetz regression tests. Separate timestamp and timestamptz regression test.
2001-09-28 10:09:14 +02:00
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
j2date(val, &tm->tm_year, &tm->tm_mon, &tm->tm_mday);
Change TRUE/FALSE to true/false The lower case spellings are C and C++ standard and are used in most parts of the PostgreSQL sources. The upper case spellings are only used in some files/modules. So standardize on the standard spellings. The APIs for ICU, Perl, and Windows define their own TRUE and FALSE, so those are left as is when using those APIs. In code comments, we use the lower-case spelling for the C concepts and keep the upper-case spelling for the SQL concepts. Reviewed-by: Michael Paquier <michael.paquier@gmail.com>
2017-08-16 06:22:32 +02:00
isjulian = true;
Re-allow input of Julian dates prior to 0001-01-01 AD. This was unintentionally broken in 8.4 while tightening up checking of ordinary non-Julian date inputs to forbid references to "year zero". Per bug #5672 from Benjamin Gigot.
2010-09-23 05:48:07 +02:00
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
/* Get the time zone from the end of the string */
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
dterr = DecodeTimezone(cp, tzp);
if (dterr)
return dterr;
Measure the current transaction time to milliseconds. Define a new function, GetCurrentTransactionStartTimeUsec() to get the time to this precision. Allow now() and timestamp 'now' to use this higher precision result so we now have fractional seconds in this "constant". Add timestamp without time zone type. Move previous timestamp type to timestamp with time zone. Accept another ISO variant for date/time values: yyyy-mm-ddThh:mm:ss (note the "T" separating the day from hours information). Remove 'current' from date/time types; convert to 'now' in input. Separate time and timetz regression tests. Separate timestamp and timestamptz regression test.
2001-09-28 10:09:14 +02:00
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
tmask = DTK_DATE_M | DTK_TIME_M | DTK_M(TZ);
ptype = 0;
break;
}
pgindent run for 9.4 This includes removing tabs after periods in C comments, which was applied to back branches, so this change should not effect backpatching.
2014-05-06 18:12:18 +02:00
Adjust C comments that would be wrap-able.
2013-10-02 01:44:53 +02:00
/*
Fix up some problems in handling of zic-style time zone names in datetime input routines. Remove the former "DecodePosixTimezone" function in favor of letting the zic code handle POSIX-style zone specs (see tzparse()). In particular this means that "PST+3" now means the same as "-03", whereas it used to mean "-11" --- the zone abbreviation is effectively just a noise word in this syntax. Make sure that all named and POSIX-style zone names will be parsed as a single token. Fix long-standing bogosities in printing and input of fractional-hour timezone offsets (since the tzparse() code will accept these, we'd better make 'em work). Also correct an error in the original coding of the zic-zone-name patch: in "timestamp without time zone" input, zone names are supposed to be allowed but ignored, but the coding was such that the zone changed the interpretation anyway.
2006-10-17 23:03:21 +02:00
* Already have a date? Then this might be a time zone name
Fix DecodeDateTime to allow timezone to appear before year. This had historically worked in some but not all cases, but as of 8.2 it failed for all timezone formats. Fix, and add regression test cases to catch future regressions in this area. Per gripe from Adam Witney.
2007-06-12 17:58:32 +02:00
* with embedded punctuation (e.g. "America/New_York") or a
* run-together time with trailing time zone (e.g. hhmmss-zz).
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
* - thomas 2001-12-25
Fix DecodeDateTime to allow timezone to appear before year. This had historically worked in some but not all cases, but as of 8.2 it failed for all timezone formats. Fix, and add regression test cases to catch future regressions in this area. Per gripe from Adam Witney.
2007-06-12 17:58:32 +02:00
*
* We consider it a time zone if we already have month & day.
* This is to allow the form "mmm dd hhmmss tz year", which
* we've historically accepted.
Adjust C comments that would be wrap-able.
2013-10-02 01:44:53 +02:00
*/
Fix DecodeDateTime to allow timezone to appear before year. This had historically worked in some but not all cases, but as of 8.2 it failed for all timezone formats. Fix, and add regression test cases to catch future regressions in this area. Per gripe from Adam Witney.
2007-06-12 17:58:32 +02:00
else if (ptype != 0 ||
((fmask & (DTK_M(MONTH) | DTK_M(DAY))) ==
(DTK_M(MONTH) | DTK_M(DAY))))
Support full POSIX-style time zone: EST+3, PST-3, etc. We probably support a superset of the spec, but I don't have the spec to confirm this. Update regression tests to include tests for this format.
2000-03-16 15:38:42 +01:00
{
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
/* No time zone accepted? Then quit... */
if (tzp == NULL)
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
return DTERR_BAD_FORMAT;
Support full POSIX-style time zone: EST+3, PST-3, etc. We probably support a superset of the spec, but I don't have the spec to confirm this. Update regression tests to include tests for this format.
2000-03-16 15:38:42 +01:00
Make sure that all <ctype.h> routines are called with unsigned char values; it's not portable to call them with signed chars. I recall doing this for the last release, but a few more uncasted calls have snuck in.
2001-12-31 00:09:42 +01:00
if (isdigit((unsigned char) *field[i]) || ptype != 0)
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
{
pgindent run.
2002-09-04 22:31:48 +02:00
char *cp;
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
Check a bit more carefully for preceeding ISO field tags when decoding date fields.
2001-12-29 19:40:58 +01:00
if (ptype != 0)
{
/* Sanity check; should not fail this test */
if (ptype != DTK_TIME)
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
return DTERR_BAD_FORMAT;
Check a bit more carefully for preceeding ISO field tags when decoding date fields.
2001-12-29 19:40:58 +01:00
ptype = 0;
}
pgindent run.
2002-09-04 22:31:48 +02:00
/*
* Starts with a digit but we already have a time
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* field? Then we are in trouble with a date and time
* already...
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
*/
if ((fmask & DTK_TIME_M) == DTK_TIME_M)
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
return DTERR_BAD_FORMAT;
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
if ((cp = strchr(field[i], '-')) == NULL)
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
return DTERR_BAD_FORMAT;
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
/* Get the time zone from the end of the string */
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
dterr = DecodeTimezone(cp, tzp);
if (dterr)
return dterr;
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
*cp = '\0';
pgindent run.
2002-09-04 22:31:48 +02:00
/*
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* Then read the rest of the field as a concatenated
* time
pgindent run.
2002-09-04 22:31:48 +02:00
*/
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
dterr = DecodeNumberField(strlen(field[i]), field[i],
fmask,
&tmask, tm,
fsec, &is2digits);
if (dterr < 0)
return dterr;
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
pgindent run.
2002-09-04 22:31:48 +02:00
/*
* modify tmask after returning from
* DecodeNumberField()
*/
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
tmask |= DTK_M(TZ);
}
else
{
Fix up some problems in handling of zic-style time zone names in datetime input routines. Remove the former "DecodePosixTimezone" function in favor of letting the zic code handle POSIX-style zone specs (see tzparse()). In particular this means that "PST+3" now means the same as "-03", whereas it used to mean "-11" --- the zone abbreviation is effectively just a noise word in this syntax. Make sure that all named and POSIX-style zone names will be parsed as a single token. Fix long-standing bogosities in printing and input of fractional-hour timezone offsets (since the tzparse() code will accept these, we'd better make 'em work). Also correct an error in the original coding of the zic-zone-name patch: in "timestamp without time zone" input, zone names are supposed to be allowed but ignored, but the coding was such that the zone changed the interpretation anyway.
2006-10-17 23:03:21 +02:00
namedTz = pg_tzset(field[i]);
if (!namedTz)
{
/*
* We should return an error code instead of
pgindent run for 8.3.
2007-11-15 22:14:46 +01:00
* ereport'ing directly, but then there is no way
* to report the bad time zone name.
Fix up some problems in handling of zic-style time zone names in datetime input routines. Remove the former "DecodePosixTimezone" function in favor of letting the zic code handle POSIX-style zone specs (see tzparse()). In particular this means that "PST+3" now means the same as "-03", whereas it used to mean "-11" --- the zone abbreviation is effectively just a noise word in this syntax. Make sure that all named and POSIX-style zone names will be parsed as a single token. Fix long-standing bogosities in printing and input of fractional-hour timezone offsets (since the tzparse() code will accept these, we'd better make 'em work). Also correct an error in the original coding of the zic-zone-name patch: in "timestamp without time zone" input, zone names are supposed to be allowed but ignored, but the coding was such that the zone changed the interpretation anyway.
2006-10-17 23:03:21 +02:00
*/
ereport(ERROR,
(errcode(ERRCODE_INVALID_PARAMETER_VALUE),
errmsg("time zone \"%s\" not recognized",
field[i])));
}
/* we'll apply the zone setting below */
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
tmask = DTK_M(TZ);
}
Support full POSIX-style time zone: EST+3, PST-3, etc. We probably support a superset of the spec, but I don't have the spec to confirm this. Update regression tests to include tests for this format.
2000-03-16 15:38:42 +01:00
}
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
else
{
Fix datetime input to behave correctly for Feb 29 in years BC. Formerly, DecodeDate attempted to verify the day-of-the-month exactly, but it was under the misapprehension that it would know whether we were looking at a BC year or not. In reality this check can't be made until the calling function (eg DecodeDateTime) has processed all the fields. So, split the BC adjustment and validity checks out into a new function ValidateDate that is called only after processing all the fields. In passing, this patch makes DecodeTimeOnly work for BC inputs, which it never did before. (The historical veracity of all this is nonexistent, of course, but if we're going to say we support proleptic Gregorian calendar then we should do it correctly. In any case the unpatched code is broken because it could emit dates that it would then reject on re-inputting.) Per report from Bernd Helmle. Back-patch as far as 8.0; in 7.x we were not using our own calendar support and so this seems a bit too risky to put into 7.4.
2008-02-26 00:21:01 +01:00
dterr = DecodeDate(field[i], fmask,
&tmask, &is2digits, tm);
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
if (dterr)
return dterr;
}
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
break;
case DTK_TIME:
pgindent run for release 9.3 This is the first run of the Perl-based pgindent script. Also update pgindent instructions.
2013-05-29 22:58:43 +02:00
Allow text timezone designations, e.g. "America/Chicago", when using the ISO "T" timestamptz format.
2012-08-25 23:44:53 +02:00
/*
* This might be an ISO time following a "t" field.
*/
if (ptype != 0)
{
/* Sanity check; should not fail this test */
if (ptype != DTK_TIME)
return DTERR_BAD_FORMAT;
ptype = 0;
}
Make our parsing of INTERVAL literals spec-compliant (or at least a heck of a lot closer than it was before). To do this, tweak coerce_type() to pass through the typmod information when invoking interval_in() on an UNKNOWN constant; then fix DecodeInterval to pay attention to the typmod when deciding how to interpret a units-less integer value. I changed one or two other details as well. I believe the code now reacts as expected by spec for all the literal syntaxes that are specifically enumerated in the spec. There are corner cases involving strings that don't exactly match the set of fields called out by the typmod, for which we might want to tweak the behavior some more; but I think this is an area of user friendliness rather than spec compliance. There remain some non-compliant details about the SQL syntax (as opposed to what's inside the literal string); but at least we'll throw error rather than silently doing the wrong thing in those cases.
2008-09-10 20:29:41 +02:00
dterr = DecodeTime(field[i], fmask, INTERVAL_FULL_RANGE,
&tmask, tm, fsec);
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
if (dterr)
return dterr;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
pgindent run on all C files. Java run to follow. initdb/regression tests pass.
2001-10-25 07:50:21 +02:00
/*
* Check upper limit on hours; other limits checked in
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
* DecodeTime()
*/
Allow times of 24:00:00 to match rounding behavior: regression=# select '23:59:59.9'::time(0); time ---------- 24:00:00 (1 row) This is bad because: regression=# select '24:00:00'::time(0); ERROR: date/time field value out of range: "24:00:00" The last example now works.
2005-10-14 13:47:57 +02:00
/* test for > 24:00:00 */
Use macros for time-based constants, rather than constants.
2011-03-12 15:31:18 +01:00
if (tm->tm_hour > HOURS_PER_DAY ||
(tm->tm_hour == HOURS_PER_DAY &&
When checking for datetime field overflow, we should allow a fractional-second part that rounds up to exactly 1.0 second. The previous coding rejected input like "00:12:57.9999999999999999999999999999", with the exact number of nines needed to cause failure varying depending on float-timestamp option and possibly on platform. Obviously this should round up to the next integral second, if we don't have enough precision to distinguish the value from that. Per bug #4789 from Robert Kruus. In passing, fix a missed check for fractional seconds in one copy of the "is it greater than 24:00:00" code. Broken all the way back, so patch all the way back.
2009-05-01 21:29:07 +02:00
(tm->tm_min > 0 || tm->tm_sec > 0 || *fsec > 0)))
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
return DTERR_FIELD_OVERFLOW;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
break;
case DTK_TZ:
Support full POSIX-style time zone: EST+3, PST-3, etc. We probably support a superset of the spec, but I don't have the spec to confirm this. Update regression tests to include tests for this format.
2000-03-16 15:38:42 +01:00
{
pgindent run.
2002-09-04 22:31:48 +02:00
int tz;
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
if (tzp == NULL)
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
return DTERR_BAD_FORMAT;
Support full POSIX-style time zone: EST+3, PST-3, etc. We probably support a superset of the spec, but I don't have the spec to confirm this. Update regression tests to include tests for this format.
2000-03-16 15:38:42 +01:00
Fix up some problems in handling of zic-style time zone names in datetime input routines. Remove the former "DecodePosixTimezone" function in favor of letting the zic code handle POSIX-style zone specs (see tzparse()). In particular this means that "PST+3" now means the same as "-03", whereas it used to mean "-11" --- the zone abbreviation is effectively just a noise word in this syntax. Make sure that all named and POSIX-style zone names will be parsed as a single token. Fix long-standing bogosities in printing and input of fractional-hour timezone offsets (since the tzparse() code will accept these, we'd better make 'em work). Also correct an error in the original coding of the zic-zone-name patch: in "timestamp without time zone" input, zone names are supposed to be allowed but ignored, but the coding was such that the zone changed the interpretation anyway.
2006-10-17 23:03:21 +02:00
dterr = DecodeTimezone(field[i], &tz);
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
if (dterr)
return dterr;
Fix up some problems in handling of zic-style time zone names in datetime input routines. Remove the former "DecodePosixTimezone" function in favor of letting the zic code handle POSIX-style zone specs (see tzparse()). In particular this means that "PST+3" now means the same as "-03", whereas it used to mean "-11" --- the zone abbreviation is effectively just a noise word in this syntax. Make sure that all named and POSIX-style zone names will be parsed as a single token. Fix long-standing bogosities in printing and input of fractional-hour timezone offsets (since the tzparse() code will accept these, we'd better make 'em work). Also correct an error in the original coding of the zic-zone-name patch: in "timestamp without time zone" input, zone names are supposed to be allowed but ignored, but the coding was such that the zone changed the interpretation anyway.
2006-10-17 23:03:21 +02:00
*tzp = tz;
tmask = DTK_M(TZ);
Support full POSIX-style time zone: EST+3, PST-3, etc. We probably support a superset of the spec, but I don't have the spec to confirm this. Update regression tests to include tests for this format.
2000-03-16 15:38:42 +01:00
}
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
break;
case DTK_NUMBER:
pgindent run.
2002-09-04 22:31:48 +02:00
pgindent run on all C files. Java run to follow. initdb/regression tests pass.
2001-10-25 07:50:21 +02:00
/*
* Was this an "ISO date" with embedded field labels? An
* example is "y2001m02d04" - thomas 2001-02-04
Measure the current transaction time to milliseconds. Define a new function, GetCurrentTransactionStartTimeUsec() to get the time to this precision. Allow now() and timestamp 'now' to use this higher precision result so we now have fractional seconds in this "constant". Add timestamp without time zone type. Move previous timestamp type to timestamp with time zone. Accept another ISO variant for date/time values: yyyy-mm-ddThh:mm:ss (note the "T" separating the day from hours information). Remove 'current' from date/time types; convert to 'now' in input. Separate time and timetz regression tests. Separate timestamp and timestamptz regression test.
2001-09-28 10:09:14 +02:00
*/
if (ptype != 0)
{
pgindent run on all C files. Java run to follow. initdb/regression tests pass.
2001-10-25 07:50:21 +02:00
char *cp;
int val;
Measure the current transaction time to milliseconds. Define a new function, GetCurrentTransactionStartTimeUsec() to get the time to this precision. Allow now() and timestamp 'now' to use this higher precision result so we now have fractional seconds in this "constant". Add timestamp without time zone type. Move previous timestamp type to timestamp with time zone. Accept another ISO variant for date/time values: yyyy-mm-ddThh:mm:ss (note the "T" separating the day from hours information). Remove 'current' from date/time types; convert to 'now' in input. Separate time and timetz regression tests. Separate timestamp and timestamptz regression test.
2001-09-28 10:09:14 +02:00
Remove comment on errno=0 lines, but add mention to port/strtol.c function.
2005-12-02 03:49:11 +01:00
errno = 0;
Rename strtoi() to strtoint(). NetBSD has seen fit to invent a libc function named strtoi(), which conflicts with the long-established static functions of the same name in datetime.c and ecpg's interval.c. While muttering darkly about intrusions on application namespace, we'll rename our functions to avoid the conflict. Back-patch to all supported branches, since this would affect attempts to build any of them on recent NetBSD. Thomas Munro
2016-04-23 22:53:15 +02:00
val = strtoint(field[i], &cp, 10);
Check for overflow in strtol() while parsing datetime inputs. Michael Fuhr.
2005-12-01 18:56:34 +01:00
if (errno == ERANGE)
return DTERR_FIELD_OVERFLOW;
pgindent run.
2002-09-04 22:31:48 +02:00
/*
* only a few kinds are allowed to have an embedded
* decimal
*/
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
if (*cp == '.')
switch (ptype)
{
case DTK_JULIAN:
case DTK_TIME:
case DTK_SECOND:
break;
default:
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
return DTERR_BAD_FORMAT;
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
break;
}
else if (*cp != '\0')
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
return DTERR_BAD_FORMAT;
Measure the current transaction time to milliseconds. Define a new function, GetCurrentTransactionStartTimeUsec() to get the time to this precision. Allow now() and timestamp 'now' to use this higher precision result so we now have fractional seconds in this "constant". Add timestamp without time zone type. Move previous timestamp type to timestamp with time zone. Accept another ISO variant for date/time values: yyyy-mm-ddThh:mm:ss (note the "T" separating the day from hours information). Remove 'current' from date/time types; convert to 'now' in input. Separate time and timetz regression tests. Separate timestamp and timestamptz regression test.
2001-09-28 10:09:14 +02:00
pgindent run on all C files. Java run to follow. initdb/regression tests pass.
2001-10-25 07:50:21 +02:00
switch (ptype)
{
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
case DTK_YEAR:
Measure the current transaction time to milliseconds. Define a new function, GetCurrentTransactionStartTimeUsec() to get the time to this precision. Allow now() and timestamp 'now' to use this higher precision result so we now have fractional seconds in this "constant". Add timestamp without time zone type. Move previous timestamp type to timestamp with time zone. Accept another ISO variant for date/time values: yyyy-mm-ddThh:mm:ss (note the "T" separating the day from hours information). Remove 'current' from date/time types; convert to 'now' in input. Separate time and timetz regression tests. Separate timestamp and timestamptz regression test.
2001-09-28 10:09:14 +02:00
tm->tm_year = val;
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
tmask = DTK_M(YEAR);
Measure the current transaction time to milliseconds. Define a new function, GetCurrentTransactionStartTimeUsec() to get the time to this precision. Allow now() and timestamp 'now' to use this higher precision result so we now have fractional seconds in this "constant". Add timestamp without time zone type. Move previous timestamp type to timestamp with time zone. Accept another ISO variant for date/time values: yyyy-mm-ddThh:mm:ss (note the "T" separating the day from hours information). Remove 'current' from date/time types; convert to 'now' in input. Separate time and timetz regression tests. Separate timestamp and timestamptz regression test.
2001-09-28 10:09:14 +02:00
break;
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
case DTK_MONTH:
pgindent run.
2002-09-04 22:31:48 +02:00
/*
* already have a month and hour? then assume
* minutes
*/
Remove unnecessary parentheses in datetime/timestamp code.
2005-05-23 19:13:14 +02:00
if ((fmask & DTK_M(MONTH)) != 0 &&
(fmask & DTK_M(HOUR)) != 0)
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
{
tm->tm_min = val;
tmask = DTK_M(MINUTE);
}
else
{
tm->tm_mon = val;
tmask = DTK_M(MONTH);
}
Measure the current transaction time to milliseconds. Define a new function, GetCurrentTransactionStartTimeUsec() to get the time to this precision. Allow now() and timestamp 'now' to use this higher precision result so we now have fractional seconds in this "constant". Add timestamp without time zone type. Move previous timestamp type to timestamp with time zone. Accept another ISO variant for date/time values: yyyy-mm-ddThh:mm:ss (note the "T" separating the day from hours information). Remove 'current' from date/time types; convert to 'now' in input. Separate time and timetz regression tests. Separate timestamp and timestamptz regression test.
2001-09-28 10:09:14 +02:00
break;
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
case DTK_DAY:
Measure the current transaction time to milliseconds. Define a new function, GetCurrentTransactionStartTimeUsec() to get the time to this precision. Allow now() and timestamp 'now' to use this higher precision result so we now have fractional seconds in this "constant". Add timestamp without time zone type. Move previous timestamp type to timestamp with time zone. Accept another ISO variant for date/time values: yyyy-mm-ddThh:mm:ss (note the "T" separating the day from hours information). Remove 'current' from date/time types; convert to 'now' in input. Separate time and timetz regression tests. Separate timestamp and timestamptz regression test.
2001-09-28 10:09:14 +02:00
tm->tm_mday = val;
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
tmask = DTK_M(DAY);
Measure the current transaction time to milliseconds. Define a new function, GetCurrentTransactionStartTimeUsec() to get the time to this precision. Allow now() and timestamp 'now' to use this higher precision result so we now have fractional seconds in this "constant". Add timestamp without time zone type. Move previous timestamp type to timestamp with time zone. Accept another ISO variant for date/time values: yyyy-mm-ddThh:mm:ss (note the "T" separating the day from hours information). Remove 'current' from date/time types; convert to 'now' in input. Separate time and timetz regression tests. Separate timestamp and timestamptz regression test.
2001-09-28 10:09:14 +02:00
break;
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
case DTK_HOUR:
Measure the current transaction time to milliseconds. Define a new function, GetCurrentTransactionStartTimeUsec() to get the time to this precision. Allow now() and timestamp 'now' to use this higher precision result so we now have fractional seconds in this "constant". Add timestamp without time zone type. Move previous timestamp type to timestamp with time zone. Accept another ISO variant for date/time values: yyyy-mm-ddThh:mm:ss (note the "T" separating the day from hours information). Remove 'current' from date/time types; convert to 'now' in input. Separate time and timetz regression tests. Separate timestamp and timestamptz regression test.
2001-09-28 10:09:14 +02:00
tm->tm_hour = val;
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
tmask = DTK_M(HOUR);
Measure the current transaction time to milliseconds. Define a new function, GetCurrentTransactionStartTimeUsec() to get the time to this precision. Allow now() and timestamp 'now' to use this higher precision result so we now have fractional seconds in this "constant". Add timestamp without time zone type. Move previous timestamp type to timestamp with time zone. Accept another ISO variant for date/time values: yyyy-mm-ddThh:mm:ss (note the "T" separating the day from hours information). Remove 'current' from date/time types; convert to 'now' in input. Separate time and timetz regression tests. Separate timestamp and timestamptz regression test.
2001-09-28 10:09:14 +02:00
break;
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
case DTK_MINUTE:
Measure the current transaction time to milliseconds. Define a new function, GetCurrentTransactionStartTimeUsec() to get the time to this precision. Allow now() and timestamp 'now' to use this higher precision result so we now have fractional seconds in this "constant". Add timestamp without time zone type. Move previous timestamp type to timestamp with time zone. Accept another ISO variant for date/time values: yyyy-mm-ddThh:mm:ss (note the "T" separating the day from hours information). Remove 'current' from date/time types; convert to 'now' in input. Separate time and timetz regression tests. Separate timestamp and timestamptz regression test.
2001-09-28 10:09:14 +02:00
tm->tm_min = val;
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
tmask = DTK_M(MINUTE);
Measure the current transaction time to milliseconds. Define a new function, GetCurrentTransactionStartTimeUsec() to get the time to this precision. Allow now() and timestamp 'now' to use this higher precision result so we now have fractional seconds in this "constant". Add timestamp without time zone type. Move previous timestamp type to timestamp with time zone. Accept another ISO variant for date/time values: yyyy-mm-ddThh:mm:ss (note the "T" separating the day from hours information). Remove 'current' from date/time types; convert to 'now' in input. Separate time and timetz regression tests. Separate timestamp and timestamptz regression test.
2001-09-28 10:09:14 +02:00
break;
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
case DTK_SECOND:
Measure the current transaction time to milliseconds. Define a new function, GetCurrentTransactionStartTimeUsec() to get the time to this precision. Allow now() and timestamp 'now' to use this higher precision result so we now have fractional seconds in this "constant". Add timestamp without time zone type. Move previous timestamp type to timestamp with time zone. Accept another ISO variant for date/time values: yyyy-mm-ddThh:mm:ss (note the "T" separating the day from hours information). Remove 'current' from date/time types; convert to 'now' in input. Separate time and timetz regression tests. Separate timestamp and timestamptz regression test.
2001-09-28 10:09:14 +02:00
tm->tm_sec = val;
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
tmask = DTK_M(SECOND);
if (*cp == '.')
{
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
dterr = ParseFractionalSecond(cp, fsec);
if (dterr)
return dterr;
Fix a bug in input processing for the "interval" type. Previously, "microsecond" and "millisecond" units were not considered valid input by themselves, which caused inputs like "1 millisecond" to be rejected erroneously. Update the docs, add regression tests, and backport to 8.2 and 8.1
2007-05-29 06:58:43 +02:00
tmask = DTK_ALL_SECS_M;
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
}
Measure the current transaction time to milliseconds. Define a new function, GetCurrentTransactionStartTimeUsec() to get the time to this precision. Allow now() and timestamp 'now' to use this higher precision result so we now have fractional seconds in this "constant". Add timestamp without time zone type. Move previous timestamp type to timestamp with time zone. Accept another ISO variant for date/time values: yyyy-mm-ddThh:mm:ss (note the "T" separating the day from hours information). Remove 'current' from date/time types; convert to 'now' in input. Separate time and timetz regression tests. Separate timestamp and timestamptz regression test.
2001-09-28 10:09:14 +02:00
break;
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
case DTK_TZ:
tmask = DTK_M(TZ);
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
dterr = DecodeTimezone(field[i], tzp);
if (dterr)
return dterr;
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
break;
pgindent run on all C files. Java run to follow. initdb/regression tests pass.
2001-10-25 07:50:21 +02:00
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
case DTK_JULIAN:
Re-allow input of Julian dates prior to 0001-01-01 AD. This was unintentionally broken in 8.4 while tightening up checking of ordinary non-Julian date inputs to forbid references to "year zero". Per bug #5672 from Benjamin Gigot.
2010-09-23 05:48:07 +02:00
/* previous field was a label for "julian date" */
if (val < 0)
return DTERR_FIELD_OVERFLOW;
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
tmask = DTK_DATE_M;
Measure the current transaction time to milliseconds. Define a new function, GetCurrentTransactionStartTimeUsec() to get the time to this precision. Allow now() and timestamp 'now' to use this higher precision result so we now have fractional seconds in this "constant". Add timestamp without time zone type. Move previous timestamp type to timestamp with time zone. Accept another ISO variant for date/time values: yyyy-mm-ddThh:mm:ss (note the "T" separating the day from hours information). Remove 'current' from date/time types; convert to 'now' in input. Separate time and timetz regression tests. Separate timestamp and timestamptz regression test.
2001-09-28 10:09:14 +02:00
j2date(val, &tm->tm_year, &tm->tm_mon, &tm->tm_mday);
Change TRUE/FALSE to true/false The lower case spellings are C and C++ standard and are used in most parts of the PostgreSQL sources. The upper case spellings are only used in some files/modules. So standardize on the standard spellings. The APIs for ICU, Perl, and Windows define their own TRUE and FALSE, so those are left as is when using those APIs. In code comments, we use the lower-case spelling for the C concepts and keep the upper-case spelling for the SQL concepts. Reviewed-by: Michael Paquier <michael.paquier@gmail.com>
2017-08-16 06:22:32 +02:00
isjulian = true;
Re-allow input of Julian dates prior to 0001-01-01 AD. This was unintentionally broken in 8.4 while tightening up checking of ordinary non-Julian date inputs to forbid references to "year zero". Per bug #5672 from Benjamin Gigot.
2010-09-23 05:48:07 +02:00
Support alternate storage scheme of 64-bit integer for date/time types. Use "--enable-integer-datetimes" in configuration to use this rather than the original float8 storage. I would recommend the integer-based storage for any platform on which it is available. We perhaps should make this the default for the production release. Change timezone(timestamptz) results to return timestamp rather than a character string. Formerly, we didn't have a way to represent timestamps with an explicit time zone other than freezing the info into a string. Now, we can reasonably omit the explicit time zone from the result and return a timestamp with values appropriate for the specified time zone. Much cleaner, and if you need the time zone in the result you can put it into a character string pretty easily anyway. Allow fractional seconds in date/time types even for dates prior to 1BC. Limit timestamp data types to 6 decimal places of precision. Just right for a micro-second storage of int8 date/time types, and reduces the number of places ad-hoc rounding was occuring for the float8-based types. Use lookup tables for precision/rounding calculations for timestamp and interval types. Formerly used pow() to calculate the desired value but with a more limited range there is no reason to not type in a lookup table. Should be *much* better performance, though formerly there were some optimizations to help minimize the number of times pow() was called. Define a HAVE_INT64_TIMESTAMP variable. Based on the configure option "--enable-integer-datetimes" and the existing internal INT64_IS_BUSTED. Add explicit date/interval operators and functions for addition and subtraction. Formerly relied on implicit type promotion from date to timestamp with time zone. Change timezone conversion functions for the timetz type from "timetz()" to "timezone()". This is consistant with other time zone coersion functions for other types. Bump the catalog version to 200204201. Fix up regression tests to reflect changes in fractional seconds representation for date/times in BC eras. All regression tests pass on my Linux box.
2002-04-21 21:52:18 +02:00
/* fractional Julian Day? */
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
if (*cp == '.')
{
pgindent run.
2002-09-04 22:31:48 +02:00
double time;
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
errno = 0;
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
time = strtod(cp, &cp);
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
if (*cp != '\0' || errno != 0)
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
return DTERR_BAD_FORMAT;
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
time *= USECS_PER_DAY;
dt2time(time,
&tm->tm_hour, &tm->tm_min,
&tm->tm_sec, fsec);
tmask |= DTK_TIME_M;
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
}
break;
Measure the current transaction time to milliseconds. Define a new function, GetCurrentTransactionStartTimeUsec() to get the time to this precision. Allow now() and timestamp 'now' to use this higher precision result so we now have fractional seconds in this "constant". Add timestamp without time zone type. Move previous timestamp type to timestamp with time zone. Accept another ISO variant for date/time values: yyyy-mm-ddThh:mm:ss (note the "T" separating the day from hours information). Remove 'current' from date/time types; convert to 'now' in input. Separate time and timetz regression tests. Separate timestamp and timestamptz regression test.
2001-09-28 10:09:14 +02:00
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
case DTK_TIME:
/* previous field was "t" for ISO time */
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
dterr = DecodeNumberField(strlen(field[i]), field[i],
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
(fmask | DTK_DATE_M),
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
&tmask, tm,
fsec, &is2digits);
if (dterr < 0)
return dterr;
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
if (tmask != DTK_TIME_M)
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
return DTERR_BAD_FORMAT;
Measure the current transaction time to milliseconds. Define a new function, GetCurrentTransactionStartTimeUsec() to get the time to this precision. Allow now() and timestamp 'now' to use this higher precision result so we now have fractional seconds in this "constant". Add timestamp without time zone type. Move previous timestamp type to timestamp with time zone. Accept another ISO variant for date/time values: yyyy-mm-ddThh:mm:ss (note the "T" separating the day from hours information). Remove 'current' from date/time types; convert to 'now' in input. Separate time and timetz regression tests. Separate timestamp and timestamptz regression test.
2001-09-28 10:09:14 +02:00
break;
default:
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
return DTERR_BAD_FORMAT;
Measure the current transaction time to milliseconds. Define a new function, GetCurrentTransactionStartTimeUsec() to get the time to this precision. Allow now() and timestamp 'now' to use this higher precision result so we now have fractional seconds in this "constant". Add timestamp without time zone type. Move previous timestamp type to timestamp with time zone. Accept another ISO variant for date/time values: yyyy-mm-ddThh:mm:ss (note the "T" separating the day from hours information). Remove 'current' from date/time types; convert to 'now' in input. Separate time and timetz regression tests. Separate timestamp and timestamptz regression test.
2001-09-28 10:09:14 +02:00
break;
}
ptype = 0;
*dtype = DTK_DATE;
}
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
else
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
{
pgindent run.
2002-09-04 22:31:48 +02:00
char *cp;
int flen;
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
flen = strlen(field[i]);
cp = strchr(field[i], '.');
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
/* Embedded decimal and no date yet? */
Remove unnecessary parentheses in datetime/timestamp code.
2005-05-23 19:13:14 +02:00
if (cp != NULL && !(fmask & DTK_DATE_M))
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
{
Fix datetime input to behave correctly for Feb 29 in years BC. Formerly, DecodeDate attempted to verify the day-of-the-month exactly, but it was under the misapprehension that it would know whether we were looking at a BC year or not. In reality this check can't be made until the calling function (eg DecodeDateTime) has processed all the fields. So, split the BC adjustment and validity checks out into a new function ValidateDate that is called only after processing all the fields. In passing, this patch makes DecodeTimeOnly work for BC inputs, which it never did before. (The historical veracity of all this is nonexistent, of course, but if we're going to say we support proleptic Gregorian calendar then we should do it correctly. In any case the unpatched code is broken because it could emit dates that it would then reject on re-inputting.) Per report from Bernd Helmle. Back-patch as far as 8.0; in 7.x we were not using our own calendar support and so this seems a bit too risky to put into 7.4.
2008-02-26 00:21:01 +01:00
dterr = DecodeDate(field[i], fmask,
&tmask, &is2digits, tm);
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
if (dterr)
return dterr;
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
}
/* embedded decimal and several digits before? */
Remove unnecessary parentheses in datetime/timestamp code.
2005-05-23 19:13:14 +02:00
else if (cp != NULL && flen - strlen(cp) > 2)
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
{
pgindent run.
2002-09-04 22:31:48 +02:00
/*
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* Interpret as a concatenated date or time Set the
* type field to allow decoding other fields later.
* Example: 20011223 or 040506
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
*/
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
dterr = DecodeNumberField(flen, field[i], fmask,
&tmask, tm,
fsec, &is2digits);
if (dterr < 0)
return dterr;
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
}
pgindent run for 9.4 This includes removing tabs after periods in C comments, which was applied to back branches, so this change should not effect backpatching.
2014-05-06 18:12:18 +02:00
Allow 5+ digit years for non-ISO timestamp/date strings, where appropriate Report from Haribabu Kommi
2013-10-16 19:22:55 +02:00
/*
* Is this a YMD or HMS specification, or a year number?
* YMD and HMS are required to be six digits or more, so
* if it is 5 digits, it is a year. If it is six or more
Remove duplicated words split across lines in comments This has been detected using some interesting tricks with sed, and the method used is mentioned in details in the discussion below. Author: Justin Pryzby Discussion: https://postgr.es/m/20180908013109.GB15350@telsasoft.com
2018-09-08 21:24:19 +02:00
* digits, we assume it is YMD or HMS unless no date and
* no time values have been specified. This forces 6+
pgindent run for 9.4 This includes removing tabs after periods in C comments, which was applied to back branches, so this change should not effect backpatching.
2014-05-06 18:12:18 +02:00
* digit years to be at the end of the string, or to use
Allow 5+ digit years for non-ISO timestamp/date strings, where appropriate Report from Haribabu Kommi
2013-10-16 19:22:55 +02:00
* the ISO date specification.
*/
else if (flen >= 6 && (!(fmask & DTK_DATE_M) ||
pgindent run for 9.4 This includes removing tabs after periods in C comments, which was applied to back branches, so this change should not effect backpatching.
2014-05-06 18:12:18 +02:00
!(fmask & DTK_TIME_M)))
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
{
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
dterr = DecodeNumberField(flen, field[i], fmask,
&tmask, tm,
fsec, &is2digits);
if (dterr < 0)
return dterr;
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
}
/* otherwise it is a single date/time field... */
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
else
{
Fix datetime input parsing to accept YYYY-MONTHNAME-DD and related syntaxes, which had been unintentionally broken by recent changes to tighten up the DateStyle rules for all-numeric date input. Add documentation and regression tests for this, too.
2003-11-16 21:29:16 +01:00
dterr = DecodeNumber(flen, field[i],
haveTextMonth, fmask,
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
&tmask, tm,
fsec, &is2digits);
if (dterr)
return dterr;
}
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
}
break;
case DTK_STRING:
case DTK_SPECIAL:
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
/* timezone abbrevs take precedence over built-in tokens */
type = DecodeTimezoneAbbrev(i, field[i], &val, &valtz);
if (type == UNKNOWN_FIELD)
type = DecodeSpecial(i, field[i], &val);
Katherine Ward wrote: > Changes to avoid collisions with WIN32 & MFC names... > 1. Renamed: > a. PROC => PGPROC > b. GetUserName() => GetUserNameFromId() > c. GetCurrentTime() => GetCurrentDateTime() > d. IGNORE => IGNORE_DTF in include/utils/datetime.h & utils/adt/datetim > > 2. Added _P to some lex/yacc tokens: > CONST, CHAR, DELETE, FLOAT, GROUP, IN, OUT Jan
2002-06-11 15:40:53 +02:00
if (type == IGNORE_DTF)
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
continue;
tmask = DTK_M(type);
switch (type)
{
case RESERV:
switch (val)
{
case DTK_NOW:
tmask = (DTK_DATE_M | DTK_TIME_M | DTK_M(TZ));
*dtype = DTK_DATE;
Fix timestamptz_in so that parsing of 'now'::timestamptz gives right answer when SET TIMEZONE has been done since the start of the current transaction. Per bug report from Robert Haas. I plan some futher cleanup in HEAD, but this is a low-risk patch for the immediate issue in 7.3.
2003-02-20 06:24:55 +01:00
GetCurrentTimeUsec(tm, fsec, tzp);
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
break;
case DTK_YESTERDAY:
tmask = DTK_DATE_M;
*dtype = DTK_DATE;
Fix parsing of time string followed by yesterday/today/tomorrow. Previously, 'yesterday 04:00:00'::timestamp didn't do the same thing as '04:00:00 yesterday'::timestamp, and the return value from the latter was midnight rather than the specified time. Dean Rasheed, with some stylistic changes
2011-08-30 17:34:29 +02:00
GetCurrentDateTime(&cur_tm);
j2date(date2j(cur_tm.tm_year, cur_tm.tm_mon, cur_tm.tm_mday) - 1,
Phase 3 of pgindent updates. Don't move parenthesized lines to the left, even if that means they flow past the right margin. By default, BSD indent lines up statement continuation lines that are within parentheses so that they start just to the right of the preceding left parenthesis. However, traditionally, if that resulted in the continuation line extending to the right of the desired right margin, then indent would push it left just far enough to not overrun the margin, if it could do so without making the continuation line start to the left of the current statement indent. That makes for a weird mix of indentations unless one has been completely rigid about never violating the 80-column limit. This behavior has been pretty universally panned by Postgres developers. Hence, disable it with indent's new -lpl switch, so that parenthesized lines are always lined up with the preceding left paren. This patch is much less interesting than the first round of indent changes, but also bulkier, so I thought it best to separate the effects. Discussion: https://postgr.es/m/E1dAmxK-0006EE-1r@gemulon.postgresql.org Discussion: https://postgr.es/m/30527.1495162840@sss.pgh.pa.us
2017-06-21 21:35:54 +02:00
&tm->tm_year, &tm->tm_mon, &tm->tm_mday);
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
break;
case DTK_TODAY:
tmask = DTK_DATE_M;
*dtype = DTK_DATE;
Fix parsing of time string followed by yesterday/today/tomorrow. Previously, 'yesterday 04:00:00'::timestamp didn't do the same thing as '04:00:00 yesterday'::timestamp, and the return value from the latter was midnight rather than the specified time. Dean Rasheed, with some stylistic changes
2011-08-30 17:34:29 +02:00
GetCurrentDateTime(&cur_tm);
tm->tm_year = cur_tm.tm_year;
tm->tm_mon = cur_tm.tm_mon;
tm->tm_mday = cur_tm.tm_mday;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
break;
case DTK_TOMORROW:
tmask = DTK_DATE_M;
*dtype = DTK_DATE;
Fix parsing of time string followed by yesterday/today/tomorrow. Previously, 'yesterday 04:00:00'::timestamp didn't do the same thing as '04:00:00 yesterday'::timestamp, and the return value from the latter was midnight rather than the specified time. Dean Rasheed, with some stylistic changes
2011-08-30 17:34:29 +02:00
GetCurrentDateTime(&cur_tm);
j2date(date2j(cur_tm.tm_year, cur_tm.tm_mon, cur_tm.tm_mday) + 1,
Phase 3 of pgindent updates. Don't move parenthesized lines to the left, even if that means they flow past the right margin. By default, BSD indent lines up statement continuation lines that are within parentheses so that they start just to the right of the preceding left parenthesis. However, traditionally, if that resulted in the continuation line extending to the right of the desired right margin, then indent would push it left just far enough to not overrun the margin, if it could do so without making the continuation line start to the left of the current statement indent. That makes for a weird mix of indentations unless one has been completely rigid about never violating the 80-column limit. This behavior has been pretty universally panned by Postgres developers. Hence, disable it with indent's new -lpl switch, so that parenthesized lines are always lined up with the preceding left paren. This patch is much less interesting than the first round of indent changes, but also bulkier, so I thought it best to separate the effects. Discussion: https://postgr.es/m/E1dAmxK-0006EE-1r@gemulon.postgresql.org Discussion: https://postgr.es/m/30527.1495162840@sss.pgh.pa.us
2017-06-21 21:35:54 +02:00
&tm->tm_year, &tm->tm_mon, &tm->tm_mday);
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
break;
case DTK_ZULU:
tmask = (DTK_TIME_M | DTK_M(TZ));
*dtype = DTK_DATE;
tm->tm_hour = 0;
tm->tm_min = 0;
tm->tm_sec = 0;
if (tzp != NULL)
*tzp = 0;
break;
default:
*dtype = val;
}
break;
case MONTH:
Ye-old pgindent run. Same 4-space tabs.
2000-04-12 19:17:23 +02:00
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
/*
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* already have a (numeric) month? then see if we can
* substitute...
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
*/
Remove unnecessary parentheses in datetime/timestamp code.
2005-05-23 19:13:14 +02:00
if ((fmask & DTK_M(MONTH)) && !haveTextMonth &&
!(fmask & DTK_M(DAY)) && tm->tm_mon >= 1 &&
tm->tm_mon <= 31)
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
{
tm->tm_mday = tm->tm_mon;
tmask = DTK_M(DAY);
}
Change TRUE/FALSE to true/false The lower case spellings are C and C++ standard and are used in most parts of the PostgreSQL sources. The upper case spellings are only used in some files/modules. So standardize on the standard spellings. The APIs for ICU, Perl, and Windows define their own TRUE and FALSE, so those are left as is when using those APIs. In code comments, we use the lower-case spelling for the C concepts and keep the upper-case spelling for the SQL concepts. Reviewed-by: Michael Paquier <michael.paquier@gmail.com>
2017-08-16 06:22:32 +02:00
haveTextMonth = true;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
tm->tm_mon = val;
break;
case DTZMOD:
/*
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* daylight savings time modifier (solves "MET DST"
* syntax)
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
*/
tmask |= DTK_M(DTZ);
tm->tm_isdst = 1;
if (tzp == NULL)
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
return DTERR_BAD_FORMAT;
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
*tzp -= val;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
break;
case DTZ:
/*
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* set mask for TZ here _or_ check for DTZ later when
* getting default timezone
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
*/
tmask |= DTK_M(TZ);
tm->tm_isdst = 1;
if (tzp == NULL)
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
return DTERR_BAD_FORMAT;
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
*tzp = -val;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
break;
case TZ:
tm->tm_isdst = 0;
if (tzp == NULL)
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
return DTERR_BAD_FORMAT;
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
*tzp = -val;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
break;
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
case DYNTZ:
tmask |= DTK_M(TZ);
if (tzp == NULL)
return DTERR_BAD_FORMAT;
/* we'll determine the actual offset later */
abbrevTz = valtz;
abbrev = field[i];
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
break;
case AMPM:
mer = val;
break;
case ADBC:
bc = (val == BC);
break;
case DOW:
tm->tm_wday = val;
break;
Measure the current transaction time to milliseconds. Define a new function, GetCurrentTransactionStartTimeUsec() to get the time to this precision. Allow now() and timestamp 'now' to use this higher precision result so we now have fractional seconds in this "constant". Add timestamp without time zone type. Move previous timestamp type to timestamp with time zone. Accept another ISO variant for date/time values: yyyy-mm-ddThh:mm:ss (note the "T" separating the day from hours information). Remove 'current' from date/time types; convert to 'now' in input. Separate time and timetz regression tests. Separate timestamp and timestamptz regression test.
2001-09-28 10:09:14 +02:00
case UNITS:
tmask = 0;
Fix transposed arguments for typmod for one INTERVAL production. Mask both typmod subfields for INTERVAL to avoid setting the high bit, per dire warning from Tom Lane. Clear tmask for DTK_ISO_TIME case to avoid time zone troubles. Symptom reported by Tom Lane. Clean up checking for valid time zone info in output routine. This should now work for both SQL99 and Unix-style time zones. Put in explicit check for INTERVAL() typmod rounding to avoid accumulating cruft in the lower bits. Not sure that this helps, but we'll need to do something. The symptom is visible with a query like select interval(2) '10000 days 01:02:03.040506'; Regression tests are patched to repair the Tom Lane symptom, and all pass.
2001-10-20 03:02:22 +02:00
ptype = val;
Measure the current transaction time to milliseconds. Define a new function, GetCurrentTransactionStartTimeUsec() to get the time to this precision. Allow now() and timestamp 'now' to use this higher precision result so we now have fractional seconds in this "constant". Add timestamp without time zone type. Move previous timestamp type to timestamp with time zone. Accept another ISO variant for date/time values: yyyy-mm-ddThh:mm:ss (note the "T" separating the day from hours information). Remove 'current' from date/time types; convert to 'now' in input. Separate time and timetz regression tests. Separate timestamp and timestamptz regression test.
2001-09-28 10:09:14 +02:00
break;
Rename TIME token to ISOTIME to eliminate conflict with gram.y parser token. Seems to be isolated to datetime.c and datetime.h.
2002-01-01 03:54:33 +01:00
case ISOTIME:
pgindent run.
2002-09-04 22:31:48 +02:00
/*
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* This is a filler field "t" indicating that the next
* field is time. Try to verify that this is sensible.
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
*/
Fix transposed arguments for typmod for one INTERVAL production. Mask both typmod subfields for INTERVAL to avoid setting the high bit, per dire warning from Tom Lane. Clear tmask for DTK_ISO_TIME case to avoid time zone troubles. Symptom reported by Tom Lane. Clean up checking for valid time zone info in output routine. This should now work for both SQL99 and Unix-style time zones. Put in explicit check for INTERVAL() typmod rounding to avoid accumulating cruft in the lower bits. Not sure that this helps, but we'll need to do something. The symptom is visible with a query like select interval(2) '10000 days 01:02:03.040506'; Regression tests are patched to repair the Tom Lane symptom, and all pass.
2001-10-20 03:02:22 +02:00
tmask = 0;
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
/* No preceding date? Then quit... */
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
if ((fmask & DTK_DATE_M) != DTK_DATE_M)
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
return DTERR_BAD_FORMAT;
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
/***
* We will need one of the following fields:
pgindent run.
2002-09-04 22:31:48 +02:00
* DTK_NUMBER should be hhmmss.fff
* DTK_TIME should be hh:mm:ss.fff
* DTK_DATE should be hhmmss-zz
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
***/
Remove unnecessary parentheses in datetime/timestamp code.
2005-05-23 19:13:14 +02:00
if (i >= nf - 1 ||
(ftype[i + 1] != DTK_NUMBER &&
ftype[i + 1] != DTK_TIME &&
ftype[i + 1] != DTK_DATE))
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
return DTERR_BAD_FORMAT;
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
ptype = val;
Measure the current transaction time to milliseconds. Define a new function, GetCurrentTransactionStartTimeUsec() to get the time to this precision. Allow now() and timestamp 'now' to use this higher precision result so we now have fractional seconds in this "constant". Add timestamp without time zone type. Move previous timestamp type to timestamp with time zone. Accept another ISO variant for date/time values: yyyy-mm-ddThh:mm:ss (note the "T" separating the day from hours information). Remove 'current' from date/time types; convert to 'now' in input. Separate time and timetz regression tests. Separate timestamp and timestamptz regression test.
2001-09-28 10:09:14 +02:00
break;
Fix up timetz input so that a date is required only when the specified timezone actually has a daylight-savings rule. This avoids breaking cases that used to work because they went through the DecodePosixTimezone code path. Per contrib regression failures (mea culpa for not running those yesterday...). Also document the already-applied change to allow GMT offsets up to 14 hours.
2006-10-18 18:43:14 +02:00
case UNKNOWN_FIELD:
pgindent run for 8.3.
2007-11-15 22:14:46 +01:00
Fix up timetz input so that a date is required only when the specified timezone actually has a daylight-savings rule. This avoids breaking cases that used to work because they went through the DecodePosixTimezone code path. Per contrib regression failures (mea culpa for not running those yesterday...). Also document the already-applied change to allow GMT offsets up to 14 hours.
2006-10-18 18:43:14 +02:00
/*
* Before giving up and declaring error, check to see
* if it is an all-alpha timezone name.
*/
namedTz = pg_tzset(field[i]);
if (!namedTz)
return DTERR_BAD_FORMAT;
/* we'll apply the zone setting below */
tmask = DTK_M(TZ);
break;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
default:
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
return DTERR_BAD_FORMAT;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
}
break;
default:
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
return DTERR_BAD_FORMAT;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
}
if (tmask & fmask)
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
return DTERR_BAD_FORMAT;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
fmask |= tmask;
8.4 pgindent run, with new combined Linux/FreeBSD/MinGW typedef list provided by Andrew.
2009-06-11 16:49:15 +02:00
} /* end loop over fields */
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
Fix datetime input to behave correctly for Feb 29 in years BC. Formerly, DecodeDate attempted to verify the day-of-the-month exactly, but it was under the misapprehension that it would know whether we were looking at a BC year or not. In reality this check can't be made until the calling function (eg DecodeDateTime) has processed all the fields. So, split the BC adjustment and validity checks out into a new function ValidateDate that is called only after processing all the fields. In passing, this patch makes DecodeTimeOnly work for BC inputs, which it never did before. (The historical veracity of all this is nonexistent, of course, but if we're going to say we support proleptic Gregorian calendar then we should do it correctly. In any case the unpatched code is broken because it could emit dates that it would then reject on re-inputting.) Per report from Bernd Helmle. Back-patch as far as 8.0; in 7.x we were not using our own calendar support and so this seems a bit too risky to put into 7.4.
2008-02-26 00:21:01 +01:00
/* do final checking/adjustment of Y/M/D fields */
Re-allow input of Julian dates prior to 0001-01-01 AD. This was unintentionally broken in 8.4 while tightening up checking of ordinary non-Julian date inputs to forbid references to "year zero". Per bug #5672 from Benjamin Gigot.
2010-09-23 05:48:07 +02:00
dterr = ValidateDate(fmask, isjulian, is2digits, bc, tm);
Fix datetime input to behave correctly for Feb 29 in years BC. Formerly, DecodeDate attempted to verify the day-of-the-month exactly, but it was under the misapprehension that it would know whether we were looking at a BC year or not. In reality this check can't be made until the calling function (eg DecodeDateTime) has processed all the fields. So, split the BC adjustment and validity checks out into a new function ValidateDate that is called only after processing all the fields. In passing, this patch makes DecodeTimeOnly work for BC inputs, which it never did before. (The historical veracity of all this is nonexistent, of course, but if we're going to say we support proleptic Gregorian calendar then we should do it correctly. In any case the unpatched code is broken because it could emit dates that it would then reject on re-inputting.) Per report from Bernd Helmle. Back-patch as far as 8.0; in 7.x we were not using our own calendar support and so this seems a bit too risky to put into 7.4.
2008-02-26 00:21:01 +01:00
if (dterr)
return dterr;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
Fix datetime input to behave correctly for Feb 29 in years BC. Formerly, DecodeDate attempted to verify the day-of-the-month exactly, but it was under the misapprehension that it would know whether we were looking at a BC year or not. In reality this check can't be made until the calling function (eg DecodeDateTime) has processed all the fields. So, split the BC adjustment and validity checks out into a new function ValidateDate that is called only after processing all the fields. In passing, this patch makes DecodeTimeOnly work for BC inputs, which it never did before. (The historical veracity of all this is nonexistent, of course, but if we're going to say we support proleptic Gregorian calendar then we should do it correctly. In any case the unpatched code is broken because it could emit dates that it would then reject on re-inputting.) Per report from Bernd Helmle. Back-patch as far as 8.0; in 7.x we were not using our own calendar support and so this seems a bit too risky to put into 7.4.
2008-02-26 00:21:01 +01:00
/* handle AM/PM */
Use macros for time-based constants, rather than constants.
2011-03-12 15:31:18 +01:00
if (mer != HR24 && tm->tm_hour > HOURS_PER_DAY / 2)
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
return DTERR_FIELD_OVERFLOW;
Use macros for time-based constants, rather than constants.
2011-03-12 15:31:18 +01:00
if (mer == AM && tm->tm_hour == HOURS_PER_DAY / 2)
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
tm->tm_hour = 0;
Use macros for time-based constants, rather than constants.
2011-03-12 15:31:18 +01:00
else if (mer == PM && tm->tm_hour != HOURS_PER_DAY / 2)
tm->tm_hour += HOURS_PER_DAY / 2;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
/* do additional checking for full date specs... */
if (*dtype == DTK_DATE)
{
if ((fmask & DTK_DATE_M) != DTK_DATE_M)
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
{
if ((fmask & DTK_TIME_M) == DTK_TIME_M)
return 1;
return DTERR_BAD_FORMAT;
}
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
Fix up some problems in handling of zic-style time zone names in datetime input routines. Remove the former "DecodePosixTimezone" function in favor of letting the zic code handle POSIX-style zone specs (see tzparse()). In particular this means that "PST+3" now means the same as "-03", whereas it used to mean "-11" --- the zone abbreviation is effectively just a noise word in this syntax. Make sure that all named and POSIX-style zone names will be parsed as a single token. Fix long-standing bogosities in printing and input of fractional-hour timezone offsets (since the tzparse() code will accept these, we'd better make 'em work). Also correct an error in the original coding of the zic-zone-name patch: in "timestamp without time zone" input, zone names are supposed to be allowed but ignored, but the coding was such that the zone changed the interpretation anyway.
2006-10-17 23:03:21 +02:00
/*
pgindent run for 8.3.
2007-11-15 22:14:46 +01:00
* If we had a full timezone spec, compute the offset (we could not do
* it before, because we need the date to resolve DST status).
Fix up some problems in handling of zic-style time zone names in datetime input routines. Remove the former "DecodePosixTimezone" function in favor of letting the zic code handle POSIX-style zone specs (see tzparse()). In particular this means that "PST+3" now means the same as "-03", whereas it used to mean "-11" --- the zone abbreviation is effectively just a noise word in this syntax. Make sure that all named and POSIX-style zone names will be parsed as a single token. Fix long-standing bogosities in printing and input of fractional-hour timezone offsets (since the tzparse() code will accept these, we'd better make 'em work). Also correct an error in the original coding of the zic-zone-name patch: in "timestamp without time zone" input, zone names are supposed to be allowed but ignored, but the coding was such that the zone changed the interpretation anyway.
2006-10-17 23:03:21 +02:00
*/
if (namedTz != NULL)
Allow timezone names in SQL strings, '2006-05-24 21:11 Americas/New_York'::timestamptz Joachim Wieland
2006-06-08 00:32:31 +02:00
{
Fix up some problems in handling of zic-style time zone names in datetime input routines. Remove the former "DecodePosixTimezone" function in favor of letting the zic code handle POSIX-style zone specs (see tzparse()). In particular this means that "PST+3" now means the same as "-03", whereas it used to mean "-11" --- the zone abbreviation is effectively just a noise word in this syntax. Make sure that all named and POSIX-style zone names will be parsed as a single token. Fix long-standing bogosities in printing and input of fractional-hour timezone offsets (since the tzparse() code will accept these, we'd better make 'em work). Also correct an error in the original coding of the zic-zone-name patch: in "timestamp without time zone" input, zone names are supposed to be allowed but ignored, but the coding was such that the zone changed the interpretation anyway.
2006-10-17 23:03:21 +02:00
/* daylight savings time modifier disallowed with full TZ */
if (fmask & DTK_M(DTZMOD))
return DTERR_BAD_FORMAT;
*tzp = DetermineTimeZoneOffset(tm, namedTz);
Allow timezone names in SQL strings, '2006-05-24 21:11 Americas/New_York'::timestamptz Joachim Wieland
2006-06-08 00:32:31 +02:00
}
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
/*
* Likewise, if we had a dynamic timezone abbreviation, resolve it
* now.
*/
if (abbrevTz != NULL)
{
/* daylight savings time modifier disallowed with dynamic TZ */
if (fmask & DTK_M(DTZMOD))
return DTERR_BAD_FORMAT;
*tzp = DetermineTimeZoneAbbrevOffset(tm, abbrev, abbrevTz);
}
/* timezone not specified? then use session timezone */
Remove excess parens, use Abs instead of : ?.
2005-05-21 05:38:05 +02:00
if (tzp != NULL && !(fmask & DTK_M(TZ)))
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
{
/*
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* daylight savings time modifier but no standard timezone? then
* error
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
*/
if (fmask & DTK_M(DTZMOD))
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
return DTERR_BAD_FORMAT;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
Switch over to using the src/timezone functions for formatting timestamps displayed in the postmaster log. This avoids Windows-specific problems with localized time zone names that are in the wrong encoding, and generally seems like a good idea to forestall other potential platform-dependent issues. To preserve the existing behavior that all backends will log in the same time zone, create a new GUC variable log_timezone that can only be changed on a system-wide basis, and reference log-related calculations to that zone instead of the TimeZone variable. This fixes the issue reported by Hiroshi Saito that timestamps printed by xlog.c startup could be improperly localized on Windows. We still need a simpler patch for that problem in the back branches, however.
2007-08-04 03:26:54 +02:00
*tzp = DetermineTimeZoneOffset(tm, session_timezone);
Consolidate several near-identical uses of mktime() into a single routine DetermineLocalTimeZone(). In that routine, be more wary of broken mktime() implementations than the original code was: don't allow mktime to change the already-set y/m/d/h/m/s information, and don't use tm_gmtoff if mktime failed. Possibly this will resolve some of the complaints we've been hearing from users of Middle Eastern timezones on RedHat.
2001-05-04 00:53:07 +02:00
}
}
return 0;
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
}
Consolidate several near-identical uses of mktime() into a single routine DetermineLocalTimeZone(). In that routine, be more wary of broken mktime() implementations than the original code was: don't allow mktime to change the already-set y/m/d/h/m/s information, and don't use tm_gmtoff if mktime failed. Possibly this will resolve some of the complaints we've been hearing from users of Middle Eastern timezones on RedHat.
2001-05-04 00:53:07 +02:00
Andrew pointed out that the current fix didn't handle dates that were near daylight savings time boudaries. This handles it properly, e.g. test=> select '2005-04-03 04:00:00'::timestamp at time zone 'America/Los_Angeles'; timezone ------------------------ 2005-04-03 07:00:00-04 (1 row)
2005-07-23 16:25:34 +02:00
/* DetermineTimeZoneOffset()
Work around mktime() brain damage in recent versions of glibc by using a series of localtime() calls to determine the local timezone offset when mktime() fails. This eliminates regression failures on RHL 7.3, and should continue to work until it occurs to the glibc boys to break localtime() as well. By then I hope we'll have our own timezone code...
2002-09-03 21:46:32 +02:00
*
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
* Given a struct pg_tm in which tm_year, tm_mon, tm_mday, tm_hour, tm_min,
* and tm_sec fields are set, and a zic-style time zone definition, determine
* the applicable GMT offset and daylight-savings status at that time.
* Set the struct pg_tm's tm_isdst field accordingly, and return the GMT
* offset as the function result.
*
* Note: if the date is out of the range we can deal with, we return zero
* as the GMT offset and set tm_isdst = 0. We don't throw an error here,
* though probably some higher-level code will.
*/
int
Initial pgindent run with pg_bsd_indent version 2.0. The new indent version includes numerous fixes thanks to Piotr Stefaniak. The main changes visible in this commit are: * Nicer formatting of function-pointer declarations. * No longer unexpectedly removes spaces in expressions using casts, sizeof, or offsetof. * No longer wants to add a space in "struct structname *varname", as well as some similar cases for const- or volatile-qualified pointers. * Declarations using PG_USED_FOR_ASSERTS_ONLY are formatted more nicely. * Fixes bug where comments following declarations were sometimes placed with no space separating them from the code. * Fixes some odd decisions for comments following case labels. * Fixes some cases where comments following code were indented to less than the expected column 33. On the less good side, it now tends to put more whitespace around typedef names that are not listed in typedefs.list. This might encourage us to put more effort into typedef name collection; it's not really a bug in indent itself. There are more changes coming after this round, having to do with comment indentation and alignment of lines appearing within parentheses. I wanted to limit the size of the diffs to something that could be reviewed without one's eyes completely glazing over, so it seemed better to split up the changes as much as practical. Discussion: https://postgr.es/m/E1dAmxK-0006EE-1r@gemulon.postgresql.org Discussion: https://postgr.es/m/30527.1495162840@sss.pgh.pa.us
2017-06-21 20:39:04 +02:00
DetermineTimeZoneOffset(struct pg_tm *tm, pg_tz *tzp)
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
{
pg_time_t t;
return DetermineTimeZoneOffsetInternal(tm, tzp, &t);
}
/* DetermineTimeZoneOffsetInternal()
*
* As above, but also return the actual UTC time imputed to the date/time
* into *tp.
*
* In event of an out-of-range date, we punt by returning zero into *tp.
* This is okay for the immediate callers but is a good reason for not
* exposing this worker function globally.
Consolidate several near-identical uses of mktime() into a single routine DetermineLocalTimeZone(). In that routine, be more wary of broken mktime() implementations than the original code was: don't allow mktime to change the already-set y/m/d/h/m/s information, and don't use tm_gmtoff if mktime failed. Possibly this will resolve some of the complaints we've been hearing from users of Middle Eastern timezones on RedHat.
2001-05-04 00:53:07 +02:00
*
Adjust our timezone library to use pg_time_t (typedef'd as int64) in place of time_t, as per prior discussion. The behavior does not change on machines without a 64-bit-int type, but on machines with one, which is most, we are rid of the bizarre boundary behavior at the edges of the 32-bit-time_t range (1901 and 2038). The system will now treat times over the full supported timestamp range as being in your local time zone. It may seem a little bizarre to consider that times in 4000 BC are PST or EST, but this is surely at least as reasonable as propagating Gregorian calendar rules back that far. I did not modify the format of the zic timezone database files, which means that for the moment the system will not know about daylight-savings periods outside the range 1901-2038. Given the way the files are set up, it's not a simple decision like 'widen to 64 bits'; we have to actually think about the range of years that need to be supported. We should probably inquire what the plans of the upstream zic people are before making any decisions of our own.
2004-06-03 04:08:07 +02:00
* Note: it might seem that we should use mktime() for this, but bitter
Invent pg_next_dst_boundary() and rewrite DetermineLocalTimeZone() to use it, as per my proposal of yesterday. This gives us a means of determining the zone offset to impute to an unlabeled timestamp that is both efficient and reliable, unlike all our previous tries involving mktime() and localtime(). The behavior for invalid or ambiguous times at a DST transition is fixed to be really and truly "assume standard time", fixing a bug that has come and gone repeatedly but was back again in 7.4. (There is some ongoing discussion about whether we should raise an error instead, but for the moment I'll make it do what it was previously intended to do.)
2004-11-01 22:34:44 +01:00
* experience teaches otherwise. This code is much faster than most versions
* of mktime(), anyway.
Consolidate several near-identical uses of mktime() into a single routine DetermineLocalTimeZone(). In that routine, be more wary of broken mktime() implementations than the original code was: don't allow mktime to change the already-set y/m/d/h/m/s information, and don't use tm_gmtoff if mktime failed. Possibly this will resolve some of the complaints we've been hearing from users of Middle Eastern timezones on RedHat.
2001-05-04 00:53:07 +02:00
*/
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
static int
Initial pgindent run with pg_bsd_indent version 2.0. The new indent version includes numerous fixes thanks to Piotr Stefaniak. The main changes visible in this commit are: * Nicer formatting of function-pointer declarations. * No longer unexpectedly removes spaces in expressions using casts, sizeof, or offsetof. * No longer wants to add a space in "struct structname *varname", as well as some similar cases for const- or volatile-qualified pointers. * Declarations using PG_USED_FOR_ASSERTS_ONLY are formatted more nicely. * Fixes bug where comments following declarations were sometimes placed with no space separating them from the code. * Fixes some odd decisions for comments following case labels. * Fixes some cases where comments following code were indented to less than the expected column 33. On the less good side, it now tends to put more whitespace around typedef names that are not listed in typedefs.list. This might encourage us to put more effort into typedef name collection; it's not really a bug in indent itself. There are more changes coming after this round, having to do with comment indentation and alignment of lines appearing within parentheses. I wanted to limit the size of the diffs to something that could be reviewed without one's eyes completely glazing over, so it seemed better to split up the changes as much as practical. Discussion: https://postgr.es/m/E1dAmxK-0006EE-1r@gemulon.postgresql.org Discussion: https://postgr.es/m/30527.1495162840@sss.pgh.pa.us
2017-06-21 20:39:04 +02:00
DetermineTimeZoneOffsetInternal(struct pg_tm *tm, pg_tz *tzp, pg_time_t *tp)
Consolidate several near-identical uses of mktime() into a single routine DetermineLocalTimeZone(). In that routine, be more wary of broken mktime() implementations than the original code was: don't allow mktime to change the already-set y/m/d/h/m/s information, and don't use tm_gmtoff if mktime failed. Possibly this will resolve some of the complaints we've been hearing from users of Middle Eastern timezones on RedHat.
2001-05-04 00:53:07 +02:00
{
Invent pg_next_dst_boundary() and rewrite DetermineLocalTimeZone() to use it, as per my proposal of yesterday. This gives us a means of determining the zone offset to impute to an unlabeled timestamp that is both efficient and reliable, unlike all our previous tries involving mktime() and localtime(). The behavior for invalid or ambiguous times at a DST transition is fixed to be really and truly "assume standard time", fixing a bug that has come and gone repeatedly but was back again in 7.4. (There is some ongoing discussion about whether we should raise an error instead, but for the moment I'll make it do what it was previously intended to do.)
2004-11-01 22:34:44 +01:00
int date,
Adjust our timezone library to use pg_time_t (typedef'd as int64) in place of time_t, as per prior discussion. The behavior does not change on machines without a 64-bit-int type, but on machines with one, which is most, we are rid of the bizarre boundary behavior at the edges of the 32-bit-time_t range (1901 and 2038). The system will now treat times over the full supported timestamp range as being in your local time zone. It may seem a little bizarre to consider that times in 4000 BC are PST or EST, but this is surely at least as reasonable as propagating Gregorian calendar rules back that far. I did not modify the format of the zic timezone database files, which means that for the moment the system will not know about daylight-savings periods outside the range 1901-2038. Given the way the files are set up, it's not a simple decision like 'widen to 64 bits'; we have to actually think about the range of years that need to be supported. We should probably inquire what the plans of the upstream zic people are before making any decisions of our own.
2004-06-03 04:08:07 +02:00
sec;
pg_time_t day,
Invent pg_next_dst_boundary() and rewrite DetermineLocalTimeZone() to use it, as per my proposal of yesterday. This gives us a means of determining the zone offset to impute to an unlabeled timestamp that is both efficient and reliable, unlike all our previous tries involving mktime() and localtime(). The behavior for invalid or ambiguous times at a DST transition is fixed to be really and truly "assume standard time", fixing a bug that has come and gone repeatedly but was back again in 7.4. (There is some ongoing discussion about whether we should raise an error instead, but for the moment I'll make it do what it was previously intended to do.)
2004-11-01 22:34:44 +01:00
mytime,
prevtime,
boundary,
beforetime,
aftertime;
long int before_gmtoff,
after_gmtoff;
int before_isdst,
after_isdst;
int res;
Consolidate several near-identical uses of mktime() into a single routine DetermineLocalTimeZone(). In that routine, be more wary of broken mktime() implementations than the original code was: don't allow mktime to change the already-set y/m/d/h/m/s information, and don't use tm_gmtoff if mktime failed. Possibly this will resolve some of the complaints we've been hearing from users of Middle Eastern timezones on RedHat.
2001-05-04 00:53:07 +02:00
Adjust our timezone library to use pg_time_t (typedef'd as int64) in place of time_t, as per prior discussion. The behavior does not change on machines without a 64-bit-int type, but on machines with one, which is most, we are rid of the bizarre boundary behavior at the edges of the 32-bit-time_t range (1901 and 2038). The system will now treat times over the full supported timestamp range as being in your local time zone. It may seem a little bizarre to consider that times in 4000 BC are PST or EST, but this is surely at least as reasonable as propagating Gregorian calendar rules back that far. I did not modify the format of the zic timezone database files, which means that for the moment the system will not know about daylight-savings periods outside the range 1901-2038. Given the way the files are set up, it's not a simple decision like 'widen to 64 bits'; we have to actually think about the range of years that need to be supported. We should probably inquire what the plans of the upstream zic people are before making any decisions of our own.
2004-06-03 04:08:07 +02:00
/*
* First, generate the pg_time_t value corresponding to the given
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* y/m/d/h/m/s taken as GMT time. If this overflows, punt and decide the
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
* timezone is GMT. (For a valid Julian date, integer overflow should be
* impossible with 64-bit pg_time_t, but let's check for safety.)
Adjust our timezone library to use pg_time_t (typedef'd as int64) in place of time_t, as per prior discussion. The behavior does not change on machines without a 64-bit-int type, but on machines with one, which is most, we are rid of the bizarre boundary behavior at the edges of the 32-bit-time_t range (1901 and 2038). The system will now treat times over the full supported timestamp range as being in your local time zone. It may seem a little bizarre to consider that times in 4000 BC are PST or EST, but this is surely at least as reasonable as propagating Gregorian calendar rules back that far. I did not modify the format of the zic timezone database files, which means that for the moment the system will not know about daylight-savings periods outside the range 1901-2038. Given the way the files are set up, it's not a simple decision like 'widen to 64 bits'; we have to actually think about the range of years that need to be supported. We should probably inquire what the plans of the upstream zic people are before making any decisions of our own.
2004-06-03 04:08:07 +02:00
*/
if (!IS_VALID_JULIAN(tm->tm_year, tm->tm_mon, tm->tm_mday))
goto overflow;
date = date2j(tm->tm_year, tm->tm_mon, tm->tm_mday) - UNIX_EPOCH_JDATE;
Another pgindent run with lib typedefs added.
2004-08-30 04:54:42 +02:00
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
day = ((pg_time_t) date) * SECS_PER_DAY;
More macro cleanups for date/time.
2005-05-23 23:54:02 +02:00
if (day / SECS_PER_DAY != date)
Adjust our timezone library to use pg_time_t (typedef'd as int64) in place of time_t, as per prior discussion. The behavior does not change on machines without a 64-bit-int type, but on machines with one, which is most, we are rid of the bizarre boundary behavior at the edges of the 32-bit-time_t range (1901 and 2038). The system will now treat times over the full supported timestamp range as being in your local time zone. It may seem a little bizarre to consider that times in 4000 BC are PST or EST, but this is surely at least as reasonable as propagating Gregorian calendar rules back that far. I did not modify the format of the zic timezone database files, which means that for the moment the system will not know about daylight-savings periods outside the range 1901-2038. Given the way the files are set up, it's not a simple decision like 'widen to 64 bits'; we have to actually think about the range of years that need to be supported. We should probably inquire what the plans of the upstream zic people are before making any decisions of our own.
2004-06-03 04:08:07 +02:00
goto overflow;
Update DAYS_PER_MONTH comment. Add SECS_PER_YEAR and MINS_PER_HOUR macros.
2005-07-21 20:06:13 +02:00
sec = tm->tm_sec + (tm->tm_min + tm->tm_hour * MINS_PER_HOUR) * SECS_PER_MINUTE;
Invent pg_next_dst_boundary() and rewrite DetermineLocalTimeZone() to use it, as per my proposal of yesterday. This gives us a means of determining the zone offset to impute to an unlabeled timestamp that is both efficient and reliable, unlike all our previous tries involving mktime() and localtime(). The behavior for invalid or ambiguous times at a DST transition is fixed to be really and truly "assume standard time", fixing a bug that has come and gone repeatedly but was back again in 7.4. (There is some ongoing discussion about whether we should raise an error instead, but for the moment I'll make it do what it was previously intended to do.)
2004-11-01 22:34:44 +01:00
mytime = day + sec;
/* since sec >= 0, overflow could only be from +day to -mytime */
if (mytime < 0 && day > 0)
Adjust our timezone library to use pg_time_t (typedef'd as int64) in place of time_t, as per prior discussion. The behavior does not change on machines without a 64-bit-int type, but on machines with one, which is most, we are rid of the bizarre boundary behavior at the edges of the 32-bit-time_t range (1901 and 2038). The system will now treat times over the full supported timestamp range as being in your local time zone. It may seem a little bizarre to consider that times in 4000 BC are PST or EST, but this is surely at least as reasonable as propagating Gregorian calendar rules back that far. I did not modify the format of the zic timezone database files, which means that for the moment the system will not know about daylight-savings periods outside the range 1901-2038. Given the way the files are set up, it's not a simple decision like 'widen to 64 bits'; we have to actually think about the range of years that need to be supported. We should probably inquire what the plans of the upstream zic people are before making any decisions of our own.
2004-06-03 04:08:07 +02:00
goto overflow;
Consolidate several near-identical uses of mktime() into a single routine DetermineLocalTimeZone(). In that routine, be more wary of broken mktime() implementations than the original code was: don't allow mktime to change the already-set y/m/d/h/m/s information, and don't use tm_gmtoff if mktime failed. Possibly this will resolve some of the complaints we've been hearing from users of Middle Eastern timezones on RedHat.
2001-05-04 00:53:07 +02:00
Adjust our timezone library to use pg_time_t (typedef'd as int64) in place of time_t, as per prior discussion. The behavior does not change on machines without a 64-bit-int type, but on machines with one, which is most, we are rid of the bizarre boundary behavior at the edges of the 32-bit-time_t range (1901 and 2038). The system will now treat times over the full supported timestamp range as being in your local time zone. It may seem a little bizarre to consider that times in 4000 BC are PST or EST, but this is surely at least as reasonable as propagating Gregorian calendar rules back that far. I did not modify the format of the zic timezone database files, which means that for the moment the system will not know about daylight-savings periods outside the range 1901-2038. Given the way the files are set up, it's not a simple decision like 'widen to 64 bits'; we have to actually think about the range of years that need to be supported. We should probably inquire what the plans of the upstream zic people are before making any decisions of our own.
2004-06-03 04:08:07 +02:00
/*
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* Find the DST time boundary just before or following the target time. We
* assume that all zones have GMT offsets less than 24 hours, and that DST
* boundaries can't be closer together than 48 hours, so backing up 24
* hours and finding the "next" boundary will work.
Adjust our timezone library to use pg_time_t (typedef'd as int64) in place of time_t, as per prior discussion. The behavior does not change on machines without a 64-bit-int type, but on machines with one, which is most, we are rid of the bizarre boundary behavior at the edges of the 32-bit-time_t range (1901 and 2038). The system will now treat times over the full supported timestamp range as being in your local time zone. It may seem a little bizarre to consider that times in 4000 BC are PST or EST, but this is surely at least as reasonable as propagating Gregorian calendar rules back that far. I did not modify the format of the zic timezone database files, which means that for the moment the system will not know about daylight-savings periods outside the range 1901-2038. Given the way the files are set up, it's not a simple decision like 'widen to 64 bits'; we have to actually think about the range of years that need to be supported. We should probably inquire what the plans of the upstream zic people are before making any decisions of our own.
2004-06-03 04:08:07 +02:00
*/
Update DAYS_PER_MONTH comment. Add SECS_PER_YEAR and MINS_PER_HOUR macros.
2005-07-21 20:06:13 +02:00
prevtime = mytime - SECS_PER_DAY;
Invent pg_next_dst_boundary() and rewrite DetermineLocalTimeZone() to use it, as per my proposal of yesterday. This gives us a means of determining the zone offset to impute to an unlabeled timestamp that is both efficient and reliable, unlike all our previous tries involving mktime() and localtime(). The behavior for invalid or ambiguous times at a DST transition is fixed to be really and truly "assume standard time", fixing a bug that has come and gone repeatedly but was back again in 7.4. (There is some ongoing discussion about whether we should raise an error instead, but for the moment I'll make it do what it was previously intended to do.)
2004-11-01 22:34:44 +01:00
if (mytime < 0 && prevtime > 0)
goto overflow;
Adjust our timezone library to use pg_time_t (typedef'd as int64) in place of time_t, as per prior discussion. The behavior does not change on machines without a 64-bit-int type, but on machines with one, which is most, we are rid of the bizarre boundary behavior at the edges of the 32-bit-time_t range (1901 and 2038). The system will now treat times over the full supported timestamp range as being in your local time zone. It may seem a little bizarre to consider that times in 4000 BC are PST or EST, but this is surely at least as reasonable as propagating Gregorian calendar rules back that far. I did not modify the format of the zic timezone database files, which means that for the moment the system will not know about daylight-savings periods outside the range 1901-2038. Given the way the files are set up, it's not a simple decision like 'widen to 64 bits'; we have to actually think about the range of years that need to be supported. We should probably inquire what the plans of the upstream zic people are before making any decisions of our own.
2004-06-03 04:08:07 +02:00
Invent pg_next_dst_boundary() and rewrite DetermineLocalTimeZone() to use it, as per my proposal of yesterday. This gives us a means of determining the zone offset to impute to an unlabeled timestamp that is both efficient and reliable, unlike all our previous tries involving mktime() and localtime(). The behavior for invalid or ambiguous times at a DST transition is fixed to be really and truly "assume standard time", fixing a bug that has come and gone repeatedly but was back again in 7.4. (There is some ongoing discussion about whether we should raise an error instead, but for the moment I'll make it do what it was previously intended to do.)
2004-11-01 22:34:44 +01:00
res = pg_next_dst_boundary(&prevtime,
&before_gmtoff, &before_isdst,
&boundary,
Attached patch gets rid of the global timezone in the following steps: * Changes the APIs to the timezone functions to take a pg_tz pointer as an argument, representing the timezone to use for the selected operation. * Adds a global_timezone variable that represents the current timezone in the backend as set by SET TIMEZONE (or guc, or env, etc). * Implements a hash-table cache of loaded tables, so we don't have to read and parse the TZ file everytime we change a timezone. While not necesasry now (we don't change timezones very often), I beleive this will be necessary (or at least good) when "multiple timezones in the same query" is eventually implemented. And code-wise, this was the time to do it. There are no user-visible changes at this time. Implementing the "multiple zones in one query" is a later step... This also gets rid of some of the cruft needed to "back out a timezone change", since we previously couldn't check a timezone unless it was activated first. Passes regression tests on win32, linux (slackware 10) and solaris x86. Magnus Hagander
2005-04-19 05:13:59 +02:00
&after_gmtoff, &after_isdst,
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
tzp);
Invent pg_next_dst_boundary() and rewrite DetermineLocalTimeZone() to use it, as per my proposal of yesterday. This gives us a means of determining the zone offset to impute to an unlabeled timestamp that is both efficient and reliable, unlike all our previous tries involving mktime() and localtime(). The behavior for invalid or ambiguous times at a DST transition is fixed to be really and truly "assume standard time", fixing a bug that has come and gone repeatedly but was back again in 7.4. (There is some ongoing discussion about whether we should raise an error instead, but for the moment I'll make it do what it was previously intended to do.)
2004-11-01 22:34:44 +01:00
if (res < 0)
goto overflow; /* failure? */
if (res == 0)
{
/* Non-DST zone, life is simple */
tm->tm_isdst = before_isdst;
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
*tp = mytime - before_gmtoff;
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
return -(int) before_gmtoff;
Invent pg_next_dst_boundary() and rewrite DetermineLocalTimeZone() to use it, as per my proposal of yesterday. This gives us a means of determining the zone offset to impute to an unlabeled timestamp that is both efficient and reliable, unlike all our previous tries involving mktime() and localtime(). The behavior for invalid or ambiguous times at a DST transition is fixed to be really and truly "assume standard time", fixing a bug that has come and gone repeatedly but was back again in 7.4. (There is some ongoing discussion about whether we should raise an error instead, but for the moment I'll make it do what it was previously intended to do.)
2004-11-01 22:34:44 +01:00
}
Adjust our timezone library to use pg_time_t (typedef'd as int64) in place of time_t, as per prior discussion. The behavior does not change on machines without a 64-bit-int type, but on machines with one, which is most, we are rid of the bizarre boundary behavior at the edges of the 32-bit-time_t range (1901 and 2038). The system will now treat times over the full supported timestamp range as being in your local time zone. It may seem a little bizarre to consider that times in 4000 BC are PST or EST, but this is surely at least as reasonable as propagating Gregorian calendar rules back that far. I did not modify the format of the zic timezone database files, which means that for the moment the system will not know about daylight-savings periods outside the range 1901-2038. Given the way the files are set up, it's not a simple decision like 'widen to 64 bits'; we have to actually think about the range of years that need to be supported. We should probably inquire what the plans of the upstream zic people are before making any decisions of our own.
2004-06-03 04:08:07 +02:00
/*
Invent pg_next_dst_boundary() and rewrite DetermineLocalTimeZone() to use it, as per my proposal of yesterday. This gives us a means of determining the zone offset to impute to an unlabeled timestamp that is both efficient and reliable, unlike all our previous tries involving mktime() and localtime(). The behavior for invalid or ambiguous times at a DST transition is fixed to be really and truly "assume standard time", fixing a bug that has come and gone repeatedly but was back again in 7.4. (There is some ongoing discussion about whether we should raise an error instead, but for the moment I'll make it do what it was previously intended to do.)
2004-11-01 22:34:44 +01:00
* Form the candidate pg_time_t values with local-time adjustment
Adjust our timezone library to use pg_time_t (typedef'd as int64) in place of time_t, as per prior discussion. The behavior does not change on machines without a 64-bit-int type, but on machines with one, which is most, we are rid of the bizarre boundary behavior at the edges of the 32-bit-time_t range (1901 and 2038). The system will now treat times over the full supported timestamp range as being in your local time zone. It may seem a little bizarre to consider that times in 4000 BC are PST or EST, but this is surely at least as reasonable as propagating Gregorian calendar rules back that far. I did not modify the format of the zic timezone database files, which means that for the moment the system will not know about daylight-savings periods outside the range 1901-2038. Given the way the files are set up, it's not a simple decision like 'widen to 64 bits'; we have to actually think about the range of years that need to be supported. We should probably inquire what the plans of the upstream zic people are before making any decisions of our own.
2004-06-03 04:08:07 +02:00
*/
Invent pg_next_dst_boundary() and rewrite DetermineLocalTimeZone() to use it, as per my proposal of yesterday. This gives us a means of determining the zone offset to impute to an unlabeled timestamp that is both efficient and reliable, unlike all our previous tries involving mktime() and localtime(). The behavior for invalid or ambiguous times at a DST transition is fixed to be really and truly "assume standard time", fixing a bug that has come and gone repeatedly but was back again in 7.4. (There is some ongoing discussion about whether we should raise an error instead, but for the moment I'll make it do what it was previously intended to do.)
2004-11-01 22:34:44 +01:00
beforetime = mytime - before_gmtoff;
Remove unnecessary parentheses in datetime/timestamp code.
2005-05-23 19:13:14 +02:00
if ((before_gmtoff > 0 &&
mytime < 0 && beforetime > 0) ||
(before_gmtoff <= 0 &&
mytime > 0 && beforetime < 0))
Invent pg_next_dst_boundary() and rewrite DetermineLocalTimeZone() to use it, as per my proposal of yesterday. This gives us a means of determining the zone offset to impute to an unlabeled timestamp that is both efficient and reliable, unlike all our previous tries involving mktime() and localtime(). The behavior for invalid or ambiguous times at a DST transition is fixed to be really and truly "assume standard time", fixing a bug that has come and gone repeatedly but was back again in 7.4. (There is some ongoing discussion about whether we should raise an error instead, but for the moment I'll make it do what it was previously intended to do.)
2004-11-01 22:34:44 +01:00
goto overflow;
aftertime = mytime - after_gmtoff;
Remove unnecessary parentheses in datetime/timestamp code.
2005-05-23 19:13:14 +02:00
if ((after_gmtoff > 0 &&
mytime < 0 && aftertime > 0) ||
(after_gmtoff <= 0 &&
mytime > 0 && aftertime < 0))
Adjust our timezone library to use pg_time_t (typedef'd as int64) in place of time_t, as per prior discussion. The behavior does not change on machines without a 64-bit-int type, but on machines with one, which is most, we are rid of the bizarre boundary behavior at the edges of the 32-bit-time_t range (1901 and 2038). The system will now treat times over the full supported timestamp range as being in your local time zone. It may seem a little bizarre to consider that times in 4000 BC are PST or EST, but this is surely at least as reasonable as propagating Gregorian calendar rules back that far. I did not modify the format of the zic timezone database files, which means that for the moment the system will not know about daylight-savings periods outside the range 1901-2038. Given the way the files are set up, it's not a simple decision like 'widen to 64 bits'; we have to actually think about the range of years that need to be supported. We should probably inquire what the plans of the upstream zic people are before making any decisions of our own.
2004-06-03 04:08:07 +02:00
goto overflow;
/*
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
* If both before or both after the boundary time, we know what to do. The
* boundary time itself is considered to be after the transition, which
* means we can accept aftertime == boundary in the second case.
Adjust our timezone library to use pg_time_t (typedef'd as int64) in place of time_t, as per prior discussion. The behavior does not change on machines without a 64-bit-int type, but on machines with one, which is most, we are rid of the bizarre boundary behavior at the edges of the 32-bit-time_t range (1901 and 2038). The system will now treat times over the full supported timestamp range as being in your local time zone. It may seem a little bizarre to consider that times in 4000 BC are PST or EST, but this is surely at least as reasonable as propagating Gregorian calendar rules back that far. I did not modify the format of the zic timezone database files, which means that for the moment the system will not know about daylight-savings periods outside the range 1901-2038. Given the way the files are set up, it's not a simple decision like 'widen to 64 bits'; we have to actually think about the range of years that need to be supported. We should probably inquire what the plans of the upstream zic people are before making any decisions of our own.
2004-06-03 04:08:07 +02:00
*/
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
if (beforetime < boundary && aftertime < boundary)
Adjust our timezone library to use pg_time_t (typedef'd as int64) in place of time_t, as per prior discussion. The behavior does not change on machines without a 64-bit-int type, but on machines with one, which is most, we are rid of the bizarre boundary behavior at the edges of the 32-bit-time_t range (1901 and 2038). The system will now treat times over the full supported timestamp range as being in your local time zone. It may seem a little bizarre to consider that times in 4000 BC are PST or EST, but this is surely at least as reasonable as propagating Gregorian calendar rules back that far. I did not modify the format of the zic timezone database files, which means that for the moment the system will not know about daylight-savings periods outside the range 1901-2038. Given the way the files are set up, it's not a simple decision like 'widen to 64 bits'; we have to actually think about the range of years that need to be supported. We should probably inquire what the plans of the upstream zic people are before making any decisions of our own.
2004-06-03 04:08:07 +02:00
{
Invent pg_next_dst_boundary() and rewrite DetermineLocalTimeZone() to use it, as per my proposal of yesterday. This gives us a means of determining the zone offset to impute to an unlabeled timestamp that is both efficient and reliable, unlike all our previous tries involving mktime() and localtime(). The behavior for invalid or ambiguous times at a DST transition is fixed to be really and truly "assume standard time", fixing a bug that has come and gone repeatedly but was back again in 7.4. (There is some ongoing discussion about whether we should raise an error instead, but for the moment I'll make it do what it was previously intended to do.)
2004-11-01 22:34:44 +01:00
tm->tm_isdst = before_isdst;
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
*tp = beforetime;
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
return -(int) before_gmtoff;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
}
Invent pg_next_dst_boundary() and rewrite DetermineLocalTimeZone() to use it, as per my proposal of yesterday. This gives us a means of determining the zone offset to impute to an unlabeled timestamp that is both efficient and reliable, unlike all our previous tries involving mktime() and localtime(). The behavior for invalid or ambiguous times at a DST transition is fixed to be really and truly "assume standard time", fixing a bug that has come and gone repeatedly but was back again in 7.4. (There is some ongoing discussion about whether we should raise an error instead, but for the moment I'll make it do what it was previously intended to do.)
2004-11-01 22:34:44 +01:00
if (beforetime > boundary && aftertime >= boundary)
{
tm->tm_isdst = after_isdst;
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
*tp = aftertime;
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
return -(int) after_gmtoff;
Invent pg_next_dst_boundary() and rewrite DetermineLocalTimeZone() to use it, as per my proposal of yesterday. This gives us a means of determining the zone offset to impute to an unlabeled timestamp that is both efficient and reliable, unlike all our previous tries involving mktime() and localtime(). The behavior for invalid or ambiguous times at a DST transition is fixed to be really and truly "assume standard time", fixing a bug that has come and gone repeatedly but was back again in 7.4. (There is some ongoing discussion about whether we should raise an error instead, but for the moment I'll make it do what it was previously intended to do.)
2004-11-01 22:34:44 +01:00
}
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
Invent pg_next_dst_boundary() and rewrite DetermineLocalTimeZone() to use it, as per my proposal of yesterday. This gives us a means of determining the zone offset to impute to an unlabeled timestamp that is both efficient and reliable, unlike all our previous tries involving mktime() and localtime(). The behavior for invalid or ambiguous times at a DST transition is fixed to be really and truly "assume standard time", fixing a bug that has come and gone repeatedly but was back again in 7.4. (There is some ongoing discussion about whether we should raise an error instead, but for the moment I'll make it do what it was previously intended to do.)
2004-11-01 22:34:44 +01:00
/*
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
* It's an invalid or ambiguous time due to timezone transition. In a
* spring-forward transition, prefer the "before" interpretation; in a
* fall-back transition, prefer "after". (We used to define and implement
* this test as "prefer the standard-time interpretation", but that rule
* does not help to resolve the behavior when both times are reported as
Update time zone data files to tzdata release 2018e. DST law changes in North Korea. Redefinition of "daylight savings" in Ireland, as well as for some past years in Namibia and Czechoslovakia. Additional historical corrections for Czechoslovakia. With this change, the IANA database models Irish timekeeping as following "standard time" in summer, and "daylight savings" in winter, so that the daylight savings offset is one hour behind standard time not one hour ahead. This does not change their UTC offset (+1:00 in summer, 0:00 in winter) nor their timezone abbreviations (IST in summer, GMT in winter), though now "IST" is more correctly read as "Irish Standard Time" not "Irish Summer Time". However, the "is_dst" column in the pg_timezone_names view will now be true in winter and false in summer for the Europe/Dublin zone. Similar changes were made for Namibia between 1994 and 2017, and for Czechoslovakia between 1946 and 1947. So far as I can find, no Postgres internal logic cares about which way tm_isdst is reported; in particular, since commit b2cbced9e we do not rely on it to decide how to interpret ambiguous timestamps during DST transitions. So I don't think this change will affect any Postgres behavior other than the timezone-view outputs. Discussion: https://postgr.es/m/30996.1525445902@sss.pgh.pa.us
2018-05-09 19:55:27 +02:00
* standard time; which does happen, eg Europe/Moscow in Oct 2014. Also,
* in some zones such as Europe/Dublin, there is widespread confusion
* about which time offset is "standard" time, so it's fortunate that our
* behavior doesn't depend on that.)
Invent pg_next_dst_boundary() and rewrite DetermineLocalTimeZone() to use it, as per my proposal of yesterday. This gives us a means of determining the zone offset to impute to an unlabeled timestamp that is both efficient and reliable, unlike all our previous tries involving mktime() and localtime(). The behavior for invalid or ambiguous times at a DST transition is fixed to be really and truly "assume standard time", fixing a bug that has come and gone repeatedly but was back again in 7.4. (There is some ongoing discussion about whether we should raise an error instead, but for the moment I'll make it do what it was previously intended to do.)
2004-11-01 22:34:44 +01:00
*/
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
if (beforetime > aftertime)
Invent pg_next_dst_boundary() and rewrite DetermineLocalTimeZone() to use it, as per my proposal of yesterday. This gives us a means of determining the zone offset to impute to an unlabeled timestamp that is both efficient and reliable, unlike all our previous tries involving mktime() and localtime(). The behavior for invalid or ambiguous times at a DST transition is fixed to be really and truly "assume standard time", fixing a bug that has come and gone repeatedly but was back again in 7.4. (There is some ongoing discussion about whether we should raise an error instead, but for the moment I'll make it do what it was previously intended to do.)
2004-11-01 22:34:44 +01:00
{
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
tm->tm_isdst = before_isdst;
*tp = beforetime;
return -(int) before_gmtoff;
Invent pg_next_dst_boundary() and rewrite DetermineLocalTimeZone() to use it, as per my proposal of yesterday. This gives us a means of determining the zone offset to impute to an unlabeled timestamp that is both efficient and reliable, unlike all our previous tries involving mktime() and localtime(). The behavior for invalid or ambiguous times at a DST transition is fixed to be really and truly "assume standard time", fixing a bug that has come and gone repeatedly but was back again in 7.4. (There is some ongoing discussion about whether we should raise an error instead, but for the moment I'll make it do what it was previously intended to do.)
2004-11-01 22:34:44 +01:00
}
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
tm->tm_isdst = after_isdst;
*tp = aftertime;
return -(int) after_gmtoff;
Adjust our timezone library to use pg_time_t (typedef'd as int64) in place of time_t, as per prior discussion. The behavior does not change on machines without a 64-bit-int type, but on machines with one, which is most, we are rid of the bizarre boundary behavior at the edges of the 32-bit-time_t range (1901 and 2038). The system will now treat times over the full supported timestamp range as being in your local time zone. It may seem a little bizarre to consider that times in 4000 BC are PST or EST, but this is surely at least as reasonable as propagating Gregorian calendar rules back that far. I did not modify the format of the zic timezone database files, which means that for the moment the system will not know about daylight-savings periods outside the range 1901-2038. Given the way the files are set up, it's not a simple decision like 'widen to 64 bits'; we have to actually think about the range of years that need to be supported. We should probably inquire what the plans of the upstream zic people are before making any decisions of our own.
2004-06-03 04:08:07 +02:00
overflow:
/* Given date is out of range, so assume UTC */
tm->tm_isdst = 0;
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
*tp = 0;
Adjust our timezone library to use pg_time_t (typedef'd as int64) in place of time_t, as per prior discussion. The behavior does not change on machines without a 64-bit-int type, but on machines with one, which is most, we are rid of the bizarre boundary behavior at the edges of the 32-bit-time_t range (1901 and 2038). The system will now treat times over the full supported timestamp range as being in your local time zone. It may seem a little bizarre to consider that times in 4000 BC are PST or EST, but this is surely at least as reasonable as propagating Gregorian calendar rules back that far. I did not modify the format of the zic timezone database files, which means that for the moment the system will not know about daylight-savings periods outside the range 1901-2038. Given the way the files are set up, it's not a simple decision like 'widen to 64 bits'; we have to actually think about the range of years that need to be supported. We should probably inquire what the plans of the upstream zic people are before making any decisions of our own.
2004-06-03 04:08:07 +02:00
return 0;
Consolidate several near-identical uses of mktime() into a single routine DetermineLocalTimeZone(). In that routine, be more wary of broken mktime() implementations than the original code was: don't allow mktime to change the already-set y/m/d/h/m/s information, and don't use tm_gmtoff if mktime failed. Possibly this will resolve some of the complaints we've been hearing from users of Middle Eastern timezones on RedHat.
2001-05-04 00:53:07 +02:00
}
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
/* DetermineTimeZoneAbbrevOffset()
*
* Determine the GMT offset and DST flag to be attributed to a dynamic
* time zone abbreviation, that is one whose meaning has changed over time.
* *tm contains the local time at which the meaning should be determined,
* and tm->tm_isdst receives the DST flag.
*
* This differs from the behavior of DetermineTimeZoneOffset() in that a
* standard-time or daylight-time abbreviation forces use of the corresponding
* GMT offset even when the zone was then in DS or standard time respectively.
Don't require dynamic timezone abbreviations to match underlying time zone. Previously, we threw an error if a dynamic timezone abbreviation did not match any abbreviation recorded in the referenced IANA time zone entry. That seemed like a good consistency check at the time, but it turns out that a number of the abbreviations in the IANA database are things that Olson and crew made up out of whole cloth. Their current policy is to remove such names in favor of using simple numeric offsets. Perhaps unsurprisingly, a lot of these made-up abbreviations have varied in meaning over time, which meant that our commit b2cbced9e and later changes made them into dynamic abbreviations. So with newer IANA database versions that don't mention these abbreviations at all, we fail, as reported in bug #14307 from Neil Anderson. It's worse than just a few unused-in-the-wild abbreviations not working, because the pg_timezone_abbrevs view stops working altogether (since its underlying function tries to compute the whole view result in one call). We considered deleting these abbreviations from our abbreviations list, but the problem with that is that we can't stay ahead of possible future IANA changes. Instead, let's leave the abbreviations list alone, and treat any "orphaned" dynamic abbreviation as just meaning the referenced time zone. It will behave a bit differently than it used to, in that you can't any longer override the zone's standard vs. daylight rule by using the "wrong" abbreviation of a pair, but that's better than failing entirely. (Also, this solution can be interpreted as adding a small new feature, which is that any abbreviation a user wants can be defined as referencing a time zone name.) Back-patch to all supported branches, since this problem affects all of them when using tzdata 2016f or newer. Report: <20160902031551.15674.67337@wrigleys.postgresql.org> Discussion: <6189.1472820913@sss.pgh.pa.us>
2016-09-02 23:29:31 +02:00
* (However, that happens only if we can match the given abbreviation to some
* abbreviation that appears in the IANA timezone data. Otherwise, we fall
* back to doing DetermineTimeZoneOffset().)
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
*/
int
Initial pgindent run with pg_bsd_indent version 2.0. The new indent version includes numerous fixes thanks to Piotr Stefaniak. The main changes visible in this commit are: * Nicer formatting of function-pointer declarations. * No longer unexpectedly removes spaces in expressions using casts, sizeof, or offsetof. * No longer wants to add a space in "struct structname *varname", as well as some similar cases for const- or volatile-qualified pointers. * Declarations using PG_USED_FOR_ASSERTS_ONLY are formatted more nicely. * Fixes bug where comments following declarations were sometimes placed with no space separating them from the code. * Fixes some odd decisions for comments following case labels. * Fixes some cases where comments following code were indented to less than the expected column 33. On the less good side, it now tends to put more whitespace around typedef names that are not listed in typedefs.list. This might encourage us to put more effort into typedef name collection; it's not really a bug in indent itself. There are more changes coming after this round, having to do with comment indentation and alignment of lines appearing within parentheses. I wanted to limit the size of the diffs to something that could be reviewed without one's eyes completely glazing over, so it seemed better to split up the changes as much as practical. Discussion: https://postgr.es/m/E1dAmxK-0006EE-1r@gemulon.postgresql.org Discussion: https://postgr.es/m/30527.1495162840@sss.pgh.pa.us
2017-06-21 20:39:04 +02:00
DetermineTimeZoneAbbrevOffset(struct pg_tm *tm, const char *abbr, pg_tz *tzp)
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
{
pg_time_t t;
Don't require dynamic timezone abbreviations to match underlying time zone. Previously, we threw an error if a dynamic timezone abbreviation did not match any abbreviation recorded in the referenced IANA time zone entry. That seemed like a good consistency check at the time, but it turns out that a number of the abbreviations in the IANA database are things that Olson and crew made up out of whole cloth. Their current policy is to remove such names in favor of using simple numeric offsets. Perhaps unsurprisingly, a lot of these made-up abbreviations have varied in meaning over time, which meant that our commit b2cbced9e and later changes made them into dynamic abbreviations. So with newer IANA database versions that don't mention these abbreviations at all, we fail, as reported in bug #14307 from Neil Anderson. It's worse than just a few unused-in-the-wild abbreviations not working, because the pg_timezone_abbrevs view stops working altogether (since its underlying function tries to compute the whole view result in one call). We considered deleting these abbreviations from our abbreviations list, but the problem with that is that we can't stay ahead of possible future IANA changes. Instead, let's leave the abbreviations list alone, and treat any "orphaned" dynamic abbreviation as just meaning the referenced time zone. It will behave a bit differently than it used to, in that you can't any longer override the zone's standard vs. daylight rule by using the "wrong" abbreviation of a pair, but that's better than failing entirely. (Also, this solution can be interpreted as adding a small new feature, which is that any abbreviation a user wants can be defined as referencing a time zone name.) Back-patch to all supported branches, since this problem affects all of them when using tzdata 2016f or newer. Report: <20160902031551.15674.67337@wrigleys.postgresql.org> Discussion: <6189.1472820913@sss.pgh.pa.us>
2016-09-02 23:29:31 +02:00
int zone_offset;
int abbr_offset;
int abbr_isdst;
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
/*
* Compute the UTC time we want to probe at. (In event of overflow, we'll
* probe at the epoch, which is a bit random but probably doesn't matter.)
*/
Don't require dynamic timezone abbreviations to match underlying time zone. Previously, we threw an error if a dynamic timezone abbreviation did not match any abbreviation recorded in the referenced IANA time zone entry. That seemed like a good consistency check at the time, but it turns out that a number of the abbreviations in the IANA database are things that Olson and crew made up out of whole cloth. Their current policy is to remove such names in favor of using simple numeric offsets. Perhaps unsurprisingly, a lot of these made-up abbreviations have varied in meaning over time, which meant that our commit b2cbced9e and later changes made them into dynamic abbreviations. So with newer IANA database versions that don't mention these abbreviations at all, we fail, as reported in bug #14307 from Neil Anderson. It's worse than just a few unused-in-the-wild abbreviations not working, because the pg_timezone_abbrevs view stops working altogether (since its underlying function tries to compute the whole view result in one call). We considered deleting these abbreviations from our abbreviations list, but the problem with that is that we can't stay ahead of possible future IANA changes. Instead, let's leave the abbreviations list alone, and treat any "orphaned" dynamic abbreviation as just meaning the referenced time zone. It will behave a bit differently than it used to, in that you can't any longer override the zone's standard vs. daylight rule by using the "wrong" abbreviation of a pair, but that's better than failing entirely. (Also, this solution can be interpreted as adding a small new feature, which is that any abbreviation a user wants can be defined as referencing a time zone name.) Back-patch to all supported branches, since this problem affects all of them when using tzdata 2016f or newer. Report: <20160902031551.15674.67337@wrigleys.postgresql.org> Discussion: <6189.1472820913@sss.pgh.pa.us>
2016-09-02 23:29:31 +02:00
zone_offset = DetermineTimeZoneOffsetInternal(tm, tzp, &t);
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
Don't require dynamic timezone abbreviations to match underlying time zone. Previously, we threw an error if a dynamic timezone abbreviation did not match any abbreviation recorded in the referenced IANA time zone entry. That seemed like a good consistency check at the time, but it turns out that a number of the abbreviations in the IANA database are things that Olson and crew made up out of whole cloth. Their current policy is to remove such names in favor of using simple numeric offsets. Perhaps unsurprisingly, a lot of these made-up abbreviations have varied in meaning over time, which meant that our commit b2cbced9e and later changes made them into dynamic abbreviations. So with newer IANA database versions that don't mention these abbreviations at all, we fail, as reported in bug #14307 from Neil Anderson. It's worse than just a few unused-in-the-wild abbreviations not working, because the pg_timezone_abbrevs view stops working altogether (since its underlying function tries to compute the whole view result in one call). We considered deleting these abbreviations from our abbreviations list, but the problem with that is that we can't stay ahead of possible future IANA changes. Instead, let's leave the abbreviations list alone, and treat any "orphaned" dynamic abbreviation as just meaning the referenced time zone. It will behave a bit differently than it used to, in that you can't any longer override the zone's standard vs. daylight rule by using the "wrong" abbreviation of a pair, but that's better than failing entirely. (Also, this solution can be interpreted as adding a small new feature, which is that any abbreviation a user wants can be defined as referencing a time zone name.) Back-patch to all supported branches, since this problem affects all of them when using tzdata 2016f or newer. Report: <20160902031551.15674.67337@wrigleys.postgresql.org> Discussion: <6189.1472820913@sss.pgh.pa.us>
2016-09-02 23:29:31 +02:00
/*
* Try to match the abbreviation to something in the zone definition.
*/
if (DetermineTimeZoneAbbrevOffsetInternal(t, abbr, tzp,
&abbr_offset, &abbr_isdst))
{
/* Success, so use the abbrev-specific answers. */
tm->tm_isdst = abbr_isdst;
return abbr_offset;
}
/*
* No match, so use the answers we already got from
* DetermineTimeZoneOffsetInternal.
*/
return zone_offset;
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
}
/* DetermineTimeZoneAbbrevOffsetTS()
*
* As above but the probe time is specified as a TimestampTz (hence, UTC time),
* and DST status is returned into *isdst rather than into tm->tm_isdst.
*/
int
DetermineTimeZoneAbbrevOffsetTS(TimestampTz ts, const char *abbr,
pg_tz *tzp, int *isdst)
{
pg_time_t t = timestamptz_to_time_t(ts);
Don't require dynamic timezone abbreviations to match underlying time zone. Previously, we threw an error if a dynamic timezone abbreviation did not match any abbreviation recorded in the referenced IANA time zone entry. That seemed like a good consistency check at the time, but it turns out that a number of the abbreviations in the IANA database are things that Olson and crew made up out of whole cloth. Their current policy is to remove such names in favor of using simple numeric offsets. Perhaps unsurprisingly, a lot of these made-up abbreviations have varied in meaning over time, which meant that our commit b2cbced9e and later changes made them into dynamic abbreviations. So with newer IANA database versions that don't mention these abbreviations at all, we fail, as reported in bug #14307 from Neil Anderson. It's worse than just a few unused-in-the-wild abbreviations not working, because the pg_timezone_abbrevs view stops working altogether (since its underlying function tries to compute the whole view result in one call). We considered deleting these abbreviations from our abbreviations list, but the problem with that is that we can't stay ahead of possible future IANA changes. Instead, let's leave the abbreviations list alone, and treat any "orphaned" dynamic abbreviation as just meaning the referenced time zone. It will behave a bit differently than it used to, in that you can't any longer override the zone's standard vs. daylight rule by using the "wrong" abbreviation of a pair, but that's better than failing entirely. (Also, this solution can be interpreted as adding a small new feature, which is that any abbreviation a user wants can be defined as referencing a time zone name.) Back-patch to all supported branches, since this problem affects all of them when using tzdata 2016f or newer. Report: <20160902031551.15674.67337@wrigleys.postgresql.org> Discussion: <6189.1472820913@sss.pgh.pa.us>
2016-09-02 23:29:31 +02:00
int zone_offset;
int abbr_offset;
int tz;
struct pg_tm tm;
fsec_t fsec;
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
Don't require dynamic timezone abbreviations to match underlying time zone. Previously, we threw an error if a dynamic timezone abbreviation did not match any abbreviation recorded in the referenced IANA time zone entry. That seemed like a good consistency check at the time, but it turns out that a number of the abbreviations in the IANA database are things that Olson and crew made up out of whole cloth. Their current policy is to remove such names in favor of using simple numeric offsets. Perhaps unsurprisingly, a lot of these made-up abbreviations have varied in meaning over time, which meant that our commit b2cbced9e and later changes made them into dynamic abbreviations. So with newer IANA database versions that don't mention these abbreviations at all, we fail, as reported in bug #14307 from Neil Anderson. It's worse than just a few unused-in-the-wild abbreviations not working, because the pg_timezone_abbrevs view stops working altogether (since its underlying function tries to compute the whole view result in one call). We considered deleting these abbreviations from our abbreviations list, but the problem with that is that we can't stay ahead of possible future IANA changes. Instead, let's leave the abbreviations list alone, and treat any "orphaned" dynamic abbreviation as just meaning the referenced time zone. It will behave a bit differently than it used to, in that you can't any longer override the zone's standard vs. daylight rule by using the "wrong" abbreviation of a pair, but that's better than failing entirely. (Also, this solution can be interpreted as adding a small new feature, which is that any abbreviation a user wants can be defined as referencing a time zone name.) Back-patch to all supported branches, since this problem affects all of them when using tzdata 2016f or newer. Report: <20160902031551.15674.67337@wrigleys.postgresql.org> Discussion: <6189.1472820913@sss.pgh.pa.us>
2016-09-02 23:29:31 +02:00
/*
* If the abbrev matches anything in the zone data, this is pretty easy.
*/
if (DetermineTimeZoneAbbrevOffsetInternal(t, abbr, tzp,
&abbr_offset, isdst))
return abbr_offset;
/*
* Else, break down the timestamp so we can use DetermineTimeZoneOffset.
*/
if (timestamp2tm(ts, &tz, &tm, &fsec, NULL, tzp) != 0)
ereport(ERROR,
(errcode(ERRCODE_DATETIME_VALUE_OUT_OF_RANGE),
errmsg("timestamp out of range")));
zone_offset = DetermineTimeZoneOffset(&tm, tzp);
*isdst = tm.tm_isdst;
return zone_offset;
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
}
/* DetermineTimeZoneAbbrevOffsetInternal()
*
* Workhorse for above two functions: work from a pg_time_t probe instant.
Don't require dynamic timezone abbreviations to match underlying time zone. Previously, we threw an error if a dynamic timezone abbreviation did not match any abbreviation recorded in the referenced IANA time zone entry. That seemed like a good consistency check at the time, but it turns out that a number of the abbreviations in the IANA database are things that Olson and crew made up out of whole cloth. Their current policy is to remove such names in favor of using simple numeric offsets. Perhaps unsurprisingly, a lot of these made-up abbreviations have varied in meaning over time, which meant that our commit b2cbced9e and later changes made them into dynamic abbreviations. So with newer IANA database versions that don't mention these abbreviations at all, we fail, as reported in bug #14307 from Neil Anderson. It's worse than just a few unused-in-the-wild abbreviations not working, because the pg_timezone_abbrevs view stops working altogether (since its underlying function tries to compute the whole view result in one call). We considered deleting these abbreviations from our abbreviations list, but the problem with that is that we can't stay ahead of possible future IANA changes. Instead, let's leave the abbreviations list alone, and treat any "orphaned" dynamic abbreviation as just meaning the referenced time zone. It will behave a bit differently than it used to, in that you can't any longer override the zone's standard vs. daylight rule by using the "wrong" abbreviation of a pair, but that's better than failing entirely. (Also, this solution can be interpreted as adding a small new feature, which is that any abbreviation a user wants can be defined as referencing a time zone name.) Back-patch to all supported branches, since this problem affects all of them when using tzdata 2016f or newer. Report: <20160902031551.15674.67337@wrigleys.postgresql.org> Discussion: <6189.1472820913@sss.pgh.pa.us>
2016-09-02 23:29:31 +02:00
* On success, return GMT offset and DST status into *offset and *isdst.
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
*/
Don't require dynamic timezone abbreviations to match underlying time zone. Previously, we threw an error if a dynamic timezone abbreviation did not match any abbreviation recorded in the referenced IANA time zone entry. That seemed like a good consistency check at the time, but it turns out that a number of the abbreviations in the IANA database are things that Olson and crew made up out of whole cloth. Their current policy is to remove such names in favor of using simple numeric offsets. Perhaps unsurprisingly, a lot of these made-up abbreviations have varied in meaning over time, which meant that our commit b2cbced9e and later changes made them into dynamic abbreviations. So with newer IANA database versions that don't mention these abbreviations at all, we fail, as reported in bug #14307 from Neil Anderson. It's worse than just a few unused-in-the-wild abbreviations not working, because the pg_timezone_abbrevs view stops working altogether (since its underlying function tries to compute the whole view result in one call). We considered deleting these abbreviations from our abbreviations list, but the problem with that is that we can't stay ahead of possible future IANA changes. Instead, let's leave the abbreviations list alone, and treat any "orphaned" dynamic abbreviation as just meaning the referenced time zone. It will behave a bit differently than it used to, in that you can't any longer override the zone's standard vs. daylight rule by using the "wrong" abbreviation of a pair, but that's better than failing entirely. (Also, this solution can be interpreted as adding a small new feature, which is that any abbreviation a user wants can be defined as referencing a time zone name.) Back-patch to all supported branches, since this problem affects all of them when using tzdata 2016f or newer. Report: <20160902031551.15674.67337@wrigleys.postgresql.org> Discussion: <6189.1472820913@sss.pgh.pa.us>
2016-09-02 23:29:31 +02:00
static bool
DetermineTimeZoneAbbrevOffsetInternal(pg_time_t t, const char *abbr, pg_tz *tzp,
int *offset, int *isdst)
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
{
char upabbr[TZ_STRLEN_MAX + 1];
unsigned char *p;
long int gmtoff;
/* We need to force the abbrev to upper case */
strlcpy(upabbr, abbr, sizeof(upabbr));
for (p = (unsigned char *) upabbr; *p; p++)
*p = pg_toupper(*p);
/* Look up the abbrev's meaning at this time in this zone */
Don't require dynamic timezone abbreviations to match underlying time zone. Previously, we threw an error if a dynamic timezone abbreviation did not match any abbreviation recorded in the referenced IANA time zone entry. That seemed like a good consistency check at the time, but it turns out that a number of the abbreviations in the IANA database are things that Olson and crew made up out of whole cloth. Their current policy is to remove such names in favor of using simple numeric offsets. Perhaps unsurprisingly, a lot of these made-up abbreviations have varied in meaning over time, which meant that our commit b2cbced9e and later changes made them into dynamic abbreviations. So with newer IANA database versions that don't mention these abbreviations at all, we fail, as reported in bug #14307 from Neil Anderson. It's worse than just a few unused-in-the-wild abbreviations not working, because the pg_timezone_abbrevs view stops working altogether (since its underlying function tries to compute the whole view result in one call). We considered deleting these abbreviations from our abbreviations list, but the problem with that is that we can't stay ahead of possible future IANA changes. Instead, let's leave the abbreviations list alone, and treat any "orphaned" dynamic abbreviation as just meaning the referenced time zone. It will behave a bit differently than it used to, in that you can't any longer override the zone's standard vs. daylight rule by using the "wrong" abbreviation of a pair, but that's better than failing entirely. (Also, this solution can be interpreted as adding a small new feature, which is that any abbreviation a user wants can be defined as referencing a time zone name.) Back-patch to all supported branches, since this problem affects all of them when using tzdata 2016f or newer. Report: <20160902031551.15674.67337@wrigleys.postgresql.org> Discussion: <6189.1472820913@sss.pgh.pa.us>
2016-09-02 23:29:31 +02:00
if (pg_interpret_timezone_abbrev(upabbr,
&t,
&gmtoff,
isdst,
tzp))
{
/* Change sign to agree with DetermineTimeZoneOffset() */
*offset = (int) -gmtoff;
return true;
}
return false;
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
}
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
/* DecodeTimeOnly()
* Interpret parsed string as time fields only.
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
* Returns 0 if successful, DTERR code if bogus input detected.
*
Implement column aliases on views "CREATE VIEW name (collist)". Implement TIME WITH TIME ZONE type (timetz internal type). Remap length() for character strings to CHAR_LENGTH() for SQL92 and to remove the ambiguity with geometric length() functions. Keep length() for character strings for backward compatibility. Shrink stored views by removing internal column name list from visible rte. Implement min(), max() for time and timetz data types. Implement conversion of TIME to INTERVAL. Implement abs(), mod(), fac() for the int8 data type. Rename some math functions to generic names: round(), sqrt(), cbrt(), pow(), etc. Rename NUMERIC power() function to pow(). Fix int2 factorial to calculate result in int4. Enhance the Oracle compatibility function translate() to work with string arguments (from Edwin Ramirez). Modify pg_proc system table to remove OID holes.
2000-03-15 00:06:59 +01:00
* Note that support for time zone is here for
Clean up references to SQL92 In most cases, these were just references to the SQL standard in general. In a few cases, a contrast was made between SQL92 and later standards -- those have been kept unchanged.
2013-04-20 17:04:41 +02:00
* SQL TIME WITH TIME ZONE, but it reveals
* bogosity with SQL date/time standards, since
Implement column aliases on views "CREATE VIEW name (collist)". Implement TIME WITH TIME ZONE type (timetz internal type). Remap length() for character strings to CHAR_LENGTH() for SQL92 and to remove the ambiguity with geometric length() functions. Keep length() for character strings for backward compatibility. Shrink stored views by removing internal column name list from visible rte. Implement min(), max() for time and timetz data types. Implement conversion of TIME to INTERVAL. Implement abs(), mod(), fac() for the int8 data type. Rename some math functions to generic names: round(), sqrt(), cbrt(), pow(), etc. Rename NUMERIC power() function to pow(). Fix int2 factorial to calculate result in int4. Enhance the Oracle compatibility function translate() to work with string arguments (from Edwin Ramirez). Modify pg_proc system table to remove OID holes.
2000-03-15 00:06:59 +01:00
* we must infer a time zone from current time.
* - thomas 2000-03-10
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
* Allow specifying date to get a better time zone,
* if time zones are allowed. - thomas 2001-12-26
From: Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> Subject: [HACKERS] More date time functions Here are some additional patches mostly related to the date and time data types. It includes some type conversion routines to move between the different date types and some other date manipulation routines such as date_part(units,datetime). I noticed Edmund Mergl et al's neat trick for getting function overloading for builtin functions, so started to use that for the date and time stuff. Later, if someone figures out how to get function overloading directly for internal C code, then we can move to that technique. These patches include documentation updates (don't faint!) for the built-in man page. Doesn't yet include mention of timestamp, since I don't know much about it and since it may change a bit to become a _real_ ANSI timestamp which would include parser support for the declaration syntax (what do you think, Dan?). The patches were developed on the 970330 release, but have been rebuilt off of the 970402 release. The first patch below is to get libpq to compile, on my Linux box, but is not related to the rest of the patches and you can choose not to apply that one at this time. Thanks in advance, scrappy!
1997-04-02 20:36:24 +02:00
*/
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
int
Implement column aliases on views "CREATE VIEW name (collist)". Implement TIME WITH TIME ZONE type (timetz internal type). Remap length() for character strings to CHAR_LENGTH() for SQL92 and to remove the ambiguity with geometric length() functions. Keep length() for character strings for backward compatibility. Shrink stored views by removing internal column name list from visible rte. Implement min(), max() for time and timetz data types. Implement conversion of TIME to INTERVAL. Implement abs(), mod(), fac() for the int8 data type. Rename some math functions to generic names: round(), sqrt(), cbrt(), pow(), etc. Rename NUMERIC power() function to pow(). Fix int2 factorial to calculate result in int4. Enhance the Oracle compatibility function translate() to work with string arguments (from Edwin Ramirez). Modify pg_proc system table to remove OID holes.
2000-03-15 00:06:59 +01:00
DecodeTimeOnly(char **field, int *ftype, int nf,
Initial pgindent run with pg_bsd_indent version 2.0. The new indent version includes numerous fixes thanks to Piotr Stefaniak. The main changes visible in this commit are: * Nicer formatting of function-pointer declarations. * No longer unexpectedly removes spaces in expressions using casts, sizeof, or offsetof. * No longer wants to add a space in "struct structname *varname", as well as some similar cases for const- or volatile-qualified pointers. * Declarations using PG_USED_FOR_ASSERTS_ONLY are formatted more nicely. * Fixes bug where comments following declarations were sometimes placed with no space separating them from the code. * Fixes some odd decisions for comments following case labels. * Fixes some cases where comments following code were indented to less than the expected column 33. On the less good side, it now tends to put more whitespace around typedef names that are not listed in typedefs.list. This might encourage us to put more effort into typedef name collection; it's not really a bug in indent itself. There are more changes coming after this round, having to do with comment indentation and alignment of lines appearing within parentheses. I wanted to limit the size of the diffs to something that could be reviewed without one's eyes completely glazing over, so it seemed better to split up the changes as much as practical. Discussion: https://postgr.es/m/E1dAmxK-0006EE-1r@gemulon.postgresql.org Discussion: https://postgr.es/m/30527.1495162840@sss.pgh.pa.us
2017-06-21 20:39:04 +02:00
int *dtype, struct pg_tm *tm, fsec_t *fsec, int *tzp)
From: Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> Subject: [HACKERS] More date time functions Here are some additional patches mostly related to the date and time data types. It includes some type conversion routines to move between the different date types and some other date manipulation routines such as date_part(units,datetime). I noticed Edmund Mergl et al's neat trick for getting function overloading for builtin functions, so started to use that for the date and time stuff. Later, if someone figures out how to get function overloading directly for internal C code, then we can move to that technique. These patches include documentation updates (don't faint!) for the built-in man page. Doesn't yet include mention of timestamp, since I don't know much about it and since it may change a bit to become a _real_ ANSI timestamp which would include parser support for the declaration syntax (what do you think, Dan?). The patches were developed on the 970330 release, but have been rebuilt off of the 970402 release. The first patch below is to get libpq to compile, on my Linux box, but is not related to the rest of the patches and you can choose not to apply that one at this time. Thanks in advance, scrappy!
1997-04-02 20:36:24 +02:00
{
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
int fmask = 0,
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
tmask,
type;
pgindent run.
2002-09-04 22:31:48 +02:00
int ptype = 0; /* "prefix type" for ISO h04mm05s06 format */
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
int i;
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
int val;
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
int dterr;
Change TRUE/FALSE to true/false The lower case spellings are C and C++ standard and are used in most parts of the PostgreSQL sources. The upper case spellings are only used in some files/modules. So standardize on the standard spellings. The APIs for ICU, Perl, and Windows define their own TRUE and FALSE, so those are left as is when using those APIs. In code comments, we use the lower-case spelling for the C concepts and keep the upper-case spelling for the SQL concepts. Reviewed-by: Michael Paquier <michael.paquier@gmail.com>
2017-08-16 06:22:32 +02:00
bool isjulian = false;
bool is2digits = false;
bool bc = false;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
int mer = HR24;
Fix up some problems in handling of zic-style time zone names in datetime input routines. Remove the former "DecodePosixTimezone" function in favor of letting the zic code handle POSIX-style zone specs (see tzparse()). In particular this means that "PST+3" now means the same as "-03", whereas it used to mean "-11" --- the zone abbreviation is effectively just a noise word in this syntax. Make sure that all named and POSIX-style zone names will be parsed as a single token. Fix long-standing bogosities in printing and input of fractional-hour timezone offsets (since the tzparse() code will accept these, we'd better make 'em work). Also correct an error in the original coding of the zic-zone-name patch: in "timestamp without time zone" input, zone names are supposed to be allowed but ignored, but the coding was such that the zone changed the interpretation anyway.
2006-10-17 23:03:21 +02:00
pg_tz *namedTz = NULL;
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
pg_tz *abbrevTz = NULL;
char *abbrev = NULL;
pg_tz *valtz;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
*dtype = DTK_TIME;
tm->tm_hour = 0;
tm->tm_min = 0;
tm->tm_sec = 0;
*fsec = 0;
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
/* don't know daylight savings time status apriori */
tm->tm_isdst = -1;
Implement column aliases on views "CREATE VIEW name (collist)". Implement TIME WITH TIME ZONE type (timetz internal type). Remap length() for character strings to CHAR_LENGTH() for SQL92 and to remove the ambiguity with geometric length() functions. Keep length() for character strings for backward compatibility. Shrink stored views by removing internal column name list from visible rte. Implement min(), max() for time and timetz data types. Implement conversion of TIME to INTERVAL. Implement abs(), mod(), fac() for the int8 data type. Rename some math functions to generic names: round(), sqrt(), cbrt(), pow(), etc. Rename NUMERIC power() function to pow(). Fix int2 factorial to calculate result in int4. Enhance the Oracle compatibility function translate() to work with string arguments (from Edwin Ramirez). Modify pg_proc system table to remove OID holes.
2000-03-15 00:06:59 +01:00
if (tzp != NULL)
*tzp = 0;
From: Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> Subject: [HACKERS] More date time functions Here are some additional patches mostly related to the date and time data types. It includes some type conversion routines to move between the different date types and some other date manipulation routines such as date_part(units,datetime). I noticed Edmund Mergl et al's neat trick for getting function overloading for builtin functions, so started to use that for the date and time stuff. Later, if someone figures out how to get function overloading directly for internal C code, then we can move to that technique. These patches include documentation updates (don't faint!) for the built-in man page. Doesn't yet include mention of timestamp, since I don't know much about it and since it may change a bit to become a _real_ ANSI timestamp which would include parser support for the declaration syntax (what do you think, Dan?). The patches were developed on the 970330 release, but have been rebuilt off of the 970402 release. The first patch below is to get libpq to compile, on my Linux box, but is not related to the rest of the patches and you can choose not to apply that one at this time. Thanks in advance, scrappy!
1997-04-02 20:36:24 +02:00
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
for (i = 0; i < nf; i++)
{
switch (ftype[i])
{
Support full POSIX-style time zone: EST+3, PST-3, etc. We probably support a superset of the spec, but I don't have the spec to confirm this. Update regression tests to include tests for this format.
2000-03-16 15:38:42 +01:00
case DTK_DATE:
pgindent run.
2002-09-04 22:31:48 +02:00
/*
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* Time zone not allowed? Then should not accept dates or time
* zones no matter what else!
Support full POSIX-style time zone: EST+3, PST-3, etc. We probably support a superset of the spec, but I don't have the spec to confirm this. Update regression tests to include tests for this format.
2000-03-16 15:38:42 +01:00
*/
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
if (tzp == NULL)
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
return DTERR_BAD_FORMAT;
Support full POSIX-style time zone: EST+3, PST-3, etc. We probably support a superset of the spec, but I don't have the spec to confirm this. Update regression tests to include tests for this format.
2000-03-16 15:38:42 +01:00
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
/* Under limited circumstances, we will accept a date... */
Remove unnecessary parentheses in datetime/timestamp code.
2005-05-23 19:13:14 +02:00
if (i == 0 && nf >= 2 &&
(ftype[nf - 1] == DTK_DATE || ftype[1] == DTK_TIME))
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
{
Fix datetime input to behave correctly for Feb 29 in years BC. Formerly, DecodeDate attempted to verify the day-of-the-month exactly, but it was under the misapprehension that it would know whether we were looking at a BC year or not. In reality this check can't be made until the calling function (eg DecodeDateTime) has processed all the fields. So, split the BC adjustment and validity checks out into a new function ValidateDate that is called only after processing all the fields. In passing, this patch makes DecodeTimeOnly work for BC inputs, which it never did before. (The historical veracity of all this is nonexistent, of course, but if we're going to say we support proleptic Gregorian calendar then we should do it correctly. In any case the unpatched code is broken because it could emit dates that it would then reject on re-inputting.) Per report from Bernd Helmle. Back-patch as far as 8.0; in 7.x we were not using our own calendar support and so this seems a bit too risky to put into 7.4.
2008-02-26 00:21:01 +01:00
dterr = DecodeDate(field[i], fmask,
&tmask, &is2digits, tm);
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
if (dterr)
return dterr;
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
}
/* otherwise, this is a time and/or time zone */
else
{
Make sure that all <ctype.h> routines are called with unsigned char values; it's not portable to call them with signed chars. I recall doing this for the last release, but a few more uncasted calls have snuck in.
2001-12-31 00:09:42 +01:00
if (isdigit((unsigned char) *field[i]))
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
{
pgindent run.
2002-09-04 22:31:48 +02:00
char *cp;
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
pgindent run.
2002-09-04 22:31:48 +02:00
/*
* Starts with a digit but we already have a time
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* field? Then we are in trouble with time already...
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
*/
if ((fmask & DTK_TIME_M) == DTK_TIME_M)
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
return DTERR_BAD_FORMAT;
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
pgindent run.
2002-09-04 22:31:48 +02:00
/*
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* Should not get here and fail. Sanity check only...
pgindent run.
2002-09-04 22:31:48 +02:00
*/
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
if ((cp = strchr(field[i], '-')) == NULL)
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
return DTERR_BAD_FORMAT;
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
/* Get the time zone from the end of the string */
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
dterr = DecodeTimezone(cp, tzp);
if (dterr)
return dterr;
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
*cp = '\0';
pgindent run.
2002-09-04 22:31:48 +02:00
/*
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* Then read the rest of the field as a concatenated
* time
pgindent run.
2002-09-04 22:31:48 +02:00
*/
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
dterr = DecodeNumberField(strlen(field[i]), field[i],
(fmask | DTK_DATE_M),
&tmask, tm,
fsec, &is2digits);
if (dterr < 0)
return dterr;
ftype[i] = dterr;
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
tmask |= DTK_M(TZ);
}
else
{
Fix up some problems in handling of zic-style time zone names in datetime input routines. Remove the former "DecodePosixTimezone" function in favor of letting the zic code handle POSIX-style zone specs (see tzparse()). In particular this means that "PST+3" now means the same as "-03", whereas it used to mean "-11" --- the zone abbreviation is effectively just a noise word in this syntax. Make sure that all named and POSIX-style zone names will be parsed as a single token. Fix long-standing bogosities in printing and input of fractional-hour timezone offsets (since the tzparse() code will accept these, we'd better make 'em work). Also correct an error in the original coding of the zic-zone-name patch: in "timestamp without time zone" input, zone names are supposed to be allowed but ignored, but the coding was such that the zone changed the interpretation anyway.
2006-10-17 23:03:21 +02:00
namedTz = pg_tzset(field[i]);
if (!namedTz)
{
/*
* We should return an error code instead of
pgindent run for 8.3.
2007-11-15 22:14:46 +01:00
* ereport'ing directly, but then there is no way
* to report the bad time zone name.
Fix up some problems in handling of zic-style time zone names in datetime input routines. Remove the former "DecodePosixTimezone" function in favor of letting the zic code handle POSIX-style zone specs (see tzparse()). In particular this means that "PST+3" now means the same as "-03", whereas it used to mean "-11" --- the zone abbreviation is effectively just a noise word in this syntax. Make sure that all named and POSIX-style zone names will be parsed as a single token. Fix long-standing bogosities in printing and input of fractional-hour timezone offsets (since the tzparse() code will accept these, we'd better make 'em work). Also correct an error in the original coding of the zic-zone-name patch: in "timestamp without time zone" input, zone names are supposed to be allowed but ignored, but the coding was such that the zone changed the interpretation anyway.
2006-10-17 23:03:21 +02:00
*/
ereport(ERROR,
(errcode(ERRCODE_INVALID_PARAMETER_VALUE),
errmsg("time zone \"%s\" not recognized",
field[i])));
}
/* we'll apply the zone setting below */
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
ftype[i] = DTK_TZ;
tmask = DTK_M(TZ);
}
}
Support full POSIX-style time zone: EST+3, PST-3, etc. We probably support a superset of the spec, but I don't have the spec to confirm this. Update regression tests to include tests for this format.
2000-03-16 15:38:42 +01:00
break;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
case DTK_TIME:
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
dterr = DecodeTime(field[i], (fmask | DTK_DATE_M),
Make our parsing of INTERVAL literals spec-compliant (or at least a heck of a lot closer than it was before). To do this, tweak coerce_type() to pass through the typmod information when invoking interval_in() on an UNKNOWN constant; then fix DecodeInterval to pay attention to the typmod when deciding how to interpret a units-less integer value. I changed one or two other details as well. I believe the code now reacts as expected by spec for all the literal syntaxes that are specifically enumerated in the spec. There are corner cases involving strings that don't exactly match the set of fields called out by the typmod, for which we might want to tweak the behavior some more; but I think this is an area of user friendliness rather than spec compliance. There remain some non-compliant details about the SQL syntax (as opposed to what's inside the literal string); but at least we'll throw error rather than silently doing the wrong thing in those cases.
2008-09-10 20:29:41 +02:00
INTERVAL_FULL_RANGE,
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
&tmask, tm, fsec);
if (dterr)
return dterr;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
break;
Implement column aliases on views "CREATE VIEW name (collist)". Implement TIME WITH TIME ZONE type (timetz internal type). Remap length() for character strings to CHAR_LENGTH() for SQL92 and to remove the ambiguity with geometric length() functions. Keep length() for character strings for backward compatibility. Shrink stored views by removing internal column name list from visible rte. Implement min(), max() for time and timetz data types. Implement conversion of TIME to INTERVAL. Implement abs(), mod(), fac() for the int8 data type. Rename some math functions to generic names: round(), sqrt(), cbrt(), pow(), etc. Rename NUMERIC power() function to pow(). Fix int2 factorial to calculate result in int4. Enhance the Oracle compatibility function translate() to work with string arguments (from Edwin Ramirez). Modify pg_proc system table to remove OID holes.
2000-03-15 00:06:59 +01:00
case DTK_TZ:
Support full POSIX-style time zone: EST+3, PST-3, etc. We probably support a superset of the spec, but I don't have the spec to confirm this. Update regression tests to include tests for this format.
2000-03-16 15:38:42 +01:00
{
Ye-old pgindent run. Same 4-space tabs.
2000-04-12 19:17:23 +02:00
int tz;
Support full POSIX-style time zone: EST+3, PST-3, etc. We probably support a superset of the spec, but I don't have the spec to confirm this. Update regression tests to include tests for this format.
2000-03-16 15:38:42 +01:00
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
if (tzp == NULL)
return DTERR_BAD_FORMAT;
Fix up some problems in handling of zic-style time zone names in datetime input routines. Remove the former "DecodePosixTimezone" function in favor of letting the zic code handle POSIX-style zone specs (see tzparse()). In particular this means that "PST+3" now means the same as "-03", whereas it used to mean "-11" --- the zone abbreviation is effectively just a noise word in this syntax. Make sure that all named and POSIX-style zone names will be parsed as a single token. Fix long-standing bogosities in printing and input of fractional-hour timezone offsets (since the tzparse() code will accept these, we'd better make 'em work). Also correct an error in the original coding of the zic-zone-name patch: in "timestamp without time zone" input, zone names are supposed to be allowed but ignored, but the coding was such that the zone changed the interpretation anyway.
2006-10-17 23:03:21 +02:00
dterr = DecodeTimezone(field[i], &tz);
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
if (dterr)
return dterr;
Fix up some problems in handling of zic-style time zone names in datetime input routines. Remove the former "DecodePosixTimezone" function in favor of letting the zic code handle POSIX-style zone specs (see tzparse()). In particular this means that "PST+3" now means the same as "-03", whereas it used to mean "-11" --- the zone abbreviation is effectively just a noise word in this syntax. Make sure that all named and POSIX-style zone names will be parsed as a single token. Fix long-standing bogosities in printing and input of fractional-hour timezone offsets (since the tzparse() code will accept these, we'd better make 'em work). Also correct an error in the original coding of the zic-zone-name patch: in "timestamp without time zone" input, zone names are supposed to be allowed but ignored, but the coding was such that the zone changed the interpretation anyway.
2006-10-17 23:03:21 +02:00
*tzp = tz;
tmask = DTK_M(TZ);
Support full POSIX-style time zone: EST+3, PST-3, etc. We probably support a superset of the spec, but I don't have the spec to confirm this. Update regression tests to include tests for this format.
2000-03-16 15:38:42 +01:00
}
Implement column aliases on views "CREATE VIEW name (collist)". Implement TIME WITH TIME ZONE type (timetz internal type). Remap length() for character strings to CHAR_LENGTH() for SQL92 and to remove the ambiguity with geometric length() functions. Keep length() for character strings for backward compatibility. Shrink stored views by removing internal column name list from visible rte. Implement min(), max() for time and timetz data types. Implement conversion of TIME to INTERVAL. Implement abs(), mod(), fac() for the int8 data type. Rename some math functions to generic names: round(), sqrt(), cbrt(), pow(), etc. Rename NUMERIC power() function to pow(). Fix int2 factorial to calculate result in int4. Enhance the Oracle compatibility function translate() to work with string arguments (from Edwin Ramirez). Modify pg_proc system table to remove OID holes.
2000-03-15 00:06:59 +01:00
break;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
case DTK_NUMBER:
pgindent run.
2002-09-04 22:31:48 +02:00
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
/*
* Was this an "ISO time" with embedded field labels? An
Fix inconsistencies and typos in the tree This is numbered take 7, and addresses a set of issues with code comments, variable names and unreferenced variables. Author: Alexander Lakhin Discussion: https://postgr.es/m/dff75442-2468-f74f-568c-6006e141062f@gmail.com
2019-07-22 03:01:50 +02:00
* example is "h04mm05s06" - thomas 2001-02-04
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
*/
if (ptype != 0)
{
char *cp;
int val;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
/* Only accept a date under limited circumstances */
switch (ptype)
{
case DTK_JULIAN:
case DTK_YEAR:
case DTK_MONTH:
case DTK_DAY:
if (tzp == NULL)
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
return DTERR_BAD_FORMAT;
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
default:
break;
}
Remove comment on errno=0 lines, but add mention to port/strtol.c function.
2005-12-02 03:49:11 +01:00
errno = 0;
Rename strtoi() to strtoint(). NetBSD has seen fit to invent a libc function named strtoi(), which conflicts with the long-established static functions of the same name in datetime.c and ecpg's interval.c. While muttering darkly about intrusions on application namespace, we'll rename our functions to avoid the conflict. Back-patch to all supported branches, since this would affect attempts to build any of them on recent NetBSD. Thomas Munro
2016-04-23 22:53:15 +02:00
val = strtoint(field[i], &cp, 10);
Check for overflow in strtol() while parsing datetime inputs. Michael Fuhr.
2005-12-01 18:56:34 +01:00
if (errno == ERANGE)
return DTERR_FIELD_OVERFLOW;
pgindent run.
2002-09-04 22:31:48 +02:00
/*
* only a few kinds are allowed to have an embedded
* decimal
*/
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
if (*cp == '.')
switch (ptype)
{
case DTK_JULIAN:
case DTK_TIME:
case DTK_SECOND:
break;
default:
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
return DTERR_BAD_FORMAT;
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
break;
}
else if (*cp != '\0')
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
return DTERR_BAD_FORMAT;
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
switch (ptype)
{
case DTK_YEAR:
tm->tm_year = val;
tmask = DTK_M(YEAR);
break;
case DTK_MONTH:
pgindent run.
2002-09-04 22:31:48 +02:00
/*
* already have a month and hour? then assume
* minutes
*/
Remove unnecessary parentheses in datetime/timestamp code.
2005-05-23 19:13:14 +02:00
if ((fmask & DTK_M(MONTH)) != 0 &&
(fmask & DTK_M(HOUR)) != 0)
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
{
tm->tm_min = val;
tmask = DTK_M(MINUTE);
}
else
{
tm->tm_mon = val;
tmask = DTK_M(MONTH);
}
break;
case DTK_DAY:
tm->tm_mday = val;
tmask = DTK_M(DAY);
break;
case DTK_HOUR:
tm->tm_hour = val;
tmask = DTK_M(HOUR);
break;
case DTK_MINUTE:
tm->tm_min = val;
tmask = DTK_M(MINUTE);
break;
case DTK_SECOND:
tm->tm_sec = val;
tmask = DTK_M(SECOND);
if (*cp == '.')
{
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
dterr = ParseFractionalSecond(cp, fsec);
if (dterr)
return dterr;
Fix a bug in input processing for the "interval" type. Previously, "microsecond" and "millisecond" units were not considered valid input by themselves, which caused inputs like "1 millisecond" to be rejected erroneously. Update the docs, add regression tests, and backport to 8.2 and 8.1
2007-05-29 06:58:43 +02:00
tmask = DTK_ALL_SECS_M;
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
}
break;
case DTK_TZ:
tmask = DTK_M(TZ);
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
dterr = DecodeTimezone(field[i], tzp);
if (dterr)
return dterr;
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
break;
case DTK_JULIAN:
Re-allow input of Julian dates prior to 0001-01-01 AD. This was unintentionally broken in 8.4 while tightening up checking of ordinary non-Julian date inputs to forbid references to "year zero". Per bug #5672 from Benjamin Gigot.
2010-09-23 05:48:07 +02:00
/* previous field was a label for "julian date" */
if (val < 0)
return DTERR_FIELD_OVERFLOW;
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
tmask = DTK_DATE_M;
j2date(val, &tm->tm_year, &tm->tm_mon, &tm->tm_mday);
Change TRUE/FALSE to true/false The lower case spellings are C and C++ standard and are used in most parts of the PostgreSQL sources. The upper case spellings are only used in some files/modules. So standardize on the standard spellings. The APIs for ICU, Perl, and Windows define their own TRUE and FALSE, so those are left as is when using those APIs. In code comments, we use the lower-case spelling for the C concepts and keep the upper-case spelling for the SQL concepts. Reviewed-by: Michael Paquier <michael.paquier@gmail.com>
2017-08-16 06:22:32 +02:00
isjulian = true;
Re-allow input of Julian dates prior to 0001-01-01 AD. This was unintentionally broken in 8.4 while tightening up checking of ordinary non-Julian date inputs to forbid references to "year zero". Per bug #5672 from Benjamin Gigot.
2010-09-23 05:48:07 +02:00
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
if (*cp == '.')
{
pgindent run.
2002-09-04 22:31:48 +02:00
double time;
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
errno = 0;
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
time = strtod(cp, &cp);
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
if (*cp != '\0' || errno != 0)
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
return DTERR_BAD_FORMAT;
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
time *= USECS_PER_DAY;
dt2time(time,
&tm->tm_hour, &tm->tm_min,
&tm->tm_sec, fsec);
tmask |= DTK_TIME_M;
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
}
break;
case DTK_TIME:
/* previous field was "t" for ISO time */
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
dterr = DecodeNumberField(strlen(field[i]), field[i],
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
(fmask | DTK_DATE_M),
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
&tmask, tm,
fsec, &is2digits);
if (dterr < 0)
return dterr;
ftype[i] = dterr;
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
if (tmask != DTK_TIME_M)
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
return DTERR_BAD_FORMAT;
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
break;
default:
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
return DTERR_BAD_FORMAT;
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
break;
}
ptype = 0;
*dtype = DTK_DATE;
}
else
{
pgindent run.
2002-09-04 22:31:48 +02:00
char *cp;
int flen;
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
flen = strlen(field[i]);
cp = strchr(field[i], '.');
/* Embedded decimal? */
if (cp != NULL)
{
pgindent run.
2002-09-04 22:31:48 +02:00
/*
* Under limited circumstances, we will accept a
* date...
*/
Remove unnecessary parentheses in datetime/timestamp code.
2005-05-23 19:13:14 +02:00
if (i == 0 && nf >= 2 && ftype[nf - 1] == DTK_DATE)
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
{
Fix datetime input to behave correctly for Feb 29 in years BC. Formerly, DecodeDate attempted to verify the day-of-the-month exactly, but it was under the misapprehension that it would know whether we were looking at a BC year or not. In reality this check can't be made until the calling function (eg DecodeDateTime) has processed all the fields. So, split the BC adjustment and validity checks out into a new function ValidateDate that is called only after processing all the fields. In passing, this patch makes DecodeTimeOnly work for BC inputs, which it never did before. (The historical veracity of all this is nonexistent, of course, but if we're going to say we support proleptic Gregorian calendar then we should do it correctly. In any case the unpatched code is broken because it could emit dates that it would then reject on re-inputting.) Per report from Bernd Helmle. Back-patch as far as 8.0; in 7.x we were not using our own calendar support and so this seems a bit too risky to put into 7.4.
2008-02-26 00:21:01 +01:00
dterr = DecodeDate(field[i], fmask,
&tmask, &is2digits, tm);
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
if (dterr)
return dterr;
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
}
/* embedded decimal and several digits before? */
Remove unnecessary parentheses in datetime/timestamp code.
2005-05-23 19:13:14 +02:00
else if (flen - strlen(cp) > 2)
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
{
pgindent run.
2002-09-04 22:31:48 +02:00
/*
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* Interpret as a concatenated date or time Set
* the type field to allow decoding other fields
* later. Example: 20011223 or 040506
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
*/
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
dterr = DecodeNumberField(flen, field[i],
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
(fmask | DTK_DATE_M),
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
&tmask, tm,
fsec, &is2digits);
if (dterr < 0)
return dterr;
ftype[i] = dterr;
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
}
else
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
return DTERR_BAD_FORMAT;
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
}
else if (flen > 4)
{
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
dterr = DecodeNumberField(flen, field[i],
(fmask | DTK_DATE_M),
&tmask, tm,
fsec, &is2digits);
if (dterr < 0)
return dterr;
ftype[i] = dterr;
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
}
/* otherwise it is a single date/time field... */
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
else
{
dterr = DecodeNumber(flen, field[i],
Change TRUE/FALSE to true/false The lower case spellings are C and C++ standard and are used in most parts of the PostgreSQL sources. The upper case spellings are only used in some files/modules. So standardize on the standard spellings. The APIs for ICU, Perl, and Windows define their own TRUE and FALSE, so those are left as is when using those APIs. In code comments, we use the lower-case spelling for the C concepts and keep the upper-case spelling for the SQL concepts. Reviewed-by: Michael Paquier <michael.paquier@gmail.com>
2017-08-16 06:22:32 +02:00
false,
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
(fmask | DTK_DATE_M),
&tmask, tm,
fsec, &is2digits);
if (dterr)
return dterr;
}
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
}
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
break;
case DTK_STRING:
case DTK_SPECIAL:
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
/* timezone abbrevs take precedence over built-in tokens */
type = DecodeTimezoneAbbrev(i, field[i], &val, &valtz);
if (type == UNKNOWN_FIELD)
type = DecodeSpecial(i, field[i], &val);
Katherine Ward wrote: > Changes to avoid collisions with WIN32 & MFC names... > 1. Renamed: > a. PROC => PGPROC > b. GetUserName() => GetUserNameFromId() > c. GetCurrentTime() => GetCurrentDateTime() > d. IGNORE => IGNORE_DTF in include/utils/datetime.h & utils/adt/datetim > > 2. Added _P to some lex/yacc tokens: > CONST, CHAR, DELETE, FLOAT, GROUP, IN, OUT Jan
2002-06-11 15:40:53 +02:00
if (type == IGNORE_DTF)
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
continue;
tmask = DTK_M(type);
switch (type)
{
case RESERV:
switch (val)
{
case DTK_NOW:
tmask = DTK_TIME_M;
*dtype = DTK_TIME;
Fix timestamptz_in so that parsing of 'now'::timestamptz gives right answer when SET TIMEZONE has been done since the start of the current transaction. Per bug report from Robert Haas. I plan some futher cleanup in HEAD, but this is a low-risk patch for the immediate issue in 7.3.
2003-02-20 06:24:55 +01:00
GetCurrentTimeUsec(tm, fsec, NULL);
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
break;
case DTK_ZULU:
tmask = (DTK_TIME_M | DTK_M(TZ));
*dtype = DTK_TIME;
tm->tm_hour = 0;
tm->tm_min = 0;
tm->tm_sec = 0;
tm->tm_isdst = 0;
break;
default:
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
return DTERR_BAD_FORMAT;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
}
break;
Support full POSIX-style time zone: EST+3, PST-3, etc. We probably support a superset of the spec, but I don't have the spec to confirm this. Update regression tests to include tests for this format.
2000-03-16 15:38:42 +01:00
case DTZMOD:
/*
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* daylight savings time modifier (solves "MET DST"
* syntax)
Support full POSIX-style time zone: EST+3, PST-3, etc. We probably support a superset of the spec, but I don't have the spec to confirm this. Update regression tests to include tests for this format.
2000-03-16 15:38:42 +01:00
*/
tmask |= DTK_M(DTZ);
tm->tm_isdst = 1;
if (tzp == NULL)
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
return DTERR_BAD_FORMAT;
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
*tzp -= val;
Support full POSIX-style time zone: EST+3, PST-3, etc. We probably support a superset of the spec, but I don't have the spec to confirm this. Update regression tests to include tests for this format.
2000-03-16 15:38:42 +01:00
break;
case DTZ:
/*
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* set mask for TZ here _or_ check for DTZ later when
* getting default timezone
Support full POSIX-style time zone: EST+3, PST-3, etc. We probably support a superset of the spec, but I don't have the spec to confirm this. Update regression tests to include tests for this format.
2000-03-16 15:38:42 +01:00
*/
tmask |= DTK_M(TZ);
tm->tm_isdst = 1;
if (tzp == NULL)
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
return DTERR_BAD_FORMAT;
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
*tzp = -val;
Support full POSIX-style time zone: EST+3, PST-3, etc. We probably support a superset of the spec, but I don't have the spec to confirm this. Update regression tests to include tests for this format.
2000-03-16 15:38:42 +01:00
ftype[i] = DTK_TZ;
break;
case TZ:
tm->tm_isdst = 0;
if (tzp == NULL)
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
return DTERR_BAD_FORMAT;
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
*tzp = -val;
Support full POSIX-style time zone: EST+3, PST-3, etc. We probably support a superset of the spec, but I don't have the spec to confirm this. Update regression tests to include tests for this format.
2000-03-16 15:38:42 +01:00
ftype[i] = DTK_TZ;
break;
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
case DYNTZ:
tmask |= DTK_M(TZ);
if (tzp == NULL)
return DTERR_BAD_FORMAT;
/* we'll determine the actual offset later */
abbrevTz = valtz;
abbrev = field[i];
ftype[i] = DTK_TZ;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
break;
case AMPM:
mer = val;
break;
Fix datetime input to behave correctly for Feb 29 in years BC. Formerly, DecodeDate attempted to verify the day-of-the-month exactly, but it was under the misapprehension that it would know whether we were looking at a BC year or not. In reality this check can't be made until the calling function (eg DecodeDateTime) has processed all the fields. So, split the BC adjustment and validity checks out into a new function ValidateDate that is called only after processing all the fields. In passing, this patch makes DecodeTimeOnly work for BC inputs, which it never did before. (The historical veracity of all this is nonexistent, of course, but if we're going to say we support proleptic Gregorian calendar then we should do it correctly. In any case the unpatched code is broken because it could emit dates that it would then reject on re-inputting.) Per report from Bernd Helmle. Back-patch as far as 8.0; in 7.x we were not using our own calendar support and so this seems a bit too risky to put into 7.4.
2008-02-26 00:21:01 +01:00
case ADBC:
bc = (val == BC);
break;
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
case UNITS:
tmask = 0;
ptype = val;
break;
Rename TIME token to ISOTIME to eliminate conflict with gram.y parser token. Seems to be isolated to datetime.c and datetime.h.
2002-01-01 03:54:33 +01:00
case ISOTIME:
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
tmask = 0;
/***
* We will need one of the following fields:
pgindent run.
2002-09-04 22:31:48 +02:00
* DTK_NUMBER should be hhmmss.fff
* DTK_TIME should be hh:mm:ss.fff
* DTK_DATE should be hhmmss-zz
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
***/
Remove unnecessary parentheses in datetime/timestamp code.
2005-05-23 19:13:14 +02:00
if (i >= nf - 1 ||
(ftype[i + 1] != DTK_NUMBER &&
ftype[i + 1] != DTK_TIME &&
ftype[i + 1] != DTK_DATE))
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
return DTERR_BAD_FORMAT;
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
ptype = val;
break;
Fix up timetz input so that a date is required only when the specified timezone actually has a daylight-savings rule. This avoids breaking cases that used to work because they went through the DecodePosixTimezone code path. Per contrib regression failures (mea culpa for not running those yesterday...). Also document the already-applied change to allow GMT offsets up to 14 hours.
2006-10-18 18:43:14 +02:00
case UNKNOWN_FIELD:
pgindent run for 8.3.
2007-11-15 22:14:46 +01:00
Fix up timetz input so that a date is required only when the specified timezone actually has a daylight-savings rule. This avoids breaking cases that used to work because they went through the DecodePosixTimezone code path. Per contrib regression failures (mea culpa for not running those yesterday...). Also document the already-applied change to allow GMT offsets up to 14 hours.
2006-10-18 18:43:14 +02:00
/*
* Before giving up and declaring error, check to see
* if it is an all-alpha timezone name.
*/
namedTz = pg_tzset(field[i]);
if (!namedTz)
return DTERR_BAD_FORMAT;
/* we'll apply the zone setting below */
tmask = DTK_M(TZ);
break;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
default:
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
return DTERR_BAD_FORMAT;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
}
break;
default:
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
return DTERR_BAD_FORMAT;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
}
if (tmask & fmask)
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
return DTERR_BAD_FORMAT;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
fmask |= tmask;
8.4 pgindent run, with new combined Linux/FreeBSD/MinGW typedef list provided by Andrew.
2009-06-11 16:49:15 +02:00
} /* end loop over fields */
Fix datetime input to behave correctly for Feb 29 in years BC. Formerly, DecodeDate attempted to verify the day-of-the-month exactly, but it was under the misapprehension that it would know whether we were looking at a BC year or not. In reality this check can't be made until the calling function (eg DecodeDateTime) has processed all the fields. So, split the BC adjustment and validity checks out into a new function ValidateDate that is called only after processing all the fields. In passing, this patch makes DecodeTimeOnly work for BC inputs, which it never did before. (The historical veracity of all this is nonexistent, of course, but if we're going to say we support proleptic Gregorian calendar then we should do it correctly. In any case the unpatched code is broken because it could emit dates that it would then reject on re-inputting.) Per report from Bernd Helmle. Back-patch as far as 8.0; in 7.x we were not using our own calendar support and so this seems a bit too risky to put into 7.4.
2008-02-26 00:21:01 +01:00
/* do final checking/adjustment of Y/M/D fields */
Re-allow input of Julian dates prior to 0001-01-01 AD. This was unintentionally broken in 8.4 while tightening up checking of ordinary non-Julian date inputs to forbid references to "year zero". Per bug #5672 from Benjamin Gigot.
2010-09-23 05:48:07 +02:00
dterr = ValidateDate(fmask, isjulian, is2digits, bc, tm);
Fix datetime input to behave correctly for Feb 29 in years BC. Formerly, DecodeDate attempted to verify the day-of-the-month exactly, but it was under the misapprehension that it would know whether we were looking at a BC year or not. In reality this check can't be made until the calling function (eg DecodeDateTime) has processed all the fields. So, split the BC adjustment and validity checks out into a new function ValidateDate that is called only after processing all the fields. In passing, this patch makes DecodeTimeOnly work for BC inputs, which it never did before. (The historical veracity of all this is nonexistent, of course, but if we're going to say we support proleptic Gregorian calendar then we should do it correctly. In any case the unpatched code is broken because it could emit dates that it would then reject on re-inputting.) Per report from Bernd Helmle. Back-patch as far as 8.0; in 7.x we were not using our own calendar support and so this seems a bit too risky to put into 7.4.
2008-02-26 00:21:01 +01:00
if (dterr)
return dterr;
Fix datetime and abstime conversions to and from date. Bring optional new-storage date and time up to date and test. This new storage format should fix the "Sparc gcc -O2 bug". (Enable new code with USE_NEW_DATE and USE_NEW_TIME in dt.h)
1997-05-16 09:19:50 +02:00
Fix datetime input to behave correctly for Feb 29 in years BC. Formerly, DecodeDate attempted to verify the day-of-the-month exactly, but it was under the misapprehension that it would know whether we were looking at a BC year or not. In reality this check can't be made until the calling function (eg DecodeDateTime) has processed all the fields. So, split the BC adjustment and validity checks out into a new function ValidateDate that is called only after processing all the fields. In passing, this patch makes DecodeTimeOnly work for BC inputs, which it never did before. (The historical veracity of all this is nonexistent, of course, but if we're going to say we support proleptic Gregorian calendar then we should do it correctly. In any case the unpatched code is broken because it could emit dates that it would then reject on re-inputting.) Per report from Bernd Helmle. Back-patch as far as 8.0; in 7.x we were not using our own calendar support and so this seems a bit too risky to put into 7.4.
2008-02-26 00:21:01 +01:00
/* handle AM/PM */
Use macros for time-based constants, rather than constants.
2011-03-12 15:31:18 +01:00
if (mer != HR24 && tm->tm_hour > HOURS_PER_DAY / 2)
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
return DTERR_FIELD_OVERFLOW;
Use macros for time-based constants, rather than constants.
2011-03-12 15:31:18 +01:00
if (mer == AM && tm->tm_hour == HOURS_PER_DAY / 2)
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
tm->tm_hour = 0;
Use macros for time-based constants, rather than constants.
2011-03-12 15:31:18 +01:00
else if (mer == PM && tm->tm_hour != HOURS_PER_DAY / 2)
tm->tm_hour += HOURS_PER_DAY / 2;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
Constructors for interval, timestamp, timestamptz Author: Pavel Stěhule, editorialized somewhat by Álvaro Herrera Reviewed-by: Tomáš Vondra, Marko Tiikkaja With input from Fabrízio de Royes Mello, Jim Nasby
2014-03-04 19:09:43 +01:00
/*
* This should match the checks in make_timestamp_internal
*/
Use macros for time-based constants, rather than constants.
2011-03-12 15:31:18 +01:00
if (tm->tm_hour < 0 || tm->tm_min < 0 || tm->tm_min > MINS_PER_HOUR - 1 ||
tm->tm_sec < 0 || tm->tm_sec > SECS_PER_MINUTE ||
tm->tm_hour > HOURS_PER_DAY ||
8.4 pgindent run, with new combined Linux/FreeBSD/MinGW typedef list provided by Andrew.
2009-06-11 16:49:15 +02:00
/* test for > 24:00:00 */
Use macros for time-based constants, rather than constants.
2011-03-12 15:31:18 +01:00
(tm->tm_hour == HOURS_PER_DAY &&
When checking for datetime field overflow, we should allow a fractional-second part that rounds up to exactly 1.0 second. The previous coding rejected input like "00:12:57.9999999999999999999999999999", with the exact number of nines needed to cause failure varying depending on float-timestamp option and possibly on platform. Obviously this should round up to the next integral second, if we don't have enough precision to distinguish the value from that. Per bug #4789 from Robert Kruus. In passing, fix a missed check for fractional seconds in one copy of the "is it greater than 24:00:00" code. Broken all the way back, so patch all the way back.
2009-05-01 21:29:07 +02:00
(tm->tm_min > 0 || tm->tm_sec > 0 || *fsec > 0)) ||
Remove now-dead code for !HAVE_INT64_TIMESTAMP. This is a basically mechanical removal of #ifdef HAVE_INT64_TIMESTAMP tests and the negative-case controlled code. Discussion: https://postgr.es/m/26788.1487455319@sss.pgh.pa.us
2017-02-23 20:04:43 +01:00
*fsec < INT64CONST(0) || *fsec > USECS_PER_SEC)
Fix up some problems in handling of zic-style time zone names in datetime input routines. Remove the former "DecodePosixTimezone" function in favor of letting the zic code handle POSIX-style zone specs (see tzparse()). In particular this means that "PST+3" now means the same as "-03", whereas it used to mean "-11" --- the zone abbreviation is effectively just a noise word in this syntax. Make sure that all named and POSIX-style zone names will be parsed as a single token. Fix long-standing bogosities in printing and input of fractional-hour timezone offsets (since the tzparse() code will accept these, we'd better make 'em work). Also correct an error in the original coding of the zic-zone-name patch: in "timestamp without time zone" input, zone names are supposed to be allowed but ignored, but the coding was such that the zone changed the interpretation anyway.
2006-10-17 23:03:21 +02:00
return DTERR_FIELD_OVERFLOW;
From: Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> Subject: [HACKERS] More date time functions Here are some additional patches mostly related to the date and time data types. It includes some type conversion routines to move between the different date types and some other date manipulation routines such as date_part(units,datetime). I noticed Edmund Mergl et al's neat trick for getting function overloading for builtin functions, so started to use that for the date and time stuff. Later, if someone figures out how to get function overloading directly for internal C code, then we can move to that technique. These patches include documentation updates (don't faint!) for the built-in man page. Doesn't yet include mention of timestamp, since I don't know much about it and since it may change a bit to become a _real_ ANSI timestamp which would include parser support for the declaration syntax (what do you think, Dan?). The patches were developed on the 970330 release, but have been rebuilt off of the 970402 release. The first patch below is to get libpq to compile, on my Linux box, but is not related to the rest of the patches and you can choose not to apply that one at this time. Thanks in advance, scrappy!
1997-04-02 20:36:24 +02:00
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
if ((fmask & DTK_TIME_M) != DTK_TIME_M)
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
return DTERR_BAD_FORMAT;
From: Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> Subject: [HACKERS] More date time functions Here are some additional patches mostly related to the date and time data types. It includes some type conversion routines to move between the different date types and some other date manipulation routines such as date_part(units,datetime). I noticed Edmund Mergl et al's neat trick for getting function overloading for builtin functions, so started to use that for the date and time stuff. Later, if someone figures out how to get function overloading directly for internal C code, then we can move to that technique. These patches include documentation updates (don't faint!) for the built-in man page. Doesn't yet include mention of timestamp, since I don't know much about it and since it may change a bit to become a _real_ ANSI timestamp which would include parser support for the declaration syntax (what do you think, Dan?). The patches were developed on the 970330 release, but have been rebuilt off of the 970402 release. The first patch below is to get libpq to compile, on my Linux box, but is not related to the rest of the patches and you can choose not to apply that one at this time. Thanks in advance, scrappy!
1997-04-02 20:36:24 +02:00
Fix up some problems in handling of zic-style time zone names in datetime input routines. Remove the former "DecodePosixTimezone" function in favor of letting the zic code handle POSIX-style zone specs (see tzparse()). In particular this means that "PST+3" now means the same as "-03", whereas it used to mean "-11" --- the zone abbreviation is effectively just a noise word in this syntax. Make sure that all named and POSIX-style zone names will be parsed as a single token. Fix long-standing bogosities in printing and input of fractional-hour timezone offsets (since the tzparse() code will accept these, we'd better make 'em work). Also correct an error in the original coding of the zic-zone-name patch: in "timestamp without time zone" input, zone names are supposed to be allowed but ignored, but the coding was such that the zone changed the interpretation anyway.
2006-10-17 23:03:21 +02:00
/*
pgindent run for 8.3.
2007-11-15 22:14:46 +01:00
* If we had a full timezone spec, compute the offset (we could not do it
* before, because we may need the date to resolve DST status).
Fix up some problems in handling of zic-style time zone names in datetime input routines. Remove the former "DecodePosixTimezone" function in favor of letting the zic code handle POSIX-style zone specs (see tzparse()). In particular this means that "PST+3" now means the same as "-03", whereas it used to mean "-11" --- the zone abbreviation is effectively just a noise word in this syntax. Make sure that all named and POSIX-style zone names will be parsed as a single token. Fix long-standing bogosities in printing and input of fractional-hour timezone offsets (since the tzparse() code will accept these, we'd better make 'em work). Also correct an error in the original coding of the zic-zone-name patch: in "timestamp without time zone" input, zone names are supposed to be allowed but ignored, but the coding was such that the zone changed the interpretation anyway.
2006-10-17 23:03:21 +02:00
*/
if (namedTz != NULL)
{
Fix up timetz input so that a date is required only when the specified timezone actually has a daylight-savings rule. This avoids breaking cases that used to work because they went through the DecodePosixTimezone code path. Per contrib regression failures (mea culpa for not running those yesterday...). Also document the already-applied change to allow GMT offsets up to 14 hours.
2006-10-18 18:43:14 +02:00
long int gmtoff;
Fix up some problems in handling of zic-style time zone names in datetime input routines. Remove the former "DecodePosixTimezone" function in favor of letting the zic code handle POSIX-style zone specs (see tzparse()). In particular this means that "PST+3" now means the same as "-03", whereas it used to mean "-11" --- the zone abbreviation is effectively just a noise word in this syntax. Make sure that all named and POSIX-style zone names will be parsed as a single token. Fix long-standing bogosities in printing and input of fractional-hour timezone offsets (since the tzparse() code will accept these, we'd better make 'em work). Also correct an error in the original coding of the zic-zone-name patch: in "timestamp without time zone" input, zone names are supposed to be allowed but ignored, but the coding was such that the zone changed the interpretation anyway.
2006-10-17 23:03:21 +02:00
/* daylight savings time modifier disallowed with full TZ */
if (fmask & DTK_M(DTZMOD))
return DTERR_BAD_FORMAT;
Fix up timetz input so that a date is required only when the specified timezone actually has a daylight-savings rule. This avoids breaking cases that used to work because they went through the DecodePosixTimezone code path. Per contrib regression failures (mea culpa for not running those yesterday...). Also document the already-applied change to allow GMT offsets up to 14 hours.
2006-10-18 18:43:14 +02:00
/* if non-DST zone, we do not need to know the date */
if (pg_get_timezone_offset(namedTz, &gmtoff))
{
*tzp = -(int) gmtoff;
}
else
{
/* a date has to be specified */
if ((fmask & DTK_DATE_M) != DTK_DATE_M)
return DTERR_BAD_FORMAT;
*tzp = DetermineTimeZoneOffset(tm, namedTz);
}
Fix up some problems in handling of zic-style time zone names in datetime input routines. Remove the former "DecodePosixTimezone" function in favor of letting the zic code handle POSIX-style zone specs (see tzparse()). In particular this means that "PST+3" now means the same as "-03", whereas it used to mean "-11" --- the zone abbreviation is effectively just a noise word in this syntax. Make sure that all named and POSIX-style zone names will be parsed as a single token. Fix long-standing bogosities in printing and input of fractional-hour timezone offsets (since the tzparse() code will accept these, we'd better make 'em work). Also correct an error in the original coding of the zic-zone-name patch: in "timestamp without time zone" input, zone names are supposed to be allowed but ignored, but the coding was such that the zone changed the interpretation anyway.
2006-10-17 23:03:21 +02:00
}
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
/*
* Likewise, if we had a dynamic timezone abbreviation, resolve it now.
*/
if (abbrevTz != NULL)
{
struct pg_tm tt,
*tmp = &tt;
/*
* daylight savings time modifier but no standard timezone? then error
*/
if (fmask & DTK_M(DTZMOD))
return DTERR_BAD_FORMAT;
if ((fmask & DTK_DATE_M) == 0)
GetCurrentDateTime(tmp);
else
{
Fix some incorrect parsing of time with time zone strings When parsing a timetz string with a dynamic timezone abbreviation or a timezone not specified, it was possible to generate incorrect timestamps based on a date which uses some non-initialized variables if the input string did not specify fully a date to parse. This is already checked when a full timezone spec is included in the input string, but the two other cases mentioned above missed the same checks. This gets fixed by generating an error as this input is invalid, or in short when a date is not fully specified. Valgrind was complaining about this problem. Bug: #15910 Author: Alexander Lakhin Discussion: https://postgr.es/m/15910-2eba5106b9aa0c61@postgresql.org Backpatch-through: 9.4
2019-08-07 11:16:31 +02:00
/* a date has to be specified */
if ((fmask & DTK_DATE_M) != DTK_DATE_M)
return DTERR_BAD_FORMAT;
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
tmp->tm_year = tm->tm_year;
tmp->tm_mon = tm->tm_mon;
tmp->tm_mday = tm->tm_mday;
}
tmp->tm_hour = tm->tm_hour;
tmp->tm_min = tm->tm_min;
tmp->tm_sec = tm->tm_sec;
*tzp = DetermineTimeZoneAbbrevOffset(tmp, abbrev, abbrevTz);
tm->tm_isdst = tmp->tm_isdst;
}
/* timezone not specified? then use session timezone */
Remove excess parens, use Abs instead of : ?.
2005-05-21 05:38:05 +02:00
if (tzp != NULL && !(fmask & DTK_M(TZ)))
Implement column aliases on views "CREATE VIEW name (collist)". Implement TIME WITH TIME ZONE type (timetz internal type). Remap length() for character strings to CHAR_LENGTH() for SQL92 and to remove the ambiguity with geometric length() functions. Keep length() for character strings for backward compatibility. Shrink stored views by removing internal column name list from visible rte. Implement min(), max() for time and timetz data types. Implement conversion of TIME to INTERVAL. Implement abs(), mod(), fac() for the int8 data type. Rename some math functions to generic names: round(), sqrt(), cbrt(), pow(), etc. Rename NUMERIC power() function to pow(). Fix int2 factorial to calculate result in int4. Enhance the Oracle compatibility function translate() to work with string arguments (from Edwin Ramirez). Modify pg_proc system table to remove OID holes.
2000-03-15 00:06:59 +01:00
{
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
struct pg_tm tt,
Ye-old pgindent run. Same 4-space tabs.
2000-04-12 19:17:23 +02:00
*tmp = &tt;
Implement column aliases on views "CREATE VIEW name (collist)". Implement TIME WITH TIME ZONE type (timetz internal type). Remap length() for character strings to CHAR_LENGTH() for SQL92 and to remove the ambiguity with geometric length() functions. Keep length() for character strings for backward compatibility. Shrink stored views by removing internal column name list from visible rte. Implement min(), max() for time and timetz data types. Implement conversion of TIME to INTERVAL. Implement abs(), mod(), fac() for the int8 data type. Rename some math functions to generic names: round(), sqrt(), cbrt(), pow(), etc. Rename NUMERIC power() function to pow(). Fix int2 factorial to calculate result in int4. Enhance the Oracle compatibility function translate() to work with string arguments (from Edwin Ramirez). Modify pg_proc system table to remove OID holes.
2000-03-15 00:06:59 +01:00
/*
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* daylight savings time modifier but no standard timezone? then error
Implement column aliases on views "CREATE VIEW name (collist)". Implement TIME WITH TIME ZONE type (timetz internal type). Remap length() for character strings to CHAR_LENGTH() for SQL92 and to remove the ambiguity with geometric length() functions. Keep length() for character strings for backward compatibility. Shrink stored views by removing internal column name list from visible rte. Implement min(), max() for time and timetz data types. Implement conversion of TIME to INTERVAL. Implement abs(), mod(), fac() for the int8 data type. Rename some math functions to generic names: round(), sqrt(), cbrt(), pow(), etc. Rename NUMERIC power() function to pow(). Fix int2 factorial to calculate result in int4. Enhance the Oracle compatibility function translate() to work with string arguments (from Edwin Ramirez). Modify pg_proc system table to remove OID holes.
2000-03-15 00:06:59 +01:00
*/
if (fmask & DTK_M(DTZMOD))
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
return DTERR_BAD_FORMAT;
Implement column aliases on views "CREATE VIEW name (collist)". Implement TIME WITH TIME ZONE type (timetz internal type). Remap length() for character strings to CHAR_LENGTH() for SQL92 and to remove the ambiguity with geometric length() functions. Keep length() for character strings for backward compatibility. Shrink stored views by removing internal column name list from visible rte. Implement min(), max() for time and timetz data types. Implement conversion of TIME to INTERVAL. Implement abs(), mod(), fac() for the int8 data type. Rename some math functions to generic names: round(), sqrt(), cbrt(), pow(), etc. Rename NUMERIC power() function to pow(). Fix int2 factorial to calculate result in int4. Enhance the Oracle compatibility function translate() to work with string arguments (from Edwin Ramirez). Modify pg_proc system table to remove OID holes.
2000-03-15 00:06:59 +01:00
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
if ((fmask & DTK_DATE_M) == 0)
Katherine Ward wrote: > Changes to avoid collisions with WIN32 & MFC names... > 1. Renamed: > a. PROC => PGPROC > b. GetUserName() => GetUserNameFromId() > c. GetCurrentTime() => GetCurrentDateTime() > d. IGNORE => IGNORE_DTF in include/utils/datetime.h & utils/adt/datetim > > 2. Added _P to some lex/yacc tokens: > CONST, CHAR, DELETE, FLOAT, GROUP, IN, OUT Jan
2002-06-11 15:40:53 +02:00
GetCurrentDateTime(tmp);
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
else
{
Fix some incorrect parsing of time with time zone strings When parsing a timetz string with a dynamic timezone abbreviation or a timezone not specified, it was possible to generate incorrect timestamps based on a date which uses some non-initialized variables if the input string did not specify fully a date to parse. This is already checked when a full timezone spec is included in the input string, but the two other cases mentioned above missed the same checks. This gets fixed by generating an error as this input is invalid, or in short when a date is not fully specified. Valgrind was complaining about this problem. Bug: #15910 Author: Alexander Lakhin Discussion: https://postgr.es/m/15910-2eba5106b9aa0c61@postgresql.org Backpatch-through: 9.4
2019-08-07 11:16:31 +02:00
/* a date has to be specified */
if ((fmask & DTK_DATE_M) != DTK_DATE_M)
return DTERR_BAD_FORMAT;
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
tmp->tm_year = tm->tm_year;
tmp->tm_mon = tm->tm_mon;
tmp->tm_mday = tm->tm_mday;
}
Implement column aliases on views "CREATE VIEW name (collist)". Implement TIME WITH TIME ZONE type (timetz internal type). Remap length() for character strings to CHAR_LENGTH() for SQL92 and to remove the ambiguity with geometric length() functions. Keep length() for character strings for backward compatibility. Shrink stored views by removing internal column name list from visible rte. Implement min(), max() for time and timetz data types. Implement conversion of TIME to INTERVAL. Implement abs(), mod(), fac() for the int8 data type. Rename some math functions to generic names: round(), sqrt(), cbrt(), pow(), etc. Rename NUMERIC power() function to pow(). Fix int2 factorial to calculate result in int4. Enhance the Oracle compatibility function translate() to work with string arguments (from Edwin Ramirez). Modify pg_proc system table to remove OID holes.
2000-03-15 00:06:59 +01:00
tmp->tm_hour = tm->tm_hour;
tmp->tm_min = tm->tm_min;
tmp->tm_sec = tm->tm_sec;
Switch over to using the src/timezone functions for formatting timestamps displayed in the postmaster log. This avoids Windows-specific problems with localized time zone names that are in the wrong encoding, and generally seems like a good idea to forestall other potential platform-dependent issues. To preserve the existing behavior that all backends will log in the same time zone, create a new GUC variable log_timezone that can only be changed on a system-wide basis, and reference log-related calculations to that zone instead of the TimeZone variable. This fixes the issue reported by Hiroshi Saito that timestamps printed by xlog.c startup could be improperly localized on Windows. We still need a simpler patch for that problem in the back branches, however.
2007-08-04 03:26:54 +02:00
*tzp = DetermineTimeZoneOffset(tmp, session_timezone);
Implement column aliases on views "CREATE VIEW name (collist)". Implement TIME WITH TIME ZONE type (timetz internal type). Remap length() for character strings to CHAR_LENGTH() for SQL92 and to remove the ambiguity with geometric length() functions. Keep length() for character strings for backward compatibility. Shrink stored views by removing internal column name list from visible rte. Implement min(), max() for time and timetz data types. Implement conversion of TIME to INTERVAL. Implement abs(), mod(), fac() for the int8 data type. Rename some math functions to generic names: round(), sqrt(), cbrt(), pow(), etc. Rename NUMERIC power() function to pow(). Fix int2 factorial to calculate result in int4. Enhance the Oracle compatibility function translate() to work with string arguments (from Edwin Ramirez). Modify pg_proc system table to remove OID holes.
2000-03-15 00:06:59 +01:00
tm->tm_isdst = tmp->tm_isdst;
}
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
return 0;
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
}
From: Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> Subject: [HACKERS] More date time functions Here are some additional patches mostly related to the date and time data types. It includes some type conversion routines to move between the different date types and some other date manipulation routines such as date_part(units,datetime). I noticed Edmund Mergl et al's neat trick for getting function overloading for builtin functions, so started to use that for the date and time stuff. Later, if someone figures out how to get function overloading directly for internal C code, then we can move to that technique. These patches include documentation updates (don't faint!) for the built-in man page. Doesn't yet include mention of timestamp, since I don't know much about it and since it may change a bit to become a _real_ ANSI timestamp which would include parser support for the declaration syntax (what do you think, Dan?). The patches were developed on the 970330 release, but have been rebuilt off of the 970402 release. The first patch below is to get libpq to compile, on my Linux box, but is not related to the rest of the patches and you can choose not to apply that one at this time. Thanks in advance, scrappy!
1997-04-02 20:36:24 +02:00
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
/* DecodeDate()
* Decode date string which includes delimiters.
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
* Return 0 if okay, a DTERR code if not.
*
Fix datetime input to behave correctly for Feb 29 in years BC. Formerly, DecodeDate attempted to verify the day-of-the-month exactly, but it was under the misapprehension that it would know whether we were looking at a BC year or not. In reality this check can't be made until the calling function (eg DecodeDateTime) has processed all the fields. So, split the BC adjustment and validity checks out into a new function ValidateDate that is called only after processing all the fields. In passing, this patch makes DecodeTimeOnly work for BC inputs, which it never did before. (The historical veracity of all this is nonexistent, of course, but if we're going to say we support proleptic Gregorian calendar then we should do it correctly. In any case the unpatched code is broken because it could emit dates that it would then reject on re-inputting.) Per report from Bernd Helmle. Back-patch as far as 8.0; in 7.x we were not using our own calendar support and so this seems a bit too risky to put into 7.4.
2008-02-26 00:21:01 +01:00
* str: field to be parsed
* fmask: bitmask for field types already seen
* *tmask: receives bitmask for fields found here
Change TRUE/FALSE to true/false The lower case spellings are C and C++ standard and are used in most parts of the PostgreSQL sources. The upper case spellings are only used in some files/modules. So standardize on the standard spellings. The APIs for ICU, Perl, and Windows define their own TRUE and FALSE, so those are left as is when using those APIs. In code comments, we use the lower-case spelling for the C concepts and keep the upper-case spelling for the SQL concepts. Reviewed-by: Michael Paquier <michael.paquier@gmail.com>
2017-08-16 06:22:32 +02:00
* *is2digits: set to true if we find 2-digit year
Fix datetime input to behave correctly for Feb 29 in years BC. Formerly, DecodeDate attempted to verify the day-of-the-month exactly, but it was under the misapprehension that it would know whether we were looking at a BC year or not. In reality this check can't be made until the calling function (eg DecodeDateTime) has processed all the fields. So, split the BC adjustment and validity checks out into a new function ValidateDate that is called only after processing all the fields. In passing, this patch makes DecodeTimeOnly work for BC inputs, which it never did before. (The historical veracity of all this is nonexistent, of course, but if we're going to say we support proleptic Gregorian calendar then we should do it correctly. In any case the unpatched code is broken because it could emit dates that it would then reject on re-inputting.) Per report from Bernd Helmle. Back-patch as far as 8.0; in 7.x we were not using our own calendar support and so this seems a bit too risky to put into 7.4.
2008-02-26 00:21:01 +01:00
* *tm: field values are stored into appropriate members of this struct
From: Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> Subject: [HACKERS] More date time functions Here are some additional patches mostly related to the date and time data types. It includes some type conversion routines to move between the different date types and some other date manipulation routines such as date_part(units,datetime). I noticed Edmund Mergl et al's neat trick for getting function overloading for builtin functions, so started to use that for the date and time stuff. Later, if someone figures out how to get function overloading directly for internal C code, then we can move to that technique. These patches include documentation updates (don't faint!) for the built-in man page. Doesn't yet include mention of timestamp, since I don't know much about it and since it may change a bit to become a _real_ ANSI timestamp which would include parser support for the declaration syntax (what do you think, Dan?). The patches were developed on the 970330 release, but have been rebuilt off of the 970402 release. The first patch below is to get libpq to compile, on my Linux box, but is not related to the rest of the patches and you can choose not to apply that one at this time. Thanks in advance, scrappy!
1997-04-02 20:36:24 +02:00
*/
Mark functions as static and ifdef NOT_USED as appropriate.
2000-06-09 00:38:00 +02:00
static int
Fix datetime input to behave correctly for Feb 29 in years BC. Formerly, DecodeDate attempted to verify the day-of-the-month exactly, but it was under the misapprehension that it would know whether we were looking at a BC year or not. In reality this check can't be made until the calling function (eg DecodeDateTime) has processed all the fields. So, split the BC adjustment and validity checks out into a new function ValidateDate that is called only after processing all the fields. In passing, this patch makes DecodeTimeOnly work for BC inputs, which it never did before. (The historical veracity of all this is nonexistent, of course, but if we're going to say we support proleptic Gregorian calendar then we should do it correctly. In any case the unpatched code is broken because it could emit dates that it would then reject on re-inputting.) Per report from Bernd Helmle. Back-patch as far as 8.0; in 7.x we were not using our own calendar support and so this seems a bit too risky to put into 7.4.
2008-02-26 00:21:01 +01:00
DecodeDate(char *str, int fmask, int *tmask, bool *is2digits,
Initial pgindent run with pg_bsd_indent version 2.0. The new indent version includes numerous fixes thanks to Piotr Stefaniak. The main changes visible in this commit are: * Nicer formatting of function-pointer declarations. * No longer unexpectedly removes spaces in expressions using casts, sizeof, or offsetof. * No longer wants to add a space in "struct structname *varname", as well as some similar cases for const- or volatile-qualified pointers. * Declarations using PG_USED_FOR_ASSERTS_ONLY are formatted more nicely. * Fixes bug where comments following declarations were sometimes placed with no space separating them from the code. * Fixes some odd decisions for comments following case labels. * Fixes some cases where comments following code were indented to less than the expected column 33. On the less good side, it now tends to put more whitespace around typedef names that are not listed in typedefs.list. This might encourage us to put more effort into typedef name collection; it's not really a bug in indent itself. There are more changes coming after this round, having to do with comment indentation and alignment of lines appearing within parentheses. I wanted to limit the size of the diffs to something that could be reviewed without one's eyes completely glazing over, so it seemed better to split up the changes as much as practical. Discussion: https://postgr.es/m/E1dAmxK-0006EE-1r@gemulon.postgresql.org Discussion: https://postgr.es/m/30527.1495162840@sss.pgh.pa.us
2017-06-21 20:39:04 +02:00
struct pg_tm *tm)
From: Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> Subject: [HACKERS] More date time functions Here are some additional patches mostly related to the date and time data types. It includes some type conversion routines to move between the different date types and some other date manipulation routines such as date_part(units,datetime). I noticed Edmund Mergl et al's neat trick for getting function overloading for builtin functions, so started to use that for the date and time stuff. Later, if someone figures out how to get function overloading directly for internal C code, then we can move to that technique. These patches include documentation updates (don't faint!) for the built-in man page. Doesn't yet include mention of timestamp, since I don't know much about it and since it may change a bit to become a _real_ ANSI timestamp which would include parser support for the declaration syntax (what do you think, Dan?). The patches were developed on the 970330 release, but have been rebuilt off of the 970402 release. The first patch below is to get libpq to compile, on my Linux box, but is not related to the rest of the patches and you can choose not to apply that one at this time. Thanks in advance, scrappy!
1997-04-02 20:36:24 +02:00
{
Support alternate storage scheme of 64-bit integer for date/time types. Use "--enable-integer-datetimes" in configuration to use this rather than the original float8 storage. I would recommend the integer-based storage for any platform on which it is available. We perhaps should make this the default for the production release. Change timezone(timestamptz) results to return timestamp rather than a character string. Formerly, we didn't have a way to represent timestamps with an explicit time zone other than freezing the info into a string. Now, we can reasonably omit the explicit time zone from the result and return a timestamp with values appropriate for the specified time zone. Much cleaner, and if you need the time zone in the result you can put it into a character string pretty easily anyway. Allow fractional seconds in date/time types even for dates prior to 1BC. Limit timestamp data types to 6 decimal places of precision. Just right for a micro-second storage of int8 date/time types, and reduces the number of places ad-hoc rounding was occuring for the float8-based types. Use lookup tables for precision/rounding calculations for timestamp and interval types. Formerly used pow() to calculate the desired value but with a more limited range there is no reason to not type in a lookup table. Should be *much* better performance, though formerly there were some optimizations to help minimize the number of times pow() was called. Define a HAVE_INT64_TIMESTAMP variable. Based on the configure option "--enable-integer-datetimes" and the existing internal INT64_IS_BUSTED. Add explicit date/interval operators and functions for addition and subtraction. Formerly relied on implicit type promotion from date to timestamp with time zone. Change timezone conversion functions for the timetz type from "timetz()" to "timezone()". This is consistant with other time zone coersion functions for other types. Bump the catalog version to 200204201. Fix up regression tests to reflect changes in fractional seconds representation for date/times in BC eras. All regression tests pass on my Linux box.
2002-04-21 21:52:18 +02:00
fsec_t fsec;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
int nf = 0;
int i,
len;
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
int dterr;
Change TRUE/FALSE to true/false The lower case spellings are C and C++ standard and are used in most parts of the PostgreSQL sources. The upper case spellings are only used in some files/modules. So standardize on the standard spellings. The APIs for ICU, Perl, and Windows define their own TRUE and FALSE, so those are left as is when using those APIs. In code comments, we use the lower-case spelling for the C concepts and keep the upper-case spelling for the SQL concepts. Reviewed-by: Michael Paquier <michael.paquier@gmail.com>
2017-08-16 06:22:32 +02:00
bool haveTextMonth = false;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
int type,
val,
dmask = 0;
char *field[MAXDATEFIELDS];
Fix datetime input to behave correctly for Feb 29 in years BC. Formerly, DecodeDate attempted to verify the day-of-the-month exactly, but it was under the misapprehension that it would know whether we were looking at a BC year or not. In reality this check can't be made until the calling function (eg DecodeDateTime) has processed all the fields. So, split the BC adjustment and validity checks out into a new function ValidateDate that is called only after processing all the fields. In passing, this patch makes DecodeTimeOnly work for BC inputs, which it never did before. (The historical veracity of all this is nonexistent, of course, but if we're going to say we support proleptic Gregorian calendar then we should do it correctly. In any case the unpatched code is broken because it could emit dates that it would then reject on re-inputting.) Per report from Bernd Helmle. Back-patch as far as 8.0; in 7.x we were not using our own calendar support and so this seems a bit too risky to put into 7.4.
2008-02-26 00:21:01 +01:00
*tmask = 0;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
/* parse this string... */
Remove unnecessary parentheses in datetime/timestamp code.
2005-05-23 19:13:14 +02:00
while (*str != '\0' && nf < MAXDATEFIELDS)
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
{
/* skip field separators */
Fix access past end of string in date parsing. This affects date_in(), and a couple of other funcions that use DecodeDate(). Hitoshi Harada
2012-10-02 09:43:48 +02:00
while (*str != '\0' && !isalnum((unsigned char) *str))
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
str++;
Fix access past end of string in date parsing. This affects date_in(), and a couple of other funcions that use DecodeDate(). Hitoshi Harada
2012-10-02 09:43:48 +02:00
if (*str == '\0')
pgindent run for release 9.3 This is the first run of the Perl-based pgindent script. Also update pgindent instructions.
2013-05-29 22:58:43 +02:00
return DTERR_BAD_FORMAT; /* end of string after separator */
Fix access past end of string in date parsing. This affects date_in(), and a couple of other funcions that use DecodeDate(). Hitoshi Harada
2012-10-02 09:43:48 +02:00
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
field[nf] = str;
Ensure that all uses of <ctype.h> functions are applied to unsigned-char values, whether the local char type is signed or not. This is necessary for portability. Per discussion on pghackers around 9/16/00.
2000-12-03 21:45:40 +01:00
if (isdigit((unsigned char) *str))
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
{
Ensure that all uses of <ctype.h> functions are applied to unsigned-char values, whether the local char type is signed or not. This is necessary for portability. Per discussion on pghackers around 9/16/00.
2000-12-03 21:45:40 +01:00
while (isdigit((unsigned char) *str))
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
str++;
}
Ensure that all uses of <ctype.h> functions are applied to unsigned-char values, whether the local char type is signed or not. This is necessary for portability. Per discussion on pghackers around 9/16/00.
2000-12-03 21:45:40 +01:00
else if (isalpha((unsigned char) *str))
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
{
Ensure that all uses of <ctype.h> functions are applied to unsigned-char values, whether the local char type is signed or not. This is necessary for portability. Per discussion on pghackers around 9/16/00.
2000-12-03 21:45:40 +01:00
while (isalpha((unsigned char) *str))
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
str++;
}
Measure the current transaction time to milliseconds. Define a new function, GetCurrentTransactionStartTimeUsec() to get the time to this precision. Allow now() and timestamp 'now' to use this higher precision result so we now have fractional seconds in this "constant". Add timestamp without time zone type. Move previous timestamp type to timestamp with time zone. Accept another ISO variant for date/time values: yyyy-mm-ddThh:mm:ss (note the "T" separating the day from hours information). Remove 'current' from date/time types; convert to 'now' in input. Separate time and timetz regression tests. Separate timestamp and timestamptz regression test.
2001-09-28 10:09:14 +02:00
/* Just get rid of any non-digit, non-alpha characters... */
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
if (*str != '\0')
*str++ = '\0';
nf++;
}
/* look first for text fields, since that will be unambiguous month */
for (i = 0; i < nf; i++)
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
{
Ensure that all uses of <ctype.h> functions are applied to unsigned-char values, whether the local char type is signed or not. This is necessary for portability. Per discussion on pghackers around 9/16/00.
2000-12-03 21:45:40 +01:00
if (isalpha((unsigned char) *field[i]))
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
{
type = DecodeSpecial(i, field[i], &val);
Katherine Ward wrote: > Changes to avoid collisions with WIN32 & MFC names... > 1. Renamed: > a. PROC => PGPROC > b. GetUserName() => GetUserNameFromId() > c. GetCurrentTime() => GetCurrentDateTime() > d. IGNORE => IGNORE_DTF in include/utils/datetime.h & utils/adt/datetim > > 2. Added _P to some lex/yacc tokens: > CONST, CHAR, DELETE, FLOAT, GROUP, IN, OUT Jan
2002-06-11 15:40:53 +02:00
if (type == IGNORE_DTF)
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
continue;
dmask = DTK_M(type);
switch (type)
{
case MONTH:
tm->tm_mon = val;
Change TRUE/FALSE to true/false The lower case spellings are C and C++ standard and are used in most parts of the PostgreSQL sources. The upper case spellings are only used in some files/modules. So standardize on the standard spellings. The APIs for ICU, Perl, and Windows define their own TRUE and FALSE, so those are left as is when using those APIs. In code comments, we use the lower-case spelling for the C concepts and keep the upper-case spelling for the SQL concepts. Reviewed-by: Michael Paquier <michael.paquier@gmail.com>
2017-08-16 06:22:32 +02:00
haveTextMonth = true;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
break;
default:
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
return DTERR_BAD_FORMAT;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
}
if (fmask & dmask)
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
return DTERR_BAD_FORMAT;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
fmask |= dmask;
*tmask |= dmask;
/* mark this field as being completed */
field[i] = NULL;
}
}
/* now pick up remaining numeric fields */
for (i = 0; i < nf; i++)
{
if (field[i] == NULL)
continue;
if ((len = strlen(field[i])) <= 0)
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
return DTERR_BAD_FORMAT;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
Fix datetime input parsing to accept YYYY-MONTHNAME-DD and related syntaxes, which had been unintentionally broken by recent changes to tighten up the DateStyle rules for all-numeric date input. Add documentation and regression tests for this, too.
2003-11-16 21:29:16 +01:00
dterr = DecodeNumber(len, field[i], haveTextMonth, fmask,
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
&dmask, tm,
Fix datetime input to behave correctly for Feb 29 in years BC. Formerly, DecodeDate attempted to verify the day-of-the-month exactly, but it was under the misapprehension that it would know whether we were looking at a BC year or not. In reality this check can't be made until the calling function (eg DecodeDateTime) has processed all the fields. So, split the BC adjustment and validity checks out into a new function ValidateDate that is called only after processing all the fields. In passing, this patch makes DecodeTimeOnly work for BC inputs, which it never did before. (The historical veracity of all this is nonexistent, of course, but if we're going to say we support proleptic Gregorian calendar then we should do it correctly. In any case the unpatched code is broken because it could emit dates that it would then reject on re-inputting.) Per report from Bernd Helmle. Back-patch as far as 8.0; in 7.x we were not using our own calendar support and so this seems a bit too risky to put into 7.4.
2008-02-26 00:21:01 +01:00
&fsec, is2digits);
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
if (dterr)
return dterr;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
if (fmask & dmask)
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
return DTERR_BAD_FORMAT;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
fmask |= dmask;
*tmask |= dmask;
}
if ((fmask & ~(DTK_M(DOY) | DTK_M(TZ))) != DTK_DATE_M)
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
return DTERR_BAD_FORMAT;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
Fix datetime input to behave correctly for Feb 29 in years BC. Formerly, DecodeDate attempted to verify the day-of-the-month exactly, but it was under the misapprehension that it would know whether we were looking at a BC year or not. In reality this check can't be made until the calling function (eg DecodeDateTime) has processed all the fields. So, split the BC adjustment and validity checks out into a new function ValidateDate that is called only after processing all the fields. In passing, this patch makes DecodeTimeOnly work for BC inputs, which it never did before. (The historical veracity of all this is nonexistent, of course, but if we're going to say we support proleptic Gregorian calendar then we should do it correctly. In any case the unpatched code is broken because it could emit dates that it would then reject on re-inputting.) Per report from Bernd Helmle. Back-patch as far as 8.0; in 7.x we were not using our own calendar support and so this seems a bit too risky to put into 7.4.
2008-02-26 00:21:01 +01:00
/* validation of the field values must wait until ValidateDate() */
return 0;
}
/* ValidateDate()
* Check valid year/month/day values, handle BC and DOY cases
* Return 0 if okay, a DTERR code if not.
*/
Add make_date() and make_time() functions. Pavel Stehule, reviewed by Jeevan Chalke and Atri Sharma
2013-11-17 21:06:50 +01:00
int
Re-allow input of Julian dates prior to 0001-01-01 AD. This was unintentionally broken in 8.4 while tightening up checking of ordinary non-Julian date inputs to forbid references to "year zero". Per bug #5672 from Benjamin Gigot.
2010-09-23 05:48:07 +02:00
ValidateDate(int fmask, bool isjulian, bool is2digits, bool bc,
Initial pgindent run with pg_bsd_indent version 2.0. The new indent version includes numerous fixes thanks to Piotr Stefaniak. The main changes visible in this commit are: * Nicer formatting of function-pointer declarations. * No longer unexpectedly removes spaces in expressions using casts, sizeof, or offsetof. * No longer wants to add a space in "struct structname *varname", as well as some similar cases for const- or volatile-qualified pointers. * Declarations using PG_USED_FOR_ASSERTS_ONLY are formatted more nicely. * Fixes bug where comments following declarations were sometimes placed with no space separating them from the code. * Fixes some odd decisions for comments following case labels. * Fixes some cases where comments following code were indented to less than the expected column 33. On the less good side, it now tends to put more whitespace around typedef names that are not listed in typedefs.list. This might encourage us to put more effort into typedef name collection; it's not really a bug in indent itself. There are more changes coming after this round, having to do with comment indentation and alignment of lines appearing within parentheses. I wanted to limit the size of the diffs to something that could be reviewed without one's eyes completely glazing over, so it seemed better to split up the changes as much as practical. Discussion: https://postgr.es/m/E1dAmxK-0006EE-1r@gemulon.postgresql.org Discussion: https://postgr.es/m/30527.1495162840@sss.pgh.pa.us
2017-06-21 20:39:04 +02:00
struct pg_tm *tm)
Fix datetime input to behave correctly for Feb 29 in years BC. Formerly, DecodeDate attempted to verify the day-of-the-month exactly, but it was under the misapprehension that it would know whether we were looking at a BC year or not. In reality this check can't be made until the calling function (eg DecodeDateTime) has processed all the fields. So, split the BC adjustment and validity checks out into a new function ValidateDate that is called only after processing all the fields. In passing, this patch makes DecodeTimeOnly work for BC inputs, which it never did before. (The historical veracity of all this is nonexistent, of course, but if we're going to say we support proleptic Gregorian calendar then we should do it correctly. In any case the unpatched code is broken because it could emit dates that it would then reject on re-inputting.) Per report from Bernd Helmle. Back-patch as far as 8.0; in 7.x we were not using our own calendar support and so this seems a bit too risky to put into 7.4.
2008-02-26 00:21:01 +01:00
{
if (fmask & DTK_M(YEAR))
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
{
Re-allow input of Julian dates prior to 0001-01-01 AD. This was unintentionally broken in 8.4 while tightening up checking of ordinary non-Julian date inputs to forbid references to "year zero". Per bug #5672 from Benjamin Gigot.
2010-09-23 05:48:07 +02:00
if (isjulian)
{
/* tm_year is correct and should not be touched */
}
else if (bc)
Fix datetime input to behave correctly for Feb 29 in years BC. Formerly, DecodeDate attempted to verify the day-of-the-month exactly, but it was under the misapprehension that it would know whether we were looking at a BC year or not. In reality this check can't be made until the calling function (eg DecodeDateTime) has processed all the fields. So, split the BC adjustment and validity checks out into a new function ValidateDate that is called only after processing all the fields. In passing, this patch makes DecodeTimeOnly work for BC inputs, which it never did before. (The historical veracity of all this is nonexistent, of course, but if we're going to say we support proleptic Gregorian calendar then we should do it correctly. In any case the unpatched code is broken because it could emit dates that it would then reject on re-inputting.) Per report from Bernd Helmle. Back-patch as far as 8.0; in 7.x we were not using our own calendar support and so this seems a bit too risky to put into 7.4.
2008-02-26 00:21:01 +01:00
{
Reject year zero during datetime input, except when it's a 2-digit year (then it means 2000 AD). Formerly we silently interpreted this as 1 BC, which at best is unwarranted familiarity with the implementation. It's barely possible that some app somewhere expects the old behavior, though, so we won't back-patch this into existing release branches.
2008-02-26 00:36:28 +01:00
/* there is no year zero in AD/BC notation */
if (tm->tm_year <= 0)
return DTERR_FIELD_OVERFLOW;
/* internally, we represent 1 BC as year zero, 2 BC as -1, etc */
tm->tm_year = -(tm->tm_year - 1);
Fix datetime input to behave correctly for Feb 29 in years BC. Formerly, DecodeDate attempted to verify the day-of-the-month exactly, but it was under the misapprehension that it would know whether we were looking at a BC year or not. In reality this check can't be made until the calling function (eg DecodeDateTime) has processed all the fields. So, split the BC adjustment and validity checks out into a new function ValidateDate that is called only after processing all the fields. In passing, this patch makes DecodeTimeOnly work for BC inputs, which it never did before. (The historical veracity of all this is nonexistent, of course, but if we're going to say we support proleptic Gregorian calendar then we should do it correctly. In any case the unpatched code is broken because it could emit dates that it would then reject on re-inputting.) Per report from Bernd Helmle. Back-patch as far as 8.0; in 7.x we were not using our own calendar support and so this seems a bit too risky to put into 7.4.
2008-02-26 00:21:01 +01:00
}
else if (is2digits)
{
Improve zero-year comments.
2009-03-17 19:39:39 +01:00
/* process 1 or 2-digit input as 1970-2069 AD, allow '0' and '00' */
8.4 pgindent run, with new combined Linux/FreeBSD/MinGW typedef list provided by Andrew.
2009-06-11 16:49:15 +02:00
if (tm->tm_year < 0) /* just paranoia */
Reject year zero during datetime input, except when it's a 2-digit year (then it means 2000 AD). Formerly we silently interpreted this as 1 BC, which at best is unwarranted familiarity with the implementation. It's barely possible that some app somewhere expects the old behavior, though, so we won't back-patch this into existing release branches.
2008-02-26 00:36:28 +01:00
return DTERR_FIELD_OVERFLOW;
Fix datetime input to behave correctly for Feb 29 in years BC. Formerly, DecodeDate attempted to verify the day-of-the-month exactly, but it was under the misapprehension that it would know whether we were looking at a BC year or not. In reality this check can't be made until the calling function (eg DecodeDateTime) has processed all the fields. So, split the BC adjustment and validity checks out into a new function ValidateDate that is called only after processing all the fields. In passing, this patch makes DecodeTimeOnly work for BC inputs, which it never did before. (The historical veracity of all this is nonexistent, of course, but if we're going to say we support proleptic Gregorian calendar then we should do it correctly. In any case the unpatched code is broken because it could emit dates that it would then reject on re-inputting.) Per report from Bernd Helmle. Back-patch as far as 8.0; in 7.x we were not using our own calendar support and so this seems a bit too risky to put into 7.4.
2008-02-26 00:21:01 +01:00
if (tm->tm_year < 70)
tm->tm_year += 2000;
else if (tm->tm_year < 100)
tm->tm_year += 1900;
}
Reject year zero during datetime input, except when it's a 2-digit year (then it means 2000 AD). Formerly we silently interpreted this as 1 BC, which at best is unwarranted familiarity with the implementation. It's barely possible that some app somewhere expects the old behavior, though, so we won't back-patch this into existing release branches.
2008-02-26 00:36:28 +01:00
else
{
/* there is no year zero in AD/BC notation */
if (tm->tm_year <= 0)
return DTERR_FIELD_OVERFLOW;
}
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
}
Apply (a somewhat revised version of) Greg Mullane's patch to eliminate heuristic determination of day vs month in date/time input. Add the ability to specify that input is interpreted as yy-mm-dd order (which formerly worked, but only for yy greater than 31). DateStyle's input component now has the preferred spellings DMY, MDY, or YMD; the older keywords European and US are now aliases for the first two of these. Per recent discussions on pgsql-general.
2003-07-29 02:03:19 +02:00
/* now that we have correct year, decode DOY */
if (fmask & DTK_M(DOY))
{
j2date(date2j(tm->tm_year, 1, 1) + tm->tm_yday - 1,
&tm->tm_year, &tm->tm_mon, &tm->tm_mday);
}
/* check for valid month */
Fix datetime input to behave correctly for Feb 29 in years BC. Formerly, DecodeDate attempted to verify the day-of-the-month exactly, but it was under the misapprehension that it would know whether we were looking at a BC year or not. In reality this check can't be made until the calling function (eg DecodeDateTime) has processed all the fields. So, split the BC adjustment and validity checks out into a new function ValidateDate that is called only after processing all the fields. In passing, this patch makes DecodeTimeOnly work for BC inputs, which it never did before. (The historical veracity of all this is nonexistent, of course, but if we're going to say we support proleptic Gregorian calendar then we should do it correctly. In any case the unpatched code is broken because it could emit dates that it would then reject on re-inputting.) Per report from Bernd Helmle. Back-patch as far as 8.0; in 7.x we were not using our own calendar support and so this seems a bit too risky to put into 7.4.
2008-02-26 00:21:01 +01:00
if (fmask & DTK_M(MONTH))
{
if (tm->tm_mon < 1 || tm->tm_mon > MONTHS_PER_YEAR)
return DTERR_MD_FIELD_OVERFLOW;
}
Apply (a somewhat revised version of) Greg Mullane's patch to eliminate heuristic determination of day vs month in date/time input. Add the ability to specify that input is interpreted as yy-mm-dd order (which formerly worked, but only for yy greater than 31). DateStyle's input component now has the preferred spellings DMY, MDY, or YMD; the older keywords European and US are now aliases for the first two of these. Per recent discussions on pgsql-general.
2003-07-29 02:03:19 +02:00
Fix datetime input to behave correctly for Feb 29 in years BC. Formerly, DecodeDate attempted to verify the day-of-the-month exactly, but it was under the misapprehension that it would know whether we were looking at a BC year or not. In reality this check can't be made until the calling function (eg DecodeDateTime) has processed all the fields. So, split the BC adjustment and validity checks out into a new function ValidateDate that is called only after processing all the fields. In passing, this patch makes DecodeTimeOnly work for BC inputs, which it never did before. (The historical veracity of all this is nonexistent, of course, but if we're going to say we support proleptic Gregorian calendar then we should do it correctly. In any case the unpatched code is broken because it could emit dates that it would then reject on re-inputting.) Per report from Bernd Helmle. Back-patch as far as 8.0; in 7.x we were not using our own calendar support and so this seems a bit too risky to put into 7.4.
2008-02-26 00:21:01 +01:00
/* minimal check for valid day */
if (fmask & DTK_M(DAY))
{
if (tm->tm_mday < 1 || tm->tm_mday > 31)
return DTERR_MD_FIELD_OVERFLOW;
}
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
Fix datetime input to behave correctly for Feb 29 in years BC. Formerly, DecodeDate attempted to verify the day-of-the-month exactly, but it was under the misapprehension that it would know whether we were looking at a BC year or not. In reality this check can't be made until the calling function (eg DecodeDateTime) has processed all the fields. So, split the BC adjustment and validity checks out into a new function ValidateDate that is called only after processing all the fields. In passing, this patch makes DecodeTimeOnly work for BC inputs, which it never did before. (The historical veracity of all this is nonexistent, of course, but if we're going to say we support proleptic Gregorian calendar then we should do it correctly. In any case the unpatched code is broken because it could emit dates that it would then reject on re-inputting.) Per report from Bernd Helmle. Back-patch as far as 8.0; in 7.x we were not using our own calendar support and so this seems a bit too risky to put into 7.4.
2008-02-26 00:21:01 +01:00
if ((fmask & DTK_DATE_M) == DTK_DATE_M)
{
/*
* Check for valid day of month, now that we know for sure the month
* and year. Note we don't use MD_FIELD_OVERFLOW here, since it seems
* unlikely that "Feb 29" is a YMD-order error.
*/
if (tm->tm_mday > day_tab[isleap(tm->tm_year)][tm->tm_mon - 1])
return DTERR_FIELD_OVERFLOW;
}
Apply (a somewhat revised version of) Greg Mullane's patch to eliminate heuristic determination of day vs month in date/time input. Add the ability to specify that input is interpreted as yy-mm-dd order (which formerly worked, but only for yy greater than 31). DateStyle's input component now has the preferred spellings DMY, MDY, or YMD; the older keywords European and US are now aliases for the first two of these. Per recent discussions on pgsql-general.
2003-07-29 02:03:19 +02:00
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
return 0;
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
}
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
/* DecodeTime()
* Decode time string which includes delimiters.
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
* Return 0 if okay, a DTERR code if not.
*
Code cleanup: use "bool" for Boolean variables, rather than "int".
2007-05-27 22:32:16 +02:00
* Only check the lower limit on hours, since this same code can be
* used to represent time spans.
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
*/
Mark functions as static and ifdef NOT_USED as appropriate.
2000-06-09 00:38:00 +02:00
static int
Make our parsing of INTERVAL literals spec-compliant (or at least a heck of a lot closer than it was before). To do this, tweak coerce_type() to pass through the typmod information when invoking interval_in() on an UNKNOWN constant; then fix DecodeInterval to pay attention to the typmod when deciding how to interpret a units-less integer value. I changed one or two other details as well. I believe the code now reacts as expected by spec for all the literal syntaxes that are specifically enumerated in the spec. There are corner cases involving strings that don't exactly match the set of fields called out by the typmod, for which we might want to tweak the behavior some more; but I think this is an area of user friendliness rather than spec compliance. There remain some non-compliant details about the SQL syntax (as opposed to what's inside the literal string); but at least we'll throw error rather than silently doing the wrong thing in those cases.
2008-09-10 20:29:41 +02:00
DecodeTime(char *str, int fmask, int range,
Initial pgindent run with pg_bsd_indent version 2.0. The new indent version includes numerous fixes thanks to Piotr Stefaniak. The main changes visible in this commit are: * Nicer formatting of function-pointer declarations. * No longer unexpectedly removes spaces in expressions using casts, sizeof, or offsetof. * No longer wants to add a space in "struct structname *varname", as well as some similar cases for const- or volatile-qualified pointers. * Declarations using PG_USED_FOR_ASSERTS_ONLY are formatted more nicely. * Fixes bug where comments following declarations were sometimes placed with no space separating them from the code. * Fixes some odd decisions for comments following case labels. * Fixes some cases where comments following code were indented to less than the expected column 33. On the less good side, it now tends to put more whitespace around typedef names that are not listed in typedefs.list. This might encourage us to put more effort into typedef name collection; it's not really a bug in indent itself. There are more changes coming after this round, having to do with comment indentation and alignment of lines appearing within parentheses. I wanted to limit the size of the diffs to something that could be reviewed without one's eyes completely glazing over, so it seemed better to split up the changes as much as practical. Discussion: https://postgr.es/m/E1dAmxK-0006EE-1r@gemulon.postgresql.org Discussion: https://postgr.es/m/30527.1495162840@sss.pgh.pa.us
2017-06-21 20:39:04 +02:00
int *tmask, struct pg_tm *tm, fsec_t *fsec)
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
{
char *cp;
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
int dterr;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
*tmask = DTK_TIME_M;
Remove comment on errno=0 lines, but add mention to port/strtol.c function.
2005-12-02 03:49:11 +01:00
errno = 0;
Rename strtoi() to strtoint(). NetBSD has seen fit to invent a libc function named strtoi(), which conflicts with the long-established static functions of the same name in datetime.c and ecpg's interval.c. While muttering darkly about intrusions on application namespace, we'll rename our functions to avoid the conflict. Back-patch to all supported branches, since this would affect attempts to build any of them on recent NetBSD. Thomas Munro
2016-04-23 22:53:15 +02:00
tm->tm_hour = strtoint(str, &cp, 10);
Check for overflow in strtol() while parsing datetime inputs. Michael Fuhr.
2005-12-01 18:56:34 +01:00
if (errno == ERANGE)
return DTERR_FIELD_OVERFLOW;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
if (*cp != ':')
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
return DTERR_BAD_FORMAT;
Remove comment on errno=0 lines, but add mention to port/strtol.c function.
2005-12-02 03:49:11 +01:00
errno = 0;
Rename strtoi() to strtoint(). NetBSD has seen fit to invent a libc function named strtoi(), which conflicts with the long-established static functions of the same name in datetime.c and ecpg's interval.c. While muttering darkly about intrusions on application namespace, we'll rename our functions to avoid the conflict. Back-patch to all supported branches, since this would affect attempts to build any of them on recent NetBSD. Thomas Munro
2016-04-23 22:53:15 +02:00
tm->tm_min = strtoint(cp + 1, &cp, 10);
Check for overflow in strtol() while parsing datetime inputs. Michael Fuhr.
2005-12-01 18:56:34 +01:00
if (errno == ERANGE)
return DTERR_FIELD_OVERFLOW;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
if (*cp == '\0')
{
tm->tm_sec = 0;
*fsec = 0;
Make our parsing of INTERVAL literals spec-compliant (or at least a heck of a lot closer than it was before). To do this, tweak coerce_type() to pass through the typmod information when invoking interval_in() on an UNKNOWN constant; then fix DecodeInterval to pay attention to the typmod when deciding how to interpret a units-less integer value. I changed one or two other details as well. I believe the code now reacts as expected by spec for all the literal syntaxes that are specifically enumerated in the spec. There are corner cases involving strings that don't exactly match the set of fields called out by the typmod, for which we might want to tweak the behavior some more; but I think this is an area of user friendliness rather than spec compliance. There remain some non-compliant details about the SQL syntax (as opposed to what's inside the literal string); but at least we'll throw error rather than silently doing the wrong thing in those cases.
2008-09-10 20:29:41 +02:00
/* If it's a MINUTE TO SECOND interval, take 2 fields as being mm:ss */
if (range == (INTERVAL_MASK(MINUTE) | INTERVAL_MASK(SECOND)))
{
tm->tm_sec = tm->tm_min;
tm->tm_min = tm->tm_hour;
tm->tm_hour = 0;
}
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
}
Adjust the parser to accept the typename syntax INTERVAL ... SECOND(n) and the literal syntax INTERVAL 'string' ... SECOND(n), as required by the SQL standard. Our old syntax put (n) directly after INTERVAL, which was a mistake, but will still be accepted for backward compatibility as well as symmetry with the TIMESTAMP cases. Change intervaltypmodout to show it in the spec's way, too. (This could potentially affect clients, if there are any that analyze the typmod of an INTERVAL in any detail.) Also fix interval input to handle 'min:sec.frac' properly; I had overlooked this case in my previous patch. Document the use of the interval fields qualifier, which up to now we had never mentioned in the docs. (I think the omission was intentional because it didn't work per spec; but it does now, or at least close enough to be credible.)
2008-09-11 17:27:30 +02:00
else if (*cp == '.')
{
/* always assume mm:ss.sss is MINUTE TO SECOND */
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
dterr = ParseFractionalSecond(cp, fsec);
if (dterr)
return dterr;
Adjust the parser to accept the typename syntax INTERVAL ... SECOND(n) and the literal syntax INTERVAL 'string' ... SECOND(n), as required by the SQL standard. Our old syntax put (n) directly after INTERVAL, which was a mistake, but will still be accepted for backward compatibility as well as symmetry with the TIMESTAMP cases. Change intervaltypmodout to show it in the spec's way, too. (This could potentially affect clients, if there are any that analyze the typmod of an INTERVAL in any detail.) Also fix interval input to handle 'min:sec.frac' properly; I had overlooked this case in my previous patch. Document the use of the interval fields qualifier, which up to now we had never mentioned in the docs. (I think the omission was intentional because it didn't work per spec; but it does now, or at least close enough to be credible.)
2008-09-11 17:27:30 +02:00
tm->tm_sec = tm->tm_min;
tm->tm_min = tm->tm_hour;
tm->tm_hour = 0;
}
else if (*cp == ':')
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
{
Remove comment on errno=0 lines, but add mention to port/strtol.c function.
2005-12-02 03:49:11 +01:00
errno = 0;
Rename strtoi() to strtoint(). NetBSD has seen fit to invent a libc function named strtoi(), which conflicts with the long-established static functions of the same name in datetime.c and ecpg's interval.c. While muttering darkly about intrusions on application namespace, we'll rename our functions to avoid the conflict. Back-patch to all supported branches, since this would affect attempts to build any of them on recent NetBSD. Thomas Munro
2016-04-23 22:53:15 +02:00
tm->tm_sec = strtoint(cp + 1, &cp, 10);
Check for overflow in strtol() while parsing datetime inputs. Michael Fuhr.
2005-12-01 18:56:34 +01:00
if (errno == ERANGE)
return DTERR_FIELD_OVERFLOW;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
if (*cp == '\0')
*fsec = 0;
else if (*cp == '.')
{
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
dterr = ParseFractionalSecond(cp, fsec);
if (dterr)
return dterr;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
}
else
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
return DTERR_BAD_FORMAT;
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
}
Adjust the parser to accept the typename syntax INTERVAL ... SECOND(n) and the literal syntax INTERVAL 'string' ... SECOND(n), as required by the SQL standard. Our old syntax put (n) directly after INTERVAL, which was a mistake, but will still be accepted for backward compatibility as well as symmetry with the TIMESTAMP cases. Change intervaltypmodout to show it in the spec's way, too. (This could potentially affect clients, if there are any that analyze the typmod of an INTERVAL in any detail.) Also fix interval input to handle 'min:sec.frac' properly; I had overlooked this case in my previous patch. Document the use of the interval fields qualifier, which up to now we had never mentioned in the docs. (I think the omission was intentional because it didn't work per spec; but it does now, or at least close enough to be credible.)
2008-09-11 17:27:30 +02:00
else
return DTERR_BAD_FORMAT;
Fix datetime and abstime conversions to and from date. Bring optional new-storage date and time up to date and test. This new storage format should fix the "Sparc gcc -O2 bug". (Enable new code with USE_NEW_DATE and USE_NEW_TIME in dt.h)
1997-05-16 09:19:50 +02:00
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
/* do a sanity check */
pgindent run before PG 9.1 beta 1.
2011-04-10 17:42:00 +02:00
if (tm->tm_hour < 0 || tm->tm_min < 0 || tm->tm_min > MINS_PER_HOUR - 1 ||
Use macros for time-based constants, rather than constants.
2011-03-12 15:31:18 +01:00
tm->tm_sec < 0 || tm->tm_sec > SECS_PER_MINUTE ||
*fsec < INT64CONST(0) ||
When checking for datetime field overflow, we should allow a fractional-second part that rounds up to exactly 1.0 second. The previous coding rejected input like "00:12:57.9999999999999999999999999999", with the exact number of nines needed to cause failure varying depending on float-timestamp option and possibly on platform. Obviously this should round up to the next integral second, if we don't have enough precision to distinguish the value from that. Per bug #4789 from Robert Kruus. In passing, fix a missed check for fractional seconds in one copy of the "is it greater than 24:00:00" code. Broken all the way back, so patch all the way back.
2009-05-01 21:29:07 +02:00
*fsec > USECS_PER_SEC)
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
return DTERR_FIELD_OVERFLOW;
From: Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> Subject: [HACKERS] More date time functions Here are some additional patches mostly related to the date and time data types. It includes some type conversion routines to move between the different date types and some other date manipulation routines such as date_part(units,datetime). I noticed Edmund Mergl et al's neat trick for getting function overloading for builtin functions, so started to use that for the date and time stuff. Later, if someone figures out how to get function overloading directly for internal C code, then we can move to that technique. These patches include documentation updates (don't faint!) for the built-in man page. Doesn't yet include mention of timestamp, since I don't know much about it and since it may change a bit to become a _real_ ANSI timestamp which would include parser support for the declaration syntax (what do you think, Dan?). The patches were developed on the 970330 release, but have been rebuilt off of the 970402 release. The first patch below is to get libpq to compile, on my Linux box, but is not related to the rest of the patches and you can choose not to apply that one at this time. Thanks in advance, scrappy!
1997-04-02 20:36:24 +02:00
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
return 0;
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
}
From: Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> Subject: [HACKERS] More date time functions Here are some additional patches mostly related to the date and time data types. It includes some type conversion routines to move between the different date types and some other date manipulation routines such as date_part(units,datetime). I noticed Edmund Mergl et al's neat trick for getting function overloading for builtin functions, so started to use that for the date and time stuff. Later, if someone figures out how to get function overloading directly for internal C code, then we can move to that technique. These patches include documentation updates (don't faint!) for the built-in man page. Doesn't yet include mention of timestamp, since I don't know much about it and since it may change a bit to become a _real_ ANSI timestamp which would include parser support for the declaration syntax (what do you think, Dan?). The patches were developed on the 970330 release, but have been rebuilt off of the 970402 release. The first patch below is to get libpq to compile, on my Linux box, but is not related to the rest of the patches and you can choose not to apply that one at this time. Thanks in advance, scrappy!
1997-04-02 20:36:24 +02:00
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
/* DecodeNumber()
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
* Interpret plain numeric field as a date value in context.
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
* Return 0 if okay, a DTERR code if not.
From: Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> Subject: [HACKERS] More date time functions Here are some additional patches mostly related to the date and time data types. It includes some type conversion routines to move between the different date types and some other date manipulation routines such as date_part(units,datetime). I noticed Edmund Mergl et al's neat trick for getting function overloading for builtin functions, so started to use that for the date and time stuff. Later, if someone figures out how to get function overloading directly for internal C code, then we can move to that technique. These patches include documentation updates (don't faint!) for the built-in man page. Doesn't yet include mention of timestamp, since I don't know much about it and since it may change a bit to become a _real_ ANSI timestamp which would include parser support for the declaration syntax (what do you think, Dan?). The patches were developed on the 970330 release, but have been rebuilt off of the 970402 release. The first patch below is to get libpq to compile, on my Linux box, but is not related to the rest of the patches and you can choose not to apply that one at this time. Thanks in advance, scrappy!
1997-04-02 20:36:24 +02:00
*/
Mark functions as static and ifdef NOT_USED as appropriate.
2000-06-09 00:38:00 +02:00
static int
Fix datetime input parsing to accept YYYY-MONTHNAME-DD and related syntaxes, which had been unintentionally broken by recent changes to tighten up the DateStyle rules for all-numeric date input. Add documentation and regression tests for this, too.
2003-11-16 21:29:16 +01:00
DecodeNumber(int flen, char *str, bool haveTextMonth, int fmask,
Initial pgindent run with pg_bsd_indent version 2.0. The new indent version includes numerous fixes thanks to Piotr Stefaniak. The main changes visible in this commit are: * Nicer formatting of function-pointer declarations. * No longer unexpectedly removes spaces in expressions using casts, sizeof, or offsetof. * No longer wants to add a space in "struct structname *varname", as well as some similar cases for const- or volatile-qualified pointers. * Declarations using PG_USED_FOR_ASSERTS_ONLY are formatted more nicely. * Fixes bug where comments following declarations were sometimes placed with no space separating them from the code. * Fixes some odd decisions for comments following case labels. * Fixes some cases where comments following code were indented to less than the expected column 33. On the less good side, it now tends to put more whitespace around typedef names that are not listed in typedefs.list. This might encourage us to put more effort into typedef name collection; it's not really a bug in indent itself. There are more changes coming after this round, having to do with comment indentation and alignment of lines appearing within parentheses. I wanted to limit the size of the diffs to something that could be reviewed without one's eyes completely glazing over, so it seemed better to split up the changes as much as practical. Discussion: https://postgr.es/m/E1dAmxK-0006EE-1r@gemulon.postgresql.org Discussion: https://postgr.es/m/30527.1495162840@sss.pgh.pa.us
2017-06-21 20:39:04 +02:00
int *tmask, struct pg_tm *tm, fsec_t *fsec, bool *is2digits)
From: Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> Subject: [HACKERS] More date time functions Here are some additional patches mostly related to the date and time data types. It includes some type conversion routines to move between the different date types and some other date manipulation routines such as date_part(units,datetime). I noticed Edmund Mergl et al's neat trick for getting function overloading for builtin functions, so started to use that for the date and time stuff. Later, if someone figures out how to get function overloading directly for internal C code, then we can move to that technique. These patches include documentation updates (don't faint!) for the built-in man page. Doesn't yet include mention of timestamp, since I don't know much about it and since it may change a bit to become a _real_ ANSI timestamp which would include parser support for the declaration syntax (what do you think, Dan?). The patches were developed on the 970330 release, but have been rebuilt off of the 970402 release. The first patch below is to get libpq to compile, on my Linux box, but is not related to the rest of the patches and you can choose not to apply that one at this time. Thanks in advance, scrappy!
1997-04-02 20:36:24 +02:00
{
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
int val;
char *cp;
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
int dterr;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
*tmask = 0;
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
Remove comment on errno=0 lines, but add mention to port/strtol.c function.
2005-12-02 03:49:11 +01:00
errno = 0;
Rename strtoi() to strtoint(). NetBSD has seen fit to invent a libc function named strtoi(), which conflicts with the long-established static functions of the same name in datetime.c and ecpg's interval.c. While muttering darkly about intrusions on application namespace, we'll rename our functions to avoid the conflict. Back-patch to all supported branches, since this would affect attempts to build any of them on recent NetBSD. Thomas Munro
2016-04-23 22:53:15 +02:00
val = strtoint(str, &cp, 10);
Check for overflow in strtol() while parsing datetime inputs. Michael Fuhr.
2005-12-01 18:56:34 +01:00
if (errno == ERANGE)
return DTERR_FIELD_OVERFLOW;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
if (cp == str)
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
return DTERR_BAD_FORMAT;
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
if (*cp == '.')
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
{
pgindent run.
2002-09-04 22:31:48 +02:00
/*
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* More than two digits before decimal point? Then could be a date or
* a run-together time: 2001.360 20011225 040506.789
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
*/
Remove unnecessary parentheses in datetime/timestamp code.
2005-05-23 19:13:14 +02:00
if (cp - str > 2)
Allow parsing of time and timetz inputs to accept the documented input syntax '040506' for '04:05:06', as well as '0405' for '04:05:00'. This has been broken since 7.2 but was only recently complained of.
2003-08-26 00:47:34 +02:00
{
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
dterr = DecodeNumberField(flen, str,
(fmask | DTK_DATE_M),
tmask, tm,
fsec, is2digits);
if (dterr < 0)
return dterr;
Allow parsing of time and timetz inputs to accept the documented input syntax '040506' for '04:05:06', as well as '0405' for '04:05:00'. This has been broken since 7.2 but was only recently complained of.
2003-08-26 00:47:34 +02:00
return 0;
}
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
dterr = ParseFractionalSecond(cp, fsec);
if (dterr)
return dterr;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
}
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
else if (*cp != '\0')
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
return DTERR_BAD_FORMAT;
Another PGINDENT run that changes variable indenting and case label indenting. Also static variable indenting.
1997-09-08 04:41:22 +02:00
Apply (a somewhat revised version of) Greg Mullane's patch to eliminate heuristic determination of day vs month in date/time input. Add the ability to specify that input is interpreted as yy-mm-dd order (which formerly worked, but only for yy greater than 31). DateStyle's input component now has the preferred spellings DMY, MDY, or YMD; the older keywords European and US are now aliases for the first two of these. Per recent discussions on pgsql-general.
2003-07-29 02:03:19 +02:00
/* Special case for day of year */
Remove unnecessary parentheses in datetime/timestamp code.
2005-05-23 19:13:14 +02:00
if (flen == 3 && (fmask & DTK_DATE_M) == DTK_M(YEAR) && val >= 1 &&
val <= 366)
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
{
*tmask = (DTK_M(DOY) | DTK_M(MONTH) | DTK_M(DAY));
tm->tm_yday = val;
Apply (a somewhat revised version of) Greg Mullane's patch to eliminate heuristic determination of day vs month in date/time input. Add the ability to specify that input is interpreted as yy-mm-dd order (which formerly worked, but only for yy greater than 31). DateStyle's input component now has the preferred spellings DMY, MDY, or YMD; the older keywords European and US are now aliases for the first two of these. Per recent discussions on pgsql-general.
2003-07-29 02:03:19 +02:00
/* tm_mon and tm_mday can't actually be set yet ... */
return 0;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
}
Another PGINDENT run that changes variable indenting and case label indenting. Also static variable indenting.
1997-09-08 04:41:22 +02:00
Apply (a somewhat revised version of) Greg Mullane's patch to eliminate heuristic determination of day vs month in date/time input. Add the ability to specify that input is interpreted as yy-mm-dd order (which formerly worked, but only for yy greater than 31). DateStyle's input component now has the preferred spellings DMY, MDY, or YMD; the older keywords European and US are now aliases for the first two of these. Per recent discussions on pgsql-general.
2003-07-29 02:03:19 +02:00
/* Switch based on what we have so far */
switch (fmask & DTK_DATE_M)
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
{
Apply (a somewhat revised version of) Greg Mullane's patch to eliminate heuristic determination of day vs month in date/time input. Add the ability to specify that input is interpreted as yy-mm-dd order (which formerly worked, but only for yy greater than 31). DateStyle's input component now has the preferred spellings DMY, MDY, or YMD; the older keywords European and US are now aliases for the first two of these. Per recent discussions on pgsql-general.
2003-07-29 02:03:19 +02:00
case 0:
pgindent run.
2003-08-04 02:43:34 +02:00
Apply (a somewhat revised version of) Greg Mullane's patch to eliminate heuristic determination of day vs month in date/time input. Add the ability to specify that input is interpreted as yy-mm-dd order (which formerly worked, but only for yy greater than 31). DateStyle's input component now has the preferred spellings DMY, MDY, or YMD; the older keywords European and US are now aliases for the first two of these. Per recent discussions on pgsql-general.
2003-07-29 02:03:19 +02:00
/*
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* Nothing so far; make a decision about what we think the input
pgindent run for 9.4 This includes removing tabs after periods in C comments, which was applied to back branches, so this change should not effect backpatching.
2014-05-06 18:12:18 +02:00
* is. There used to be lots of heuristics here, but the
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* consensus now is to be paranoid. It *must* be either
Apply (a somewhat revised version of) Greg Mullane's patch to eliminate heuristic determination of day vs month in date/time input. Add the ability to specify that input is interpreted as yy-mm-dd order (which formerly worked, but only for yy greater than 31). DateStyle's input component now has the preferred spellings DMY, MDY, or YMD; the older keywords European and US are now aliases for the first two of these. Per recent discussions on pgsql-general.
2003-07-29 02:03:19 +02:00
* YYYY-MM-DD (with a more-than-two-digit year field), or the
* field order defined by DateOrder.
*/
if (flen >= 3 || DateOrder == DATEORDER_YMD)
{
*tmask = DTK_M(YEAR);
tm->tm_year = val;
}
else if (DateOrder == DATEORDER_DMY)
{
*tmask = DTK_M(DAY);
tm->tm_mday = val;
}
else
{
*tmask = DTK_M(MONTH);
tm->tm_mon = val;
}
break;
Another PGINDENT run that changes variable indenting and case label indenting. Also static variable indenting.
1997-09-08 04:41:22 +02:00
Apply (a somewhat revised version of) Greg Mullane's patch to eliminate heuristic determination of day vs month in date/time input. Add the ability to specify that input is interpreted as yy-mm-dd order (which formerly worked, but only for yy greater than 31). DateStyle's input component now has the preferred spellings DMY, MDY, or YMD; the older keywords European and US are now aliases for the first two of these. Per recent discussions on pgsql-general.
2003-07-29 02:03:19 +02:00
case (DTK_M(YEAR)):
/* Must be at second field of YY-MM-DD */
*tmask = DTK_M(MONTH);
tm->tm_mon = val;
break;
Fix datetime input parsing to accept YYYY-MONTHNAME-DD and related syntaxes, which had been unintentionally broken by recent changes to tighten up the DateStyle rules for all-numeric date input. Add documentation and regression tests for this, too.
2003-11-16 21:29:16 +01:00
case (DTK_M(MONTH)):
if (haveTextMonth)
{
/*
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* We are at the first numeric field of a date that included a
pgindent run for 9.4 This includes removing tabs after periods in C comments, which was applied to back branches, so this change should not effect backpatching.
2014-05-06 18:12:18 +02:00
* textual month name. We want to support the variants
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* MON-DD-YYYY, DD-MON-YYYY, and YYYY-MON-DD as unambiguous
pgindent run for 9.4 This includes removing tabs after periods in C comments, which was applied to back branches, so this change should not effect backpatching.
2014-05-06 18:12:18 +02:00
* inputs. We will also accept MON-DD-YY or DD-MON-YY in
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* either DMY or MDY modes, as well as YY-MON-DD in YMD mode.
Fix datetime input parsing to accept YYYY-MONTHNAME-DD and related syntaxes, which had been unintentionally broken by recent changes to tighten up the DateStyle rules for all-numeric date input. Add documentation and regression tests for this, too.
2003-11-16 21:29:16 +01:00
*/
if (flen >= 3 || DateOrder == DATEORDER_YMD)
{
*tmask = DTK_M(YEAR);
tm->tm_year = val;
}
else
{
*tmask = DTK_M(DAY);
tm->tm_mday = val;
}
}
else
{
/* Must be at second field of MM-DD-YY */
*tmask = DTK_M(DAY);
tm->tm_mday = val;
}
break;
Apply (a somewhat revised version of) Greg Mullane's patch to eliminate heuristic determination of day vs month in date/time input. Add the ability to specify that input is interpreted as yy-mm-dd order (which formerly worked, but only for yy greater than 31). DateStyle's input component now has the preferred spellings DMY, MDY, or YMD; the older keywords European and US are now aliases for the first two of these. Per recent discussions on pgsql-general.
2003-07-29 02:03:19 +02:00
case (DTK_M(YEAR) | DTK_M(MONTH)):
Fix datetime input parsing to accept YYYY-MONTHNAME-DD and related syntaxes, which had been unintentionally broken by recent changes to tighten up the DateStyle rules for all-numeric date input. Add documentation and regression tests for this, too.
2003-11-16 21:29:16 +01:00
if (haveTextMonth)
{
/* Need to accept DD-MON-YYYY even in YMD mode */
if (flen >= 3 && *is2digits)
{
/* Guess that first numeric field is day was wrong */
Phase 2 of pgindent updates. Change pg_bsd_indent to follow upstream rules for placement of comments to the right of code, and remove pgindent hack that caused comments following #endif to not obey the general rule. Commit e3860ffa4dd0dad0dd9eea4be9cc1412373a8c89 wasn't actually using the published version of pg_bsd_indent, but a hacked-up version that tried to minimize the amount of movement of comments to the right of code. The situation of interest is where such a comment has to be moved to the right of its default placement at column 33 because there's code there. BSD indent has always moved right in units of tab stops in such cases --- but in the previous incarnation, indent was working in 8-space tab stops, while now it knows we use 4-space tabs. So the net result is that in about half the cases, such comments are placed one tab stop left of before. This is better all around: it leaves more room on the line for comment text, and it means that in such cases the comment uniformly starts at the next 4-space tab stop after the code, rather than sometimes one and sometimes two tabs after. Also, ensure that comments following #endif are indented the same as comments following other preprocessor commands such as #else. That inconsistency turns out to have been self-inflicted damage from a poorly-thought-through post-indent "fixup" in pgindent. This patch is much less interesting than the first round of indent changes, but also bulkier, so I thought it best to separate the effects. Discussion: https://postgr.es/m/E1dAmxK-0006EE-1r@gemulon.postgresql.org Discussion: https://postgr.es/m/30527.1495162840@sss.pgh.pa.us
2017-06-21 21:18:54 +02:00
*tmask = DTK_M(DAY); /* YEAR is already set */
Fix datetime input parsing to accept YYYY-MONTHNAME-DD and related syntaxes, which had been unintentionally broken by recent changes to tighten up the DateStyle rules for all-numeric date input. Add documentation and regression tests for this, too.
2003-11-16 21:29:16 +01:00
tm->tm_mday = tm->tm_year;
tm->tm_year = val;
Change TRUE/FALSE to true/false The lower case spellings are C and C++ standard and are used in most parts of the PostgreSQL sources. The upper case spellings are only used in some files/modules. So standardize on the standard spellings. The APIs for ICU, Perl, and Windows define their own TRUE and FALSE, so those are left as is when using those APIs. In code comments, we use the lower-case spelling for the C concepts and keep the upper-case spelling for the SQL concepts. Reviewed-by: Michael Paquier <michael.paquier@gmail.com>
2017-08-16 06:22:32 +02:00
*is2digits = false;
Fix datetime input parsing to accept YYYY-MONTHNAME-DD and related syntaxes, which had been unintentionally broken by recent changes to tighten up the DateStyle rules for all-numeric date input. Add documentation and regression tests for this, too.
2003-11-16 21:29:16 +01:00
}
else
{
*tmask = DTK_M(DAY);
tm->tm_mday = val;
}
}
else
{
/* Must be at third field of YY-MM-DD */
*tmask = DTK_M(DAY);
tm->tm_mday = val;
}
Apply (a somewhat revised version of) Greg Mullane's patch to eliminate heuristic determination of day vs month in date/time input. Add the ability to specify that input is interpreted as yy-mm-dd order (which formerly worked, but only for yy greater than 31). DateStyle's input component now has the preferred spellings DMY, MDY, or YMD; the older keywords European and US are now aliases for the first two of these. Per recent discussions on pgsql-general.
2003-07-29 02:03:19 +02:00
break;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
Apply (a somewhat revised version of) Greg Mullane's patch to eliminate heuristic determination of day vs month in date/time input. Add the ability to specify that input is interpreted as yy-mm-dd order (which formerly worked, but only for yy greater than 31). DateStyle's input component now has the preferred spellings DMY, MDY, or YMD; the older keywords European and US are now aliases for the first two of these. Per recent discussions on pgsql-general.
2003-07-29 02:03:19 +02:00
case (DTK_M(DAY)):
/* Must be at second field of DD-MM-YY */
*tmask = DTK_M(MONTH);
tm->tm_mon = val;
break;
Change rules for interpreting date/time input to disallow 1 and 3 character years. Rejects dates like '0.085', which were accepted previously.
2000-03-29 05:57:18 +02:00
Apply (a somewhat revised version of) Greg Mullane's patch to eliminate heuristic determination of day vs month in date/time input. Add the ability to specify that input is interpreted as yy-mm-dd order (which formerly worked, but only for yy greater than 31). DateStyle's input component now has the preferred spellings DMY, MDY, or YMD; the older keywords European and US are now aliases for the first two of these. Per recent discussions on pgsql-general.
2003-07-29 02:03:19 +02:00
case (DTK_M(MONTH) | DTK_M(DAY)):
/* Must be at third field of DD-MM-YY or MM-DD-YY */
*tmask = DTK_M(YEAR);
tm->tm_year = val;
break;
Allow parsing of time and timetz inputs to accept the documented input syntax '040506' for '04:05:06', as well as '0405' for '04:05:00'. This has been broken since 7.2 but was only recently complained of.
2003-08-26 00:47:34 +02:00
case (DTK_M(YEAR) | DTK_M(MONTH) | DTK_M(DAY)):
/* we have all the date, so it must be a time field */
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
dterr = DecodeNumberField(flen, str, fmask,
tmask, tm,
fsec, is2digits);
if (dterr < 0)
return dterr;
Allow parsing of time and timetz inputs to accept the documented input syntax '040506' for '04:05:06', as well as '0405' for '04:05:00'. This has been broken since 7.2 but was only recently complained of.
2003-08-26 00:47:34 +02:00
return 0;
Apply (a somewhat revised version of) Greg Mullane's patch to eliminate heuristic determination of day vs month in date/time input. Add the ability to specify that input is interpreted as yy-mm-dd order (which formerly worked, but only for yy greater than 31). DateStyle's input component now has the preferred spellings DMY, MDY, or YMD; the older keywords European and US are now aliases for the first two of these. Per recent discussions on pgsql-general.
2003-07-29 02:03:19 +02:00
default:
/* Anything else is bogus input */
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
return DTERR_BAD_FORMAT;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
}
Ye-old pgindent run. Same 4-space tabs.
2000-04-12 19:17:23 +02:00
/*
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* When processing a year field, mark it for adjustment if it's only one
* or two digits.
Change rules for interpreting date/time input to disallow 1 and 3 character years. Rejects dates like '0.085', which were accepted previously.
2000-03-29 05:57:18 +02:00
*/
Apply (a somewhat revised version of) Greg Mullane's patch to eliminate heuristic determination of day vs month in date/time input. Add the ability to specify that input is interpreted as yy-mm-dd order (which formerly worked, but only for yy greater than 31). DateStyle's input component now has the preferred spellings DMY, MDY, or YMD; the older keywords European and US are now aliases for the first two of these. Per recent discussions on pgsql-general.
2003-07-29 02:03:19 +02:00
if (*tmask == DTK_M(YEAR))
Fix datetime input parsing to accept YYYY-MONTHNAME-DD and related syntaxes, which had been unintentionally broken by recent changes to tighten up the DateStyle rules for all-numeric date input. Add documentation and regression tests for this, too.
2003-11-16 21:29:16 +01:00
*is2digits = (flen <= 2);
From: Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> Subject: [HACKERS] More date time functions Here are some additional patches mostly related to the date and time data types. It includes some type conversion routines to move between the different date types and some other date manipulation routines such as date_part(units,datetime). I noticed Edmund Mergl et al's neat trick for getting function overloading for builtin functions, so started to use that for the date and time stuff. Later, if someone figures out how to get function overloading directly for internal C code, then we can move to that technique. These patches include documentation updates (don't faint!) for the built-in man page. Doesn't yet include mention of timestamp, since I don't know much about it and since it may change a bit to become a _real_ ANSI timestamp which would include parser support for the declaration syntax (what do you think, Dan?). The patches were developed on the 970330 release, but have been rebuilt off of the 970402 release. The first patch below is to get libpq to compile, on my Linux box, but is not related to the rest of the patches and you can choose not to apply that one at this time. Thanks in advance, scrappy!
1997-04-02 20:36:24 +02:00
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
return 0;
Apply (a somewhat revised version of) Greg Mullane's patch to eliminate heuristic determination of day vs month in date/time input. Add the ability to specify that input is interpreted as yy-mm-dd order (which formerly worked, but only for yy greater than 31). DateStyle's input component now has the preferred spellings DMY, MDY, or YMD; the older keywords European and US are now aliases for the first two of these. Per recent discussions on pgsql-general.
2003-07-29 02:03:19 +02:00
}
From: Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> Subject: [HACKERS] More date time functions Here are some additional patches mostly related to the date and time data types. It includes some type conversion routines to move between the different date types and some other date manipulation routines such as date_part(units,datetime). I noticed Edmund Mergl et al's neat trick for getting function overloading for builtin functions, so started to use that for the date and time stuff. Later, if someone figures out how to get function overloading directly for internal C code, then we can move to that technique. These patches include documentation updates (don't faint!) for the built-in man page. Doesn't yet include mention of timestamp, since I don't know much about it and since it may change a bit to become a _real_ ANSI timestamp which would include parser support for the declaration syntax (what do you think, Dan?). The patches were developed on the 970330 release, but have been rebuilt off of the 970402 release. The first patch below is to get libpq to compile, on my Linux box, but is not related to the rest of the patches and you can choose not to apply that one at this time. Thanks in advance, scrappy!
1997-04-02 20:36:24 +02:00
Fix datetime and abstime conversions to and from date. Bring optional new-storage date and time up to date and test. This new storage format should fix the "Sparc gcc -O2 bug". (Enable new code with USE_NEW_DATE and USE_NEW_TIME in dt.h)
1997-05-16 09:19:50 +02:00
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
/* DecodeNumberField()
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
* Interpret numeric string as a concatenated date or time field.
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
* Return a DTK token (>= 0) if successful, a DTERR code (< 0) if not.
*
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
* Use the context of previously decoded fields to help with
* the interpretation.
Fix datetime and abstime conversions to and from date. Bring optional new-storage date and time up to date and test. This new storage format should fix the "Sparc gcc -O2 bug". (Enable new code with USE_NEW_DATE and USE_NEW_TIME in dt.h)
1997-05-16 09:19:50 +02:00
*/
Mark functions as static and ifdef NOT_USED as appropriate.
2000-06-09 00:38:00 +02:00
static int
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
DecodeNumberField(int len, char *str, int fmask,
Initial pgindent run with pg_bsd_indent version 2.0. The new indent version includes numerous fixes thanks to Piotr Stefaniak. The main changes visible in this commit are: * Nicer formatting of function-pointer declarations. * No longer unexpectedly removes spaces in expressions using casts, sizeof, or offsetof. * No longer wants to add a space in "struct structname *varname", as well as some similar cases for const- or volatile-qualified pointers. * Declarations using PG_USED_FOR_ASSERTS_ONLY are formatted more nicely. * Fixes bug where comments following declarations were sometimes placed with no space separating them from the code. * Fixes some odd decisions for comments following case labels. * Fixes some cases where comments following code were indented to less than the expected column 33. On the less good side, it now tends to put more whitespace around typedef names that are not listed in typedefs.list. This might encourage us to put more effort into typedef name collection; it's not really a bug in indent itself. There are more changes coming after this round, having to do with comment indentation and alignment of lines appearing within parentheses. I wanted to limit the size of the diffs to something that could be reviewed without one's eyes completely glazing over, so it seemed better to split up the changes as much as practical. Discussion: https://postgr.es/m/E1dAmxK-0006EE-1r@gemulon.postgresql.org Discussion: https://postgr.es/m/30527.1495162840@sss.pgh.pa.us
2017-06-21 20:39:04 +02:00
int *tmask, struct pg_tm *tm, fsec_t *fsec, bool *is2digits)
Fix datetime and abstime conversions to and from date. Bring optional new-storage date and time up to date and test. This new storage format should fix the "Sparc gcc -O2 bug". (Enable new code with USE_NEW_DATE and USE_NEW_TIME in dt.h)
1997-05-16 09:19:50 +02:00
{
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
char *cp;
Fix datetime and abstime conversions to and from date. Bring optional new-storage date and time up to date and test. This new storage format should fix the "Sparc gcc -O2 bug". (Enable new code with USE_NEW_DATE and USE_NEW_TIME in dt.h)
1997-05-16 09:19:50 +02:00
pgindent run.
2002-09-04 22:31:48 +02:00
/*
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* Have a decimal point? Then this is a date or something with a seconds
* field...
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
*/
if ((cp = strchr(str, '.')) != NULL)
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
{
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
/*
8.4 pgindent run, with new combined Linux/FreeBSD/MinGW typedef list provided by Andrew.
2009-06-11 16:49:15 +02:00
* Can we use ParseFractionalSecond here? Not clear whether trailing
* junk should be rejected ...
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
*/
Fix several places where fractional-second inputs were misprocessed in HAVE_INT64_TIMESTAMP cases, including two potential stack smashes when more than six fractional digits were supplied. Per bug report from Philipp Reisner.
2003-08-05 19:39:19 +02:00
double frac;
Support alternate storage scheme of 64-bit integer for date/time types. Use "--enable-integer-datetimes" in configuration to use this rather than the original float8 storage. I would recommend the integer-based storage for any platform on which it is available. We perhaps should make this the default for the production release. Change timezone(timestamptz) results to return timestamp rather than a character string. Formerly, we didn't have a way to represent timestamps with an explicit time zone other than freezing the info into a string. Now, we can reasonably omit the explicit time zone from the result and return a timestamp with values appropriate for the specified time zone. Much cleaner, and if you need the time zone in the result you can put it into a character string pretty easily anyway. Allow fractional seconds in date/time types even for dates prior to 1BC. Limit timestamp data types to 6 decimal places of precision. Just right for a micro-second storage of int8 date/time types, and reduces the number of places ad-hoc rounding was occuring for the float8-based types. Use lookup tables for precision/rounding calculations for timestamp and interval types. Formerly used pow() to calculate the desired value but with a more limited range there is no reason to not type in a lookup table. Should be *much* better performance, though formerly there were some optimizations to help minimize the number of times pow() was called. Define a HAVE_INT64_TIMESTAMP variable. Based on the configure option "--enable-integer-datetimes" and the existing internal INT64_IS_BUSTED. Add explicit date/interval operators and functions for addition and subtraction. Formerly relied on implicit type promotion from date to timestamp with time zone. Change timezone conversion functions for the timetz type from "timetz()" to "timezone()". This is consistant with other time zone coersion functions for other types. Bump the catalog version to 200204201. Fix up regression tests to reflect changes in fractional seconds representation for date/times in BC eras. All regression tests pass on my Linux box.
2002-04-21 21:52:18 +02:00
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
errno = 0;
Fix several places where fractional-second inputs were misprocessed in HAVE_INT64_TIMESTAMP cases, including two potential stack smashes when more than six fractional digits were supplied. Per bug report from Philipp Reisner.
2003-08-05 19:39:19 +02:00
frac = strtod(cp, NULL);
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
if (errno != 0)
return DTERR_BAD_FORMAT;
Fix several places where fractional-second inputs were misprocessed in HAVE_INT64_TIMESTAMP cases, including two potential stack smashes when more than six fractional digits were supplied. Per bug report from Philipp Reisner.
2003-08-05 19:39:19 +02:00
*fsec = rint(frac * 1000000);
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
/* Now truncate off the fraction for further processing */
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
*cp = '\0';
len = strlen(str);
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
}
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
/* No decimal point and no complete date yet? */
else if ((fmask & DTK_DATE_M) != DTK_DATE_M)
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
{
Allow 5+ digit years for non-ISO timestamp/date strings, where appropriate Report from Haribabu Kommi
2013-10-16 19:22:55 +02:00
if (len >= 6)
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
{
*tmask = DTK_DATE_M;
pgindent run for 9.4 This includes removing tabs after periods in C comments, which was applied to back branches, so this change should not effect backpatching.
2014-05-06 18:12:18 +02:00
Allow 5+ digit years for non-ISO timestamp/date strings, where appropriate Report from Haribabu Kommi
2013-10-16 19:22:55 +02:00
/*
* Start from end and consider first 2 as Day, next 2 as Month,
* and the rest as Year.
*/
tm->tm_mday = atoi(str + (len - 2));
*(str + (len - 2)) = '\0';
tm->tm_mon = atoi(str + (len - 4));
*(str + (len - 4)) = '\0';
tm->tm_year = atoi(str);
if ((len - 4) == 2)
Change TRUE/FALSE to true/false The lower case spellings are C and C++ standard and are used in most parts of the PostgreSQL sources. The upper case spellings are only used in some files/modules. So standardize on the standard spellings. The APIs for ICU, Perl, and Windows define their own TRUE and FALSE, so those are left as is when using those APIs. In code comments, we use the lower-case spelling for the C concepts and keep the upper-case spelling for the SQL concepts. Reviewed-by: Michael Paquier <michael.paquier@gmail.com>
2017-08-16 06:22:32 +02:00
*is2digits = true;
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
return DTK_DATE;
}
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
}
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
/* not all time fields are specified? */
if ((fmask & DTK_TIME_M) != DTK_TIME_M)
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
{
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
/* hhmmss */
if (len == 6)
{
*tmask = DTK_TIME_M;
tm->tm_sec = atoi(str + 4);
*(str + 4) = '\0';
tm->tm_min = atoi(str + 2);
*(str + 2) = '\0';
Allow 5+ digit years for non-ISO timestamp/date strings, where appropriate Report from Haribabu Kommi
2013-10-16 19:22:55 +02:00
tm->tm_hour = atoi(str);
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
return DTK_TIME;
}
/* hhmm? */
else if (len == 4)
{
*tmask = DTK_TIME_M;
tm->tm_sec = 0;
tm->tm_min = atoi(str + 2);
*(str + 2) = '\0';
Allow 5+ digit years for non-ISO timestamp/date strings, where appropriate Report from Haribabu Kommi
2013-10-16 19:22:55 +02:00
tm->tm_hour = atoi(str);
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
return DTK_TIME;
}
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
}
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
return DTERR_BAD_FORMAT;
}
Fix datetime and abstime conversions to and from date. Bring optional new-storage date and time up to date and test. This new storage format should fix the "Sparc gcc -O2 bug". (Enable new code with USE_NEW_DATE and USE_NEW_TIME in dt.h)
1997-05-16 09:19:50 +02:00
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
/* DecodeTimezone()
* Interpret string as a numeric timezone.
Allow +1300 as a numeric timezone specifier; we already accept FJST as meaning +1300.
2002-11-13 18:24:05 +01:00
*
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
* Return 0 if okay (and set *tzp), a DTERR code if not okay.
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
*/
Constructors for interval, timestamp, timestamptz Author: Pavel Stěhule, editorialized somewhat by Álvaro Herrera Reviewed-by: Tomáš Vondra, Marko Tiikkaja With input from Fabrízio de Royes Mello, Jim Nasby
2014-03-04 19:09:43 +01:00
int
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
DecodeTimezone(char *str, int *tzp)
{
int tz;
int hr,
Fix up some problems in handling of zic-style time zone names in datetime input routines. Remove the former "DecodePosixTimezone" function in favor of letting the zic code handle POSIX-style zone specs (see tzparse()). In particular this means that "PST+3" now means the same as "-03", whereas it used to mean "-11" --- the zone abbreviation is effectively just a noise word in this syntax. Make sure that all named and POSIX-style zone names will be parsed as a single token. Fix long-standing bogosities in printing and input of fractional-hour timezone offsets (since the tzparse() code will accept these, we'd better make 'em work). Also correct an error in the original coding of the zic-zone-name patch: in "timestamp without time zone" input, zone names are supposed to be allowed but ignored, but the coding was such that the zone changed the interpretation anyway.
2006-10-17 23:03:21 +02:00
min,
sec = 0;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
char *cp;
Add code to test for unknown timezone names (following some ideas from Ross Reedstrom, a couple months back) and to detect timezones that are using leap-second timekeeping. The unknown-zone-name test is pretty heuristic and ugly, but it seems better than the old behavior of just switching to GMT given a bad name. Also make DecodePosixTimezone() a tad more robust.
2003-05-18 03:06:26 +02:00
/* leading character must be "+" or "-" */
if (*str != '+' && *str != '-')
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
return DTERR_BAD_FORMAT;
Add code to test for unknown timezone names (following some ideas from Ross Reedstrom, a couple months back) and to detect timezones that are using leap-second timekeeping. The unknown-zone-name test is pretty heuristic and ugly, but it seems better than the old behavior of just switching to GMT given a bad name. Also make DecodePosixTimezone() a tad more robust.
2003-05-18 03:06:26 +02:00
Remove comment on errno=0 lines, but add mention to port/strtol.c function.
2005-12-02 03:49:11 +01:00
errno = 0;
Rename strtoi() to strtoint(). NetBSD has seen fit to invent a libc function named strtoi(), which conflicts with the long-established static functions of the same name in datetime.c and ecpg's interval.c. While muttering darkly about intrusions on application namespace, we'll rename our functions to avoid the conflict. Back-patch to all supported branches, since this would affect attempts to build any of them on recent NetBSD. Thomas Munro
2016-04-23 22:53:15 +02:00
hr = strtoint(str + 1, &cp, 10);
Check for overflow in strtol() while parsing datetime inputs. Michael Fuhr.
2005-12-01 18:56:34 +01:00
if (errno == ERANGE)
return DTERR_TZDISP_OVERFLOW;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
/* explicit delimiter? */
if (*cp == ':')
Check for overflow in strtol() while parsing datetime inputs. Michael Fuhr.
2005-12-01 18:56:34 +01:00
{
Remove comment on errno=0 lines, but add mention to port/strtol.c function.
2005-12-02 03:49:11 +01:00
errno = 0;
Rename strtoi() to strtoint(). NetBSD has seen fit to invent a libc function named strtoi(), which conflicts with the long-established static functions of the same name in datetime.c and ecpg's interval.c. While muttering darkly about intrusions on application namespace, we'll rename our functions to avoid the conflict. Back-patch to all supported branches, since this would affect attempts to build any of them on recent NetBSD. Thomas Munro
2016-04-23 22:53:15 +02:00
min = strtoint(cp + 1, &cp, 10);
Check for overflow in strtol() while parsing datetime inputs. Michael Fuhr.
2005-12-01 18:56:34 +01:00
if (errno == ERANGE)
return DTERR_TZDISP_OVERFLOW;
Fix up some problems in handling of zic-style time zone names in datetime input routines. Remove the former "DecodePosixTimezone" function in favor of letting the zic code handle POSIX-style zone specs (see tzparse()). In particular this means that "PST+3" now means the same as "-03", whereas it used to mean "-11" --- the zone abbreviation is effectively just a noise word in this syntax. Make sure that all named and POSIX-style zone names will be parsed as a single token. Fix long-standing bogosities in printing and input of fractional-hour timezone offsets (since the tzparse() code will accept these, we'd better make 'em work). Also correct an error in the original coding of the zic-zone-name patch: in "timestamp without time zone" input, zone names are supposed to be allowed but ignored, but the coding was such that the zone changed the interpretation anyway.
2006-10-17 23:03:21 +02:00
if (*cp == ':')
{
errno = 0;
Rename strtoi() to strtoint(). NetBSD has seen fit to invent a libc function named strtoi(), which conflicts with the long-established static functions of the same name in datetime.c and ecpg's interval.c. While muttering darkly about intrusions on application namespace, we'll rename our functions to avoid the conflict. Back-patch to all supported branches, since this would affect attempts to build any of them on recent NetBSD. Thomas Munro
2016-04-23 22:53:15 +02:00
sec = strtoint(cp + 1, &cp, 10);
Fix up some problems in handling of zic-style time zone names in datetime input routines. Remove the former "DecodePosixTimezone" function in favor of letting the zic code handle POSIX-style zone specs (see tzparse()). In particular this means that "PST+3" now means the same as "-03", whereas it used to mean "-11" --- the zone abbreviation is effectively just a noise word in this syntax. Make sure that all named and POSIX-style zone names will be parsed as a single token. Fix long-standing bogosities in printing and input of fractional-hour timezone offsets (since the tzparse() code will accept these, we'd better make 'em work). Also correct an error in the original coding of the zic-zone-name patch: in "timestamp without time zone" input, zone names are supposed to be allowed but ignored, but the coding was such that the zone changed the interpretation anyway.
2006-10-17 23:03:21 +02:00
if (errno == ERANGE)
return DTERR_TZDISP_OVERFLOW;
}
Check for overflow in strtol() while parsing datetime inputs. Michael Fuhr.
2005-12-01 18:56:34 +01:00
}
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
/* otherwise, might have run things together... */
Remove unnecessary parentheses in datetime/timestamp code.
2005-05-23 19:13:14 +02:00
else if (*cp == '\0' && strlen(str) > 3)
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
{
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
min = hr % 100;
hr = hr / 100;
Fix up some problems in handling of zic-style time zone names in datetime input routines. Remove the former "DecodePosixTimezone" function in favor of letting the zic code handle POSIX-style zone specs (see tzparse()). In particular this means that "PST+3" now means the same as "-03", whereas it used to mean "-11" --- the zone abbreviation is effectively just a noise word in this syntax. Make sure that all named and POSIX-style zone names will be parsed as a single token. Fix long-standing bogosities in printing and input of fractional-hour timezone offsets (since the tzparse() code will accept these, we'd better make 'em work). Also correct an error in the original coding of the zic-zone-name patch: in "timestamp without time zone" input, zone names are supposed to be allowed but ignored, but the coding was such that the zone changed the interpretation anyway.
2006-10-17 23:03:21 +02:00
/* we could, but don't, support a run-together hhmmss format */
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
}
else
min = 0;
Fix datetime and abstime conversions to and from date. Bring optional new-storage date and time up to date and test. This new storage format should fix the "Sparc gcc -O2 bug". (Enable new code with USE_NEW_DATE and USE_NEW_TIME in dt.h)
1997-05-16 09:19:50 +02:00
Expand the allowed range of timezone offsets to +/-15:59:59 from Greenwich. We used to only allow offsets less than +/-13 hours, then it was +/14, then it was +/-15. That's still not good enough though, as per today's bug report from Patric Bechtel. This time I actually looked through the Olson timezone database to find the largest offsets used anywhere. The winners are Asia/Manila, at -15:56:00 until 1844, and America/Metlakatla, at +15:13:42 until 1867. So we'd better allow offsets less than +/-16 hours. Given the history, we are way overdue to have some greppable #define symbols controlling this, so make some ... and also remove an obsolete comment that didn't get fixed the last time. Back-patch to all supported branches.
2012-05-31 01:58:35 +02:00
/* Range-check the values; see notes in datatype/timestamp.h */
if (hr < 0 || hr > MAX_TZDISP_HOUR)
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
return DTERR_TZDISP_OVERFLOW;
Use macros for time-based constants, rather than constants.
2011-03-12 15:31:18 +01:00
if (min < 0 || min >= MINS_PER_HOUR)
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
return DTERR_TZDISP_OVERFLOW;
Use macros for time-based constants, rather than constants.
2011-03-12 15:31:18 +01:00
if (sec < 0 || sec >= SECS_PER_MINUTE)
Fix up some problems in handling of zic-style time zone names in datetime input routines. Remove the former "DecodePosixTimezone" function in favor of letting the zic code handle POSIX-style zone specs (see tzparse()). In particular this means that "PST+3" now means the same as "-03", whereas it used to mean "-11" --- the zone abbreviation is effectively just a noise word in this syntax. Make sure that all named and POSIX-style zone names will be parsed as a single token. Fix long-standing bogosities in printing and input of fractional-hour timezone offsets (since the tzparse() code will accept these, we'd better make 'em work). Also correct an error in the original coding of the zic-zone-name patch: in "timestamp without time zone" input, zone names are supposed to be allowed but ignored, but the coding was such that the zone changed the interpretation anyway.
2006-10-17 23:03:21 +02:00
return DTERR_TZDISP_OVERFLOW;
Add code to test for unknown timezone names (following some ideas from Ross Reedstrom, a couple months back) and to detect timezones that are using leap-second timekeeping. The unknown-zone-name test is pretty heuristic and ugly, but it seems better than the old behavior of just switching to GMT given a bad name. Also make DecodePosixTimezone() a tad more robust.
2003-05-18 03:06:26 +02:00
Fix up some problems in handling of zic-style time zone names in datetime input routines. Remove the former "DecodePosixTimezone" function in favor of letting the zic code handle POSIX-style zone specs (see tzparse()). In particular this means that "PST+3" now means the same as "-03", whereas it used to mean "-11" --- the zone abbreviation is effectively just a noise word in this syntax. Make sure that all named and POSIX-style zone names will be parsed as a single token. Fix long-standing bogosities in printing and input of fractional-hour timezone offsets (since the tzparse() code will accept these, we'd better make 'em work). Also correct an error in the original coding of the zic-zone-name patch: in "timestamp without time zone" input, zone names are supposed to be allowed but ignored, but the coding was such that the zone changed the interpretation anyway.
2006-10-17 23:03:21 +02:00
tz = (hr * MINS_PER_HOUR + min) * SECS_PER_MINUTE + sec;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
if (*str == '-')
tz = -tz;
Fix datetime and abstime conversions to and from date. Bring optional new-storage date and time up to date and test. This new storage format should fix the "Sparc gcc -O2 bug". (Enable new code with USE_NEW_DATE and USE_NEW_TIME in dt.h)
1997-05-16 09:19:50 +02:00
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
*tzp = -tz;
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
if (*cp != '\0')
return DTERR_BAD_FORMAT;
return 0;
}
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
/* DecodeTimezoneAbbrev()
* Interpret string as a timezone abbreviation, if possible.
*
* Returns an abbreviation type (TZ, DTZ, or DYNTZ), or UNKNOWN_FIELD if
* string is not any known abbreviation. On success, set *offset and *tz to
* represent the UTC offset (for TZ or DTZ) or underlying zone (for DYNTZ).
* Note that full timezone names (such as America/New_York) are not handled
* here, mostly for historical reasons.
*
* Given string must be lowercased already.
*
* Implement a cache lookup since it is likely that dates
* will be related in format.
*/
int
DecodeTimezoneAbbrev(int field, char *lowtoken,
int *offset, pg_tz **tz)
{
int type;
const datetkn *tp;
tp = abbrevcache[field];
/* use strncmp so that we match truncated tokens */
if (tp == NULL || strncmp(lowtoken, tp->token, TOKMAXLEN) != 0)
{
if (zoneabbrevtbl)
tp = datebsearch(lowtoken, zoneabbrevtbl->abbrevs,
zoneabbrevtbl->numabbrevs);
else
tp = NULL;
}
if (tp == NULL)
{
type = UNKNOWN_FIELD;
*offset = 0;
*tz = NULL;
}
else
{
abbrevcache[field] = tp;
type = tp->type;
if (type == DYNTZ)
{
*offset = 0;
*tz = FetchDynamicTimeZone(zoneabbrevtbl, tp);
}
else
{
*offset = tp->value;
*tz = NULL;
}
}
return type;
}
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
/* DecodeSpecial()
* Decode text string using lookup table.
Add code to test for unknown timezone names (following some ideas from Ross Reedstrom, a couple months back) and to detect timezones that are using leap-second timekeeping. The unknown-zone-name test is pretty heuristic and ugly, but it seems better than the old behavior of just switching to GMT given a bad name. Also make DecodePosixTimezone() a tad more robust.
2003-05-18 03:06:26 +02:00
*
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
* Recognizes the keywords listed in datetktbl.
* Note: at one time this would also recognize timezone abbreviations,
* but no more; use DecodeTimezoneAbbrev for that.
*
* Given string must be lowercased already.
*
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
* Implement a cache lookup since it is likely that dates
* will be related in format.
*/
int
DecodeSpecial(int field, char *lowtoken, int *val)
{
int type;
Remove hard-wired lists of timezone abbreviations in favor of providing configuration files that can be altered by a DBA. The australian_timezones GUC setting disappears, replaced by a timezone_abbreviations setting (set this to 'Australia' to get the effect of australian_timezones). The list of zone names defined by default has undergone a bit of cleanup, too. Documentation still needs some work --- in particular, should we fix Table B-4, or just get rid of it? Joachim Wieland, with some editorializing by moi.
2006-07-25 05:51:23 +02:00
const datetkn *tp;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
Remove hard-wired lists of timezone abbreviations in favor of providing configuration files that can be altered by a DBA. The australian_timezones GUC setting disappears, replaced by a timezone_abbreviations setting (set this to 'Australia' to get the effect of australian_timezones). The list of zone names defined by default has undergone a bit of cleanup, too. Documentation still needs some work --- in particular, should we fix Table B-4, or just get rid of it? Joachim Wieland, with some editorializing by moi.
2006-07-25 05:51:23 +02:00
tp = datecache[field];
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
/* use strncmp so that we match truncated tokens */
Remove hard-wired lists of timezone abbreviations in favor of providing configuration files that can be altered by a DBA. The australian_timezones GUC setting disappears, replaced by a timezone_abbreviations setting (set this to 'Australia' to get the effect of australian_timezones). The list of zone names defined by default has undergone a bit of cleanup, too. Documentation still needs some work --- in particular, should we fix Table B-4, or just get rid of it? Joachim Wieland, with some editorializing by moi.
2006-07-25 05:51:23 +02:00
if (tp == NULL || strncmp(lowtoken, tp->token, TOKMAXLEN) != 0)
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
{
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
tp = datebsearch(lowtoken, datetktbl, szdatetktbl);
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
}
if (tp == NULL)
{
Measure the current transaction time to milliseconds. Define a new function, GetCurrentTransactionStartTimeUsec() to get the time to this precision. Allow now() and timestamp 'now' to use this higher precision result so we now have fractional seconds in this "constant". Add timestamp without time zone type. Move previous timestamp type to timestamp with time zone. Accept another ISO variant for date/time values: yyyy-mm-ddThh:mm:ss (note the "T" separating the day from hours information). Remove 'current' from date/time types; convert to 'now' in input. Separate time and timetz regression tests. Separate timestamp and timestamptz regression test.
2001-09-28 10:09:14 +02:00
type = UNKNOWN_FIELD;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
*val = 0;
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
}
else
{
Remove hard-wired lists of timezone abbreviations in favor of providing configuration files that can be altered by a DBA. The australian_timezones GUC setting disappears, replaced by a timezone_abbreviations setting (set this to 'Australia' to get the effect of australian_timezones). The list of zone names defined by default has undergone a bit of cleanup, too. Documentation still needs some work --- in particular, should we fix Table B-4, or just get rid of it? Joachim Wieland, with some editorializing by moi.
2006-07-25 05:51:23 +02:00
datecache[field] = tp;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
type = tp->type;
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
*val = tp->value;
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
}
Fix datetime and abstime conversions to and from date. Bring optional new-storage date and time up to date and test. This new storage format should fix the "Sparc gcc -O2 bug". (Enable new code with USE_NEW_DATE and USE_NEW_TIME in dt.h)
1997-05-16 09:19:50 +02:00
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
return type;
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
}
Date/Time updates from Thomas...
1997-03-15 00:21:12 +01:00
Fix many typos and inconsistencies Author: Alexander Lakhin Discussion: https://postgr.es/m/af27d1b3-a128-9d62-46e0-88f424397f44@gmail.com
2019-07-01 03:00:23 +02:00
/* ClearPgTm
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
*
* Zero out a pg_tm and associated fsec_t
*/
8.4 pgindent run, with new combined Linux/FreeBSD/MinGW typedef list provided by Andrew.
2009-06-11 16:49:15 +02:00
static inline void
Initial pgindent run with pg_bsd_indent version 2.0. The new indent version includes numerous fixes thanks to Piotr Stefaniak. The main changes visible in this commit are: * Nicer formatting of function-pointer declarations. * No longer unexpectedly removes spaces in expressions using casts, sizeof, or offsetof. * No longer wants to add a space in "struct structname *varname", as well as some similar cases for const- or volatile-qualified pointers. * Declarations using PG_USED_FOR_ASSERTS_ONLY are formatted more nicely. * Fixes bug where comments following declarations were sometimes placed with no space separating them from the code. * Fixes some odd decisions for comments following case labels. * Fixes some cases where comments following code were indented to less than the expected column 33. On the less good side, it now tends to put more whitespace around typedef names that are not listed in typedefs.list. This might encourage us to put more effort into typedef name collection; it's not really a bug in indent itself. There are more changes coming after this round, having to do with comment indentation and alignment of lines appearing within parentheses. I wanted to limit the size of the diffs to something that could be reviewed without one's eyes completely glazing over, so it seemed better to split up the changes as much as practical. Discussion: https://postgr.es/m/E1dAmxK-0006EE-1r@gemulon.postgresql.org Discussion: https://postgr.es/m/30527.1495162840@sss.pgh.pa.us
2017-06-21 20:39:04 +02:00
ClearPgTm(struct pg_tm *tm, fsec_t *fsec)
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
{
tm->tm_year = 0;
8.4 pgindent run, with new combined Linux/FreeBSD/MinGW typedef list provided by Andrew.
2009-06-11 16:49:15 +02:00
tm->tm_mon = 0;
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
tm->tm_mday = 0;
tm->tm_hour = 0;
8.4 pgindent run, with new combined Linux/FreeBSD/MinGW typedef list provided by Andrew.
2009-06-11 16:49:15 +02:00
tm->tm_min = 0;
tm->tm_sec = 0;
*fsec = 0;
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
}
Support alternate storage scheme of 64-bit integer for date/time types. Use "--enable-integer-datetimes" in configuration to use this rather than the original float8 storage. I would recommend the integer-based storage for any platform on which it is available. We perhaps should make this the default for the production release. Change timezone(timestamptz) results to return timestamp rather than a character string. Formerly, we didn't have a way to represent timestamps with an explicit time zone other than freezing the info into a string. Now, we can reasonably omit the explicit time zone from the result and return a timestamp with values appropriate for the specified time zone. Much cleaner, and if you need the time zone in the result you can put it into a character string pretty easily anyway. Allow fractional seconds in date/time types even for dates prior to 1BC. Limit timestamp data types to 6 decimal places of precision. Just right for a micro-second storage of int8 date/time types, and reduces the number of places ad-hoc rounding was occuring for the float8-based types. Use lookup tables for precision/rounding calculations for timestamp and interval types. Formerly used pow() to calculate the desired value but with a more limited range there is no reason to not type in a lookup table. Should be *much* better performance, though formerly there were some optimizations to help minimize the number of times pow() was called. Define a HAVE_INT64_TIMESTAMP variable. Based on the configure option "--enable-integer-datetimes" and the existing internal INT64_IS_BUSTED. Add explicit date/interval operators and functions for addition and subtraction. Formerly relied on implicit type promotion from date to timestamp with time zone. Change timezone conversion functions for the timetz type from "timetz()" to "timezone()". This is consistant with other time zone coersion functions for other types. Bump the catalog version to 200204201. Fix up regression tests to reflect changes in fractional seconds representation for date/times in BC eras. All regression tests pass on my Linux box.
2002-04-21 21:52:18 +02:00
/* DecodeInterval()
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
* Interpret previously parsed fields for general time interval.
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
* Returns 0 if successful, DTERR code if bogus input detected.
Add support for input and output of interval values formatted per ISO 8601; specifically, we can input either the "format with designators" or the "alternative format", and we can output the former when IntervalStyle is set to iso_8601. Ron Mayer
2008-11-11 03:42:33 +01:00
* dtype, tm, fsec are output parameters.
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
*
* Allow "date" field DTK_DATE since this could be just
* an unsigned floating point number. - thomas 1997-11-16
*
* Allow ISO-style time span, with implicit units on number of days
Spell 'precedes', 'preceding' correctly in various places.
2001-11-21 23:57:01 +01:00
* preceding an hh:mm:ss field. - thomas 1998-04-30
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
*/
int
Make our parsing of INTERVAL literals spec-compliant (or at least a heck of a lot closer than it was before). To do this, tweak coerce_type() to pass through the typmod information when invoking interval_in() on an UNKNOWN constant; then fix DecodeInterval to pay attention to the typmod when deciding how to interpret a units-less integer value. I changed one or two other details as well. I believe the code now reacts as expected by spec for all the literal syntaxes that are specifically enumerated in the spec. There are corner cases involving strings that don't exactly match the set of fields called out by the typmod, for which we might want to tweak the behavior some more; but I think this is an area of user friendliness rather than spec compliance. There remain some non-compliant details about the SQL syntax (as opposed to what's inside the literal string); but at least we'll throw error rather than silently doing the wrong thing in those cases.
2008-09-10 20:29:41 +02:00
DecodeInterval(char **field, int *ftype, int nf, int range,
Initial pgindent run with pg_bsd_indent version 2.0. The new indent version includes numerous fixes thanks to Piotr Stefaniak. The main changes visible in this commit are: * Nicer formatting of function-pointer declarations. * No longer unexpectedly removes spaces in expressions using casts, sizeof, or offsetof. * No longer wants to add a space in "struct structname *varname", as well as some similar cases for const- or volatile-qualified pointers. * Declarations using PG_USED_FOR_ASSERTS_ONLY are formatted more nicely. * Fixes bug where comments following declarations were sometimes placed with no space separating them from the code. * Fixes some odd decisions for comments following case labels. * Fixes some cases where comments following code were indented to less than the expected column 33. On the less good side, it now tends to put more whitespace around typedef names that are not listed in typedefs.list. This might encourage us to put more effort into typedef name collection; it's not really a bug in indent itself. There are more changes coming after this round, having to do with comment indentation and alignment of lines appearing within parentheses. I wanted to limit the size of the diffs to something that could be reviewed without one's eyes completely glazing over, so it seemed better to split up the changes as much as practical. Discussion: https://postgr.es/m/E1dAmxK-0006EE-1r@gemulon.postgresql.org Discussion: https://postgr.es/m/30527.1495162840@sss.pgh.pa.us
2017-06-21 20:39:04 +02:00
int *dtype, struct pg_tm *tm, fsec_t *fsec)
Date/Time updates from Thomas...
1997-03-15 00:21:12 +01:00
{
Change TRUE/FALSE to true/false The lower case spellings are C and C++ standard and are used in most parts of the PostgreSQL sources. The upper case spellings are only used in some files/modules. So standardize on the standard spellings. The APIs for ICU, Perl, and Windows define their own TRUE and FALSE, so those are left as is when using those APIs. In code comments, we use the lower-case spelling for the C concepts and keep the upper-case spelling for the SQL concepts. Reviewed-by: Michael Paquier <michael.paquier@gmail.com>
2017-08-16 06:22:32 +02:00
bool is_before = false;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
char *cp;
int fmask = 0,
tmask,
type;
int i;
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
int dterr;
Fix INTERVAL output when year/month has different sign as day/hour etc. Previously, all fields were unsigned, with only a trailing "ago" to indicate negative intervals. Now, ISO format does not use "ago", and and the traditional PostgreSQL format has the first numeric field unsigned with "ago" supporting that field. So "1 month - 2 days ago" is two days less than a month in the past. Fix interval arithmetic across daylight savings time boundaries. Previously, most math across boundaries introduced a one hour offset. Allow some date/time functions to return NULL if called with NULL args. Implement functions for AT TIME ZONE support. Support "SAT" as an Australian time zone if USE_AUSTRALIAN_RULES is defined.
2000-11-06 16:57:00 +01:00
int val;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
double fval;
*dtype = DTK_DELTA;
Katherine Ward wrote: > Changes to avoid collisions with WIN32 & MFC names... > 1. Renamed: > a. PROC => PGPROC > b. GetUserName() => GetUserNameFromId() > c. GetCurrentTime() => GetCurrentDateTime() > d. IGNORE => IGNORE_DTF in include/utils/datetime.h & utils/adt/datetim > > 2. Added _P to some lex/yacc tokens: > CONST, CHAR, DELETE, FLOAT, GROUP, IN, OUT Jan
2002-06-11 15:40:53 +02:00
type = IGNORE_DTF;
8.4 pgindent run, with new combined Linux/FreeBSD/MinGW typedef list provided by Andrew.
2009-06-11 16:49:15 +02:00
ClearPgTm(tm, fsec);
Date/Time updates from Thomas...
1997-03-15 00:21:12 +01:00
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
/* read through list backwards to pick up units before values */
for (i = nf - 1; i >= 0; i--)
{
switch (ftype[i])
{
case DTK_TIME:
Make our parsing of INTERVAL literals spec-compliant (or at least a heck of a lot closer than it was before). To do this, tweak coerce_type() to pass through the typmod information when invoking interval_in() on an UNKNOWN constant; then fix DecodeInterval to pay attention to the typmod when deciding how to interpret a units-less integer value. I changed one or two other details as well. I believe the code now reacts as expected by spec for all the literal syntaxes that are specifically enumerated in the spec. There are corner cases involving strings that don't exactly match the set of fields called out by the typmod, for which we might want to tweak the behavior some more; but I think this is an area of user friendliness rather than spec compliance. There remain some non-compliant details about the SQL syntax (as opposed to what's inside the literal string); but at least we'll throw error rather than silently doing the wrong thing in those cases.
2008-09-10 20:29:41 +02:00
dterr = DecodeTime(field[i], fmask, range,
&tmask, tm, fsec);
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
if (dterr)
return dterr;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
type = DTK_DAY;
break;
case DTK_TZ:
pgindent run. Make it all clean.
2001-03-22 05:01:46 +01:00
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
/*
Fix bugs with parsing signed hh:mm and hh:mm:ss fields in interval input. DecodeInterval() failed to honor the "range" parameter (the special SQL syntax for indicating which fields appear in the literal string) if the time was signed. This seems inappropriate, so make it work like the not-signed case. The inconsistency was introduced in my commit f867339c0148381eb1d01f93ab5c79f9d10211de, which as noted in its log message was only really focused on making SQL-compliant literals work per spec. Including a sign here is not per spec, but if we're going to allow it then it's reasonable to expect it to work like the not-signed case. Also, remove bogus setting of tmask, which caused subsequent processing to think that what had been given was a timezone and not an hh:mm(:ss) field, thus confusing checks for redundant fields. This seems to be an aboriginal mistake in Lockhart's commit 2cf1642461536d0d8f3a1cf124ead0eac04eb760. Add regression test cases to illustrate the changed behaviors. Back-patch as far as 8.4, where support for spec-compliant interval literals was added. Range problem reported and diagnosed by Amit Kapila, tmask problem by me.
2012-08-03 23:39:39 +02:00
* Timezone means a token with a leading sign character and at
8.4 pgindent run, with new combined Linux/FreeBSD/MinGW typedef list provided by Andrew.
2009-06-11 16:49:15 +02:00
* least one digit; there could be ':', '.', '-' embedded in
* it as well.
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
*/
Remove unnecessary parentheses in datetime/timestamp code.
2005-05-23 19:13:14 +02:00
Assert(*field[i] == '-' || *field[i] == '+');
pgindent run. Make it all clean.
2001-03-22 05:01:46 +01:00
/*
pgindent run for 9.4 This includes removing tabs after periods in C comments, which was applied to back branches, so this change should not effect backpatching.
2014-05-06 18:12:18 +02:00
* Check for signed hh:mm or hh:mm:ss. If so, process exactly
Fix bugs with parsing signed hh:mm and hh:mm:ss fields in interval input. DecodeInterval() failed to honor the "range" parameter (the special SQL syntax for indicating which fields appear in the literal string) if the time was signed. This seems inappropriate, so make it work like the not-signed case. The inconsistency was introduced in my commit f867339c0148381eb1d01f93ab5c79f9d10211de, which as noted in its log message was only really focused on making SQL-compliant literals work per spec. Including a sign here is not per spec, but if we're going to allow it then it's reasonable to expect it to work like the not-signed case. Also, remove bogus setting of tmask, which caused subsequent processing to think that what had been given was a timezone and not an hh:mm(:ss) field, thus confusing checks for redundant fields. This seems to be an aboriginal mistake in Lockhart's commit 2cf1642461536d0d8f3a1cf124ead0eac04eb760. Add regression test cases to illustrate the changed behaviors. Back-patch as far as 8.4, where support for spec-compliant interval literals was added. Range problem reported and diagnosed by Amit Kapila, tmask problem by me.
2012-08-03 23:39:39 +02:00
* like DTK_TIME case above, plus handling the sign.
Fix INTERVAL output when year/month has different sign as day/hour etc. Previously, all fields were unsigned, with only a trailing "ago" to indicate negative intervals. Now, ISO format does not use "ago", and and the traditional PostgreSQL format has the first numeric field unsigned with "ago" supporting that field. So "1 month - 2 days ago" is two days less than a month in the past. Fix interval arithmetic across daylight savings time boundaries. Previously, most math across boundaries introduced a one hour offset. Allow some date/time functions to return NULL if called with NULL args. Implement functions for AT TIME ZONE support. Support "SAT" as an Australian time zone if USE_AUSTRALIAN_RULES is defined.
2000-11-06 16:57:00 +01:00
*/
Clean up a couple of weird corner cases in interval parsing: make -yyyy-mm be interpreted as expected (the sign should affect months too), and get rid of hard-wired assumption that unmarked signed values must be hours (if integers) or seconds (if floats). The former was just a bug in my previous patch, while the latter may have made sense at one time but seems illogical now that we support determination of the units from typmod information. Ron Mayer and myself.
2008-09-17 00:31:21 +02:00
if (strchr(field[i] + 1, ':') != NULL &&
Fix bugs with parsing signed hh:mm and hh:mm:ss fields in interval input. DecodeInterval() failed to honor the "range" parameter (the special SQL syntax for indicating which fields appear in the literal string) if the time was signed. This seems inappropriate, so make it work like the not-signed case. The inconsistency was introduced in my commit f867339c0148381eb1d01f93ab5c79f9d10211de, which as noted in its log message was only really focused on making SQL-compliant literals work per spec. Including a sign here is not per spec, but if we're going to allow it then it's reasonable to expect it to work like the not-signed case. Also, remove bogus setting of tmask, which caused subsequent processing to think that what had been given was a timezone and not an hh:mm(:ss) field, thus confusing checks for redundant fields. This seems to be an aboriginal mistake in Lockhart's commit 2cf1642461536d0d8f3a1cf124ead0eac04eb760. Add regression test cases to illustrate the changed behaviors. Back-patch as far as 8.4, where support for spec-compliant interval literals was added. Range problem reported and diagnosed by Amit Kapila, tmask problem by me.
2012-08-03 23:39:39 +02:00
DecodeTime(field[i] + 1, fmask, range,
Make our parsing of INTERVAL literals spec-compliant (or at least a heck of a lot closer than it was before). To do this, tweak coerce_type() to pass through the typmod information when invoking interval_in() on an UNKNOWN constant; then fix DecodeInterval to pay attention to the typmod when deciding how to interpret a units-less integer value. I changed one or two other details as well. I believe the code now reacts as expected by spec for all the literal syntaxes that are specifically enumerated in the spec. There are corner cases involving strings that don't exactly match the set of fields called out by the typmod, for which we might want to tweak the behavior some more; but I think this is an area of user friendliness rather than spec compliance. There remain some non-compliant details about the SQL syntax (as opposed to what's inside the literal string); but at least we'll throw error rather than silently doing the wrong thing in those cases.
2008-09-10 20:29:41 +02:00
&tmask, tm, fsec) == 0)
pgindent run. Make it all clean.
2001-03-22 05:01:46 +01:00
{
if (*field[i] == '-')
{
Fix INTERVAL output when year/month has different sign as day/hour etc. Previously, all fields were unsigned, with only a trailing "ago" to indicate negative intervals. Now, ISO format does not use "ago", and and the traditional PostgreSQL format has the first numeric field unsigned with "ago" supporting that field. So "1 month - 2 days ago" is two days less than a month in the past. Fix interval arithmetic across daylight savings time boundaries. Previously, most math across boundaries introduced a one hour offset. Allow some date/time functions to return NULL if called with NULL args. Implement functions for AT TIME ZONE support. Support "SAT" as an Australian time zone if USE_AUSTRALIAN_RULES is defined.
2000-11-06 16:57:00 +01:00
/* flip the sign on all fields */
tm->tm_hour = -tm->tm_hour;
tm->tm_min = -tm->tm_min;
tm->tm_sec = -tm->tm_sec;
*fsec = -(*fsec);
}
pgindent run. Make it all clean.
2001-03-22 05:01:46 +01:00
/*
* Set the next type to be a day, if units are not
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* specified. This handles the case of '1 +02:03' since we
* are reading right to left.
Fix INTERVAL output when year/month has different sign as day/hour etc. Previously, all fields were unsigned, with only a trailing "ago" to indicate negative intervals. Now, ISO format does not use "ago", and and the traditional PostgreSQL format has the first numeric field unsigned with "ago" supporting that field. So "1 month - 2 days ago" is two days less than a month in the past. Fix interval arithmetic across daylight savings time boundaries. Previously, most math across boundaries introduced a one hour offset. Allow some date/time functions to return NULL if called with NULL args. Implement functions for AT TIME ZONE support. Support "SAT" as an Australian time zone if USE_AUSTRALIAN_RULES is defined.
2000-11-06 16:57:00 +01:00
*/
type = DTK_DAY;
break;
pgindent run. Make it all clean.
2001-03-22 05:01:46 +01:00
}
Fix bugs with parsing signed hh:mm and hh:mm:ss fields in interval input. DecodeInterval() failed to honor the "range" parameter (the special SQL syntax for indicating which fields appear in the literal string) if the time was signed. This seems inappropriate, so make it work like the not-signed case. The inconsistency was introduced in my commit f867339c0148381eb1d01f93ab5c79f9d10211de, which as noted in its log message was only really focused on making SQL-compliant literals work per spec. Including a sign here is not per spec, but if we're going to allow it then it's reasonable to expect it to work like the not-signed case. Also, remove bogus setting of tmask, which caused subsequent processing to think that what had been given was a timezone and not an hh:mm(:ss) field, thus confusing checks for redundant fields. This seems to be an aboriginal mistake in Lockhart's commit 2cf1642461536d0d8f3a1cf124ead0eac04eb760. Add regression test cases to illustrate the changed behaviors. Back-patch as far as 8.4, where support for spec-compliant interval literals was added. Range problem reported and diagnosed by Amit Kapila, tmask problem by me.
2012-08-03 23:39:39 +02:00
/*
* Otherwise, fall through to DTK_NUMBER case, which can
* handle signed float numbers and signed year-month values.
*/
Clean up warnings from -Wimplicit-fallthrough. Recent gcc can warn about switch-case fall throughs that are not explicitly labeled as intentional. This seems like a good thing, so clean up the warnings exposed thereby by labeling all such cases with comments that gcc will recognize. In files that already had one or more suitable comments, I generally matched the existing style of those. Otherwise I went with /* FALLTHROUGH */, which is one of the spellings approved at the more-restrictive-than-default level -Wimplicit-fallthrough=4. (At the default level you can also spell it /* FALL ?THRU */, and it's not picky about case. What you can't do is include additional text in the same comment, so some existing comments containing versions of this aren't good enough.) Testing with gcc 8.0.1 (Fedora 28's current version), I found that I also had to put explicit "break"s after elog(ERROR) or ereport(ERROR); apparently, for this purpose gcc doesn't recognize that those don't return. That seems like possibly a gcc bug, but it's fine because in most places we did that anyway; so this amounts to a visit from the style police. Discussion: https://postgr.es/m/15083.1525207729@sss.pgh.pa.us
2018-05-02 01:35:08 +02:00
/* FALLTHROUGH */
Fix INTERVAL output when year/month has different sign as day/hour etc. Previously, all fields were unsigned, with only a trailing "ago" to indicate negative intervals. Now, ISO format does not use "ago", and and the traditional PostgreSQL format has the first numeric field unsigned with "ago" supporting that field. So "1 month - 2 days ago" is two days less than a month in the past. Fix interval arithmetic across daylight savings time boundaries. Previously, most math across boundaries introduced a one hour offset. Allow some date/time functions to return NULL if called with NULL args. Implement functions for AT TIME ZONE support. Support "SAT" as an Australian time zone if USE_AUSTRALIAN_RULES is defined.
2000-11-06 16:57:00 +01:00
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
case DTK_DATE:
case DTK_NUMBER:
Make our parsing of INTERVAL literals spec-compliant (or at least a heck of a lot closer than it was before). To do this, tweak coerce_type() to pass through the typmod information when invoking interval_in() on an UNKNOWN constant; then fix DecodeInterval to pay attention to the typmod when deciding how to interpret a units-less integer value. I changed one or two other details as well. I believe the code now reacts as expected by spec for all the literal syntaxes that are specifically enumerated in the spec. There are corner cases involving strings that don't exactly match the set of fields called out by the typmod, for which we might want to tweak the behavior some more; but I think this is an area of user friendliness rather than spec compliance. There remain some non-compliant details about the SQL syntax (as opposed to what's inside the literal string); but at least we'll throw error rather than silently doing the wrong thing in those cases.
2008-09-10 20:29:41 +02:00
if (type == IGNORE_DTF)
{
Clean up a couple of weird corner cases in interval parsing: make -yyyy-mm be interpreted as expected (the sign should affect months too), and get rid of hard-wired assumption that unmarked signed values must be hours (if integers) or seconds (if floats). The former was just a bug in my previous patch, while the latter may have made sense at one time but seems illogical now that we support determination of the units from typmod information. Ron Mayer and myself.
2008-09-17 00:31:21 +02:00
/* use typmod to decide what rightmost field is */
Make our parsing of INTERVAL literals spec-compliant (or at least a heck of a lot closer than it was before). To do this, tweak coerce_type() to pass through the typmod information when invoking interval_in() on an UNKNOWN constant; then fix DecodeInterval to pay attention to the typmod when deciding how to interpret a units-less integer value. I changed one or two other details as well. I believe the code now reacts as expected by spec for all the literal syntaxes that are specifically enumerated in the spec. There are corner cases involving strings that don't exactly match the set of fields called out by the typmod, for which we might want to tweak the behavior some more; but I think this is an area of user friendliness rather than spec compliance. There remain some non-compliant details about the SQL syntax (as opposed to what's inside the literal string); but at least we'll throw error rather than silently doing the wrong thing in those cases.
2008-09-10 20:29:41 +02:00
switch (range)
{
case INTERVAL_MASK(YEAR):
type = DTK_YEAR;
break;
case INTERVAL_MASK(MONTH):
case INTERVAL_MASK(YEAR) | INTERVAL_MASK(MONTH):
type = DTK_MONTH;
break;
case INTERVAL_MASK(DAY):
type = DTK_DAY;
break;
case INTERVAL_MASK(HOUR):
case INTERVAL_MASK(DAY) | INTERVAL_MASK(HOUR):
type = DTK_HOUR;
break;
case INTERVAL_MASK(MINUTE):
case INTERVAL_MASK(HOUR) | INTERVAL_MASK(MINUTE):
Make handling of INTERVAL DAY TO MINUTE and INTERVAL DAY TO SECOND input more consistent with other cases, by having an unlabeled integer field be treated as a number of minutes or seconds respectively. These cases are outside the spec (which insists on full "dd hh:mm" or "dd hh:mm:ss" input respectively), so it's not much help to us in deciding what to do. But with this change, it's uniformly the case that an unlabeled integer will be considered as being a number of the interval's rightmost field. The change also takes us back to the 8.3 behavior of throwing error for certain ambiguous inputs such as INTERVAL '1 2' DAY TO MINUTE. Per recent discussion.
2009-06-10 07:05:03 +02:00
case INTERVAL_MASK(DAY) | INTERVAL_MASK(HOUR) | INTERVAL_MASK(MINUTE):
Make our parsing of INTERVAL literals spec-compliant (or at least a heck of a lot closer than it was before). To do this, tweak coerce_type() to pass through the typmod information when invoking interval_in() on an UNKNOWN constant; then fix DecodeInterval to pay attention to the typmod when deciding how to interpret a units-less integer value. I changed one or two other details as well. I believe the code now reacts as expected by spec for all the literal syntaxes that are specifically enumerated in the spec. There are corner cases involving strings that don't exactly match the set of fields called out by the typmod, for which we might want to tweak the behavior some more; but I think this is an area of user friendliness rather than spec compliance. There remain some non-compliant details about the SQL syntax (as opposed to what's inside the literal string); but at least we'll throw error rather than silently doing the wrong thing in those cases.
2008-09-10 20:29:41 +02:00
type = DTK_MINUTE;
break;
case INTERVAL_MASK(SECOND):
case INTERVAL_MASK(MINUTE) | INTERVAL_MASK(SECOND):
Make handling of INTERVAL DAY TO MINUTE and INTERVAL DAY TO SECOND input more consistent with other cases, by having an unlabeled integer field be treated as a number of minutes or seconds respectively. These cases are outside the spec (which insists on full "dd hh:mm" or "dd hh:mm:ss" input respectively), so it's not much help to us in deciding what to do. But with this change, it's uniformly the case that an unlabeled integer will be considered as being a number of the interval's rightmost field. The change also takes us back to the 8.3 behavior of throwing error for certain ambiguous inputs such as INTERVAL '1 2' DAY TO MINUTE. Per recent discussion.
2009-06-10 07:05:03 +02:00
case INTERVAL_MASK(HOUR) | INTERVAL_MASK(MINUTE) | INTERVAL_MASK(SECOND):
case INTERVAL_MASK(DAY) | INTERVAL_MASK(HOUR) | INTERVAL_MASK(MINUTE) | INTERVAL_MASK(SECOND):
Make our parsing of INTERVAL literals spec-compliant (or at least a heck of a lot closer than it was before). To do this, tweak coerce_type() to pass through the typmod information when invoking interval_in() on an UNKNOWN constant; then fix DecodeInterval to pay attention to the typmod when deciding how to interpret a units-less integer value. I changed one or two other details as well. I believe the code now reacts as expected by spec for all the literal syntaxes that are specifically enumerated in the spec. There are corner cases involving strings that don't exactly match the set of fields called out by the typmod, for which we might want to tweak the behavior some more; but I think this is an area of user friendliness rather than spec compliance. There remain some non-compliant details about the SQL syntax (as opposed to what's inside the literal string); but at least we'll throw error rather than silently doing the wrong thing in those cases.
2008-09-10 20:29:41 +02:00
type = DTK_SECOND;
break;
default:
type = DTK_SECOND;
break;
}
}
Remove comment on errno=0 lines, but add mention to port/strtol.c function.
2005-12-02 03:49:11 +01:00
errno = 0;
Rename strtoi() to strtoint(). NetBSD has seen fit to invent a libc function named strtoi(), which conflicts with the long-established static functions of the same name in datetime.c and ecpg's interval.c. While muttering darkly about intrusions on application namespace, we'll rename our functions to avoid the conflict. Back-patch to all supported branches, since this would affect attempts to build any of them on recent NetBSD. Thomas Munro
2016-04-23 22:53:15 +02:00
val = strtoint(field[i], &cp, 10);
Check for overflow in strtol() while parsing datetime inputs. Michael Fuhr.
2005-12-01 18:56:34 +01:00
if (errno == ERANGE)
return DTERR_FIELD_OVERFLOW;
Fix INTERVAL output when year/month has different sign as day/hour etc. Previously, all fields were unsigned, with only a trailing "ago" to indicate negative intervals. Now, ISO format does not use "ago", and and the traditional PostgreSQL format has the first numeric field unsigned with "ago" supporting that field. So "1 month - 2 days ago" is two days less than a month in the past. Fix interval arithmetic across daylight savings time boundaries. Previously, most math across boundaries introduced a one hour offset. Allow some date/time functions to return NULL if called with NULL args. Implement functions for AT TIME ZONE support. Support "SAT" as an Australian time zone if USE_AUSTRALIAN_RULES is defined.
2000-11-06 16:57:00 +01:00
Make our parsing of INTERVAL literals spec-compliant (or at least a heck of a lot closer than it was before). To do this, tweak coerce_type() to pass through the typmod information when invoking interval_in() on an UNKNOWN constant; then fix DecodeInterval to pay attention to the typmod when deciding how to interpret a units-less integer value. I changed one or two other details as well. I believe the code now reacts as expected by spec for all the literal syntaxes that are specifically enumerated in the spec. There are corner cases involving strings that don't exactly match the set of fields called out by the typmod, for which we might want to tweak the behavior some more; but I think this is an area of user friendliness rather than spec compliance. There remain some non-compliant details about the SQL syntax (as opposed to what's inside the literal string); but at least we'll throw error rather than silently doing the wrong thing in those cases.
2008-09-10 20:29:41 +02:00
if (*cp == '-')
{
/* SQL "years-months" syntax */
8.4 pgindent run, with new combined Linux/FreeBSD/MinGW typedef list provided by Andrew.
2009-06-11 16:49:15 +02:00
int val2;
Measure the current transaction time to milliseconds. Define a new function, GetCurrentTransactionStartTimeUsec() to get the time to this precision. Allow now() and timestamp 'now' to use this higher precision result so we now have fractional seconds in this "constant". Add timestamp without time zone type. Move previous timestamp type to timestamp with time zone. Accept another ISO variant for date/time values: yyyy-mm-ddThh:mm:ss (note the "T" separating the day from hours information). Remove 'current' from date/time types; convert to 'now' in input. Separate time and timetz regression tests. Separate timestamp and timestamptz regression test.
2001-09-28 10:09:14 +02:00
Rename strtoi() to strtoint(). NetBSD has seen fit to invent a libc function named strtoi(), which conflicts with the long-established static functions of the same name in datetime.c and ecpg's interval.c. While muttering darkly about intrusions on application namespace, we'll rename our functions to avoid the conflict. Back-patch to all supported branches, since this would affect attempts to build any of them on recent NetBSD. Thomas Munro
2016-04-23 22:53:15 +02:00
val2 = strtoint(cp + 1, &cp, 10);
Make our parsing of INTERVAL literals spec-compliant (or at least a heck of a lot closer than it was before). To do this, tweak coerce_type() to pass through the typmod information when invoking interval_in() on an UNKNOWN constant; then fix DecodeInterval to pay attention to the typmod when deciding how to interpret a units-less integer value. I changed one or two other details as well. I believe the code now reacts as expected by spec for all the literal syntaxes that are specifically enumerated in the spec. There are corner cases involving strings that don't exactly match the set of fields called out by the typmod, for which we might want to tweak the behavior some more; but I think this is an area of user friendliness rather than spec compliance. There remain some non-compliant details about the SQL syntax (as opposed to what's inside the literal string); but at least we'll throw error rather than silently doing the wrong thing in those cases.
2008-09-10 20:29:41 +02:00
if (errno == ERANGE || val2 < 0 || val2 >= MONTHS_PER_YEAR)
return DTERR_FIELD_OVERFLOW;
if (*cp != '\0')
return DTERR_BAD_FORMAT;
type = DTK_MONTH;
Fix recently added code for SQL years-months interval syntax so that it behaves correctly for a leading minus sign, zero year value, and nonzero month value. Per discussion with Ron Mayer.
2008-11-08 21:51:49 +01:00
if (*field[i] == '-')
Clean up a couple of weird corner cases in interval parsing: make -yyyy-mm be interpreted as expected (the sign should affect months too), and get rid of hard-wired assumption that unmarked signed values must be hours (if integers) or seconds (if floats). The former was just a bug in my previous patch, while the latter may have made sense at one time but seems illogical now that we support determination of the units from typmod information. Ron Mayer and myself.
2008-09-17 00:31:21 +02:00
val2 = -val2;
pgindent run for 9.4 This includes removing tabs after periods in C comments, which was applied to back branches, so this change should not effect backpatching.
2014-05-06 18:12:18 +02:00
if (((double) val * MONTHS_PER_YEAR + val2) > INT_MAX ||
((double) val * MONTHS_PER_YEAR + val2) < INT_MIN)
Add checks for interval overflow/underflow New checks include input, month/day/time internal adjustments, addition, subtraction, multiplication, and negation. Also adjust docs to correctly specify interval size in bytes. Report from Rok Kralj
2014-01-30 15:41:43 +01:00
return DTERR_FIELD_OVERFLOW;
Make our parsing of INTERVAL literals spec-compliant (or at least a heck of a lot closer than it was before). To do this, tweak coerce_type() to pass through the typmod information when invoking interval_in() on an UNKNOWN constant; then fix DecodeInterval to pay attention to the typmod when deciding how to interpret a units-less integer value. I changed one or two other details as well. I believe the code now reacts as expected by spec for all the literal syntaxes that are specifically enumerated in the spec. There are corner cases involving strings that don't exactly match the set of fields called out by the typmod, for which we might want to tweak the behavior some more; but I think this is an area of user friendliness rather than spec compliance. There remain some non-compliant details about the SQL syntax (as opposed to what's inside the literal string); but at least we'll throw error rather than silently doing the wrong thing in those cases.
2008-09-10 20:29:41 +02:00
val = val * MONTHS_PER_YEAR + val2;
fval = 0;
}
else if (*cp == '.')
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
{
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
errno = 0;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
fval = strtod(cp, &cp);
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
if (*cp != '\0' || errno != 0)
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
return DTERR_BAD_FORMAT;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
Fix DecodeInterval to handle '-0.1' sanely, per gripe from Tilo Schwarz.
2003-12-17 22:45:44 +01:00
if (*field[i] == '-')
Remove more extraneous parentheses in date/time functions.
2005-05-24 04:09:45 +02:00
fval = -fval;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
}
else if (*cp == '\0')
fval = 0;
else
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
return DTERR_BAD_FORMAT;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
tmask = 0; /* DTK_M(type); */
switch (type)
{
case DTK_MICROSEC:
Fix integral timestamps so the output is consistent in all cases to round: select interval '0:0:0.7', interval '@ 0.70 secs', interval '0.7 seconds'; Ron Mayer
2008-09-24 21:46:44 +02:00
*fsec += rint(val + fval);
Fix a bug in input processing for the "interval" type. Previously, "microsecond" and "millisecond" units were not considered valid input by themselves, which caused inputs like "1 millisecond" to be rejected erroneously. Update the docs, add regression tests, and backport to 8.2 and 8.1
2007-05-29 06:58:43 +02:00
tmask = DTK_M(MICROSECOND);
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
break;
case DTK_MILLISEC:
Fix overflow for INTERVAL 'x ms' where x is more than a couple million, and integer datetimes are in use. Per bug report from Hubert Depesz Lubaczewski. Alex Hunsaker
2009-08-18 23:23:14 +02:00
/* avoid overflowing the fsec field */
tm->tm_sec += val / 1000;
val -= (val / 1000) * 1000;
Fix integral timestamps so the output is consistent in all cases to round: select interval '0:0:0.7', interval '@ 0.70 secs', interval '0.7 seconds'; Ron Mayer
2008-09-24 21:46:44 +02:00
*fsec += rint((val + fval) * 1000);
Fix a bug in input processing for the "interval" type. Previously, "microsecond" and "millisecond" units were not considered valid input by themselves, which caused inputs like "1 millisecond" to be rejected erroneously. Update the docs, add regression tests, and backport to 8.2 and 8.1
2007-05-29 06:58:43 +02:00
tmask = DTK_M(MILLISECOND);
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
break;
case DTK_SECOND:
tm->tm_sec += val;
Fix integral timestamps so the output is consistent in all cases to round: select interval '0:0:0.7', interval '@ 0.70 secs', interval '0.7 seconds'; Ron Mayer
2008-09-24 21:46:44 +02:00
*fsec += rint(fval * 1000000);
pgindent run for 8.3.
2007-11-15 22:14:46 +01:00
Fix a bug in input processing for the "interval" type. Previously, "microsecond" and "millisecond" units were not considered valid input by themselves, which caused inputs like "1 millisecond" to be rejected erroneously. Update the docs, add regression tests, and backport to 8.2 and 8.1
2007-05-29 06:58:43 +02:00
/*
pgindent run for 8.3.
2007-11-15 22:14:46 +01:00
* If any subseconds were specified, consider this
* microsecond and millisecond input as well.
Fix a bug in input processing for the "interval" type. Previously, "microsecond" and "millisecond" units were not considered valid input by themselves, which caused inputs like "1 millisecond" to be rejected erroneously. Update the docs, add regression tests, and backport to 8.2 and 8.1
2007-05-29 06:58:43 +02:00
*/
if (fval == 0)
tmask = DTK_M(SECOND);
else
tmask = DTK_ALL_SECS_M;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
break;
case DTK_MINUTE:
tm->tm_min += val;
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
AdjustFractSeconds(fval, tm, fsec, SECS_PER_MINUTE);
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
tmask = DTK_M(MINUTE);
break;
case DTK_HOUR:
tm->tm_hour += val;
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
AdjustFractSeconds(fval, tm, fsec, SECS_PER_HOUR);
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
tmask = DTK_M(HOUR);
8.4 pgindent run, with new combined Linux/FreeBSD/MinGW typedef list provided by Andrew.
2009-06-11 16:49:15 +02:00
type = DTK_DAY; /* set for next field */
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
break;
case DTK_DAY:
tm->tm_mday += val;
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
AdjustFractSeconds(fval, tm, fsec, SECS_PER_DAY);
Fix DecodeInterval to report an error for multiple occurrences of DAY, WEEK, YEAR, DECADE, CENTURY, or MILLENIUM fields, just as it always has done for other types of fields. The previous behavior seems to have been a hack to avoid defining bit-positions for all these field types in DTK_M() masks, rather than something that was really considered to be desired behavior. But there is room in the masks for these, and we really need to tighten up at least the behavior of DAY and YEAR fields to avoid unexpected behavior associated with the 8.4 changes to interpret ambiguous fields based on the interval qualifier (typmod) value. Per my example and proposed patch.
2009-06-01 18:55:11 +02:00
tmask = DTK_M(DAY);
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
break;
case DTK_WEEK:
tm->tm_mday += val * 7;
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
AdjustFractDays(fval, tm, fsec, 7);
Fix DecodeInterval to report an error for multiple occurrences of DAY, WEEK, YEAR, DECADE, CENTURY, or MILLENIUM fields, just as it always has done for other types of fields. The previous behavior seems to have been a hack to avoid defining bit-positions for all these field types in DTK_M() masks, rather than something that was really considered to be desired behavior. But there is room in the masks for these, and we really need to tighten up at least the behavior of DAY and YEAR fields to avoid unexpected behavior associated with the 8.4 changes to interpret ambiguous fields based on the interval qualifier (typmod) value. Per my example and proposed patch.
2009-06-01 18:55:11 +02:00
tmask = DTK_M(WEEK);
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
break;
case DTK_MONTH:
tm->tm_mon += val;
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
AdjustFractDays(fval, tm, fsec, DAYS_PER_MONTH);
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
tmask = DTK_M(MONTH);
break;
case DTK_YEAR:
tm->tm_year += val;
if (fval != 0)
Add time/date macros for code clarity: #define DAYS_PER_YEAR 365.25 #define MONTHS_PER_YEAR 12 #define DAYS_PER_MONTH 30 #define HOURS_PER_DAY 24
2005-07-21 05:56:25 +02:00
tm->tm_mon += fval * MONTHS_PER_YEAR;
Fix DecodeInterval to report an error for multiple occurrences of DAY, WEEK, YEAR, DECADE, CENTURY, or MILLENIUM fields, just as it always has done for other types of fields. The previous behavior seems to have been a hack to avoid defining bit-positions for all these field types in DTK_M() masks, rather than something that was really considered to be desired behavior. But there is room in the masks for these, and we really need to tighten up at least the behavior of DAY and YEAR fields to avoid unexpected behavior associated with the 8.4 changes to interpret ambiguous fields based on the interval qualifier (typmod) value. Per my example and proposed patch.
2009-06-01 18:55:11 +02:00
tmask = DTK_M(YEAR);
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
break;
case DTK_DECADE:
tm->tm_year += val * 10;
if (fval != 0)
Add time/date macros for code clarity: #define DAYS_PER_YEAR 365.25 #define MONTHS_PER_YEAR 12 #define DAYS_PER_MONTH 30 #define HOURS_PER_DAY 24
2005-07-21 05:56:25 +02:00
tm->tm_mon += fval * MONTHS_PER_YEAR * 10;
Fix DecodeInterval to report an error for multiple occurrences of DAY, WEEK, YEAR, DECADE, CENTURY, or MILLENIUM fields, just as it always has done for other types of fields. The previous behavior seems to have been a hack to avoid defining bit-positions for all these field types in DTK_M() masks, rather than something that was really considered to be desired behavior. But there is room in the masks for these, and we really need to tighten up at least the behavior of DAY and YEAR fields to avoid unexpected behavior associated with the 8.4 changes to interpret ambiguous fields based on the interval qualifier (typmod) value. Per my example and proposed patch.
2009-06-01 18:55:11 +02:00
tmask = DTK_M(DECADE);
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
break;
case DTK_CENTURY:
tm->tm_year += val * 100;
if (fval != 0)
Add time/date macros for code clarity: #define DAYS_PER_YEAR 365.25 #define MONTHS_PER_YEAR 12 #define DAYS_PER_MONTH 30 #define HOURS_PER_DAY 24
2005-07-21 05:56:25 +02:00
tm->tm_mon += fval * MONTHS_PER_YEAR * 100;
Fix DecodeInterval to report an error for multiple occurrences of DAY, WEEK, YEAR, DECADE, CENTURY, or MILLENIUM fields, just as it always has done for other types of fields. The previous behavior seems to have been a hack to avoid defining bit-positions for all these field types in DTK_M() masks, rather than something that was really considered to be desired behavior. But there is room in the masks for these, and we really need to tighten up at least the behavior of DAY and YEAR fields to avoid unexpected behavior associated with the 8.4 changes to interpret ambiguous fields based on the interval qualifier (typmod) value. Per my example and proposed patch.
2009-06-01 18:55:11 +02:00
tmask = DTK_M(CENTURY);
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
break;
Fix (well, add) support for ISO "week" in date_part(). Needed for ODBC. Fix spelling of "millennium". Thanks to Mika Nystrom <mika@camembert.cs.caltech.edu> for spotting this.
2000-04-14 17:22:10 +02:00
case DTK_MILLENNIUM:
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
tm->tm_year += val * 1000;
if (fval != 0)
Add time/date macros for code clarity: #define DAYS_PER_YEAR 365.25 #define MONTHS_PER_YEAR 12 #define DAYS_PER_MONTH 30 #define HOURS_PER_DAY 24
2005-07-21 05:56:25 +02:00
tm->tm_mon += fval * MONTHS_PER_YEAR * 1000;
Fix DecodeInterval to report an error for multiple occurrences of DAY, WEEK, YEAR, DECADE, CENTURY, or MILLENIUM fields, just as it always has done for other types of fields. The previous behavior seems to have been a hack to avoid defining bit-positions for all these field types in DTK_M() masks, rather than something that was really considered to be desired behavior. But there is room in the masks for these, and we really need to tighten up at least the behavior of DAY and YEAR fields to avoid unexpected behavior associated with the 8.4 changes to interpret ambiguous fields based on the interval qualifier (typmod) value. Per my example and proposed patch.
2009-06-01 18:55:11 +02:00
tmask = DTK_M(MILLENNIUM);
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
break;
default:
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
return DTERR_BAD_FORMAT;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
}
break;
case DTK_STRING:
case DTK_SPECIAL:
type = DecodeUnits(i, field[i], &val);
Katherine Ward wrote: > Changes to avoid collisions with WIN32 & MFC names... > 1. Renamed: > a. PROC => PGPROC > b. GetUserName() => GetUserNameFromId() > c. GetCurrentTime() => GetCurrentDateTime() > d. IGNORE => IGNORE_DTF in include/utils/datetime.h & utils/adt/datetim > > 2. Added _P to some lex/yacc tokens: > CONST, CHAR, DELETE, FLOAT, GROUP, IN, OUT Jan
2002-06-11 15:40:53 +02:00
if (type == IGNORE_DTF)
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
continue;
tmask = 0; /* DTK_M(type); */
switch (type)
{
case UNITS:
type = val;
break;
case AGO:
Change TRUE/FALSE to true/false The lower case spellings are C and C++ standard and are used in most parts of the PostgreSQL sources. The upper case spellings are only used in some files/modules. So standardize on the standard spellings. The APIs for ICU, Perl, and Windows define their own TRUE and FALSE, so those are left as is when using those APIs. In code comments, we use the lower-case spelling for the C concepts and keep the upper-case spelling for the SQL concepts. Reviewed-by: Michael Paquier <michael.paquier@gmail.com>
2017-08-16 06:22:32 +02:00
is_before = true;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
type = val;
break;
case RESERV:
Fix an ancient typo that prevented the detection of conflicting fields when interval input "invalid" was specified together with other fields. Spotted by Neil Conway with the help of a clang warning. Although this has been wrong since the interval code was written more than 10 years ago, it doesn't affect anything beyond which error message you get for a wrong input, so not worth back-patching very far.
2010-08-02 03:24:54 +02:00
tmask = (DTK_DATE_M | DTK_TIME_M);
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
*dtype = val;
break;
default:
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
return DTERR_BAD_FORMAT;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
}
break;
default:
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
return DTERR_BAD_FORMAT;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
}
if (tmask & fmask)
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
return DTERR_BAD_FORMAT;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
fmask |= tmask;
}
Date/Time updates from Thomas...
1997-03-15 00:21:12 +01:00
Add a new GUC variable called "IntervalStyle" that decouples interval output from DateStyle, and create a new interval style that produces output matching the SQL standard (at least for interval values that fall within the standard's restrictions). IntervalStyle is also used to resolve the conflict between the standard and traditional Postgres rules for interpreting negative interval input. Ron Mayer
2008-11-09 01:28:35 +01:00
/* ensure that at least one time field has been found */
if (fmask == 0)
return DTERR_BAD_FORMAT;
/* ensure fractional seconds are fractional */
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
if (*fsec != 0)
{
pgindent run.
2002-09-04 22:31:48 +02:00
int sec;
Support alternate storage scheme of 64-bit integer for date/time types. Use "--enable-integer-datetimes" in configuration to use this rather than the original float8 storage. I would recommend the integer-based storage for any platform on which it is available. We perhaps should make this the default for the production release. Change timezone(timestamptz) results to return timestamp rather than a character string. Formerly, we didn't have a way to represent timestamps with an explicit time zone other than freezing the info into a string. Now, we can reasonably omit the explicit time zone from the result and return a timestamp with values appropriate for the specified time zone. Much cleaner, and if you need the time zone in the result you can put it into a character string pretty easily anyway. Allow fractional seconds in date/time types even for dates prior to 1BC. Limit timestamp data types to 6 decimal places of precision. Just right for a micro-second storage of int8 date/time types, and reduces the number of places ad-hoc rounding was occuring for the float8-based types. Use lookup tables for precision/rounding calculations for timestamp and interval types. Formerly used pow() to calculate the desired value but with a more limited range there is no reason to not type in a lookup table. Should be *much* better performance, though formerly there were some optimizations to help minimize the number of times pow() was called. Define a HAVE_INT64_TIMESTAMP variable. Based on the configure option "--enable-integer-datetimes" and the existing internal INT64_IS_BUSTED. Add explicit date/interval operators and functions for addition and subtraction. Formerly relied on implicit type promotion from date to timestamp with time zone. Change timezone conversion functions for the timetz type from "timetz()" to "timezone()". This is consistant with other time zone coersion functions for other types. Bump the catalog version to 200204201. Fix up regression tests to reflect changes in fractional seconds representation for date/times in BC eras. All regression tests pass on my Linux box.
2002-04-21 21:52:18 +02:00
Remove unnecessary parentheses in assignments. Add spaces where needed. Reference time interval variables as tinterval.
2005-07-21 06:41:43 +02:00
sec = *fsec / USECS_PER_SEC;
*fsec -= sec * USECS_PER_SEC;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
tm->tm_sec += sec;
}
Date/Time updates from Thomas...
1997-03-15 00:21:12 +01:00
Add a new GUC variable called "IntervalStyle" that decouples interval output from DateStyle, and create a new interval style that produces output matching the SQL standard (at least for interval values that fall within the standard's restrictions). IntervalStyle is also used to resolve the conflict between the standard and traditional Postgres rules for interpreting negative interval input. Ron Mayer
2008-11-09 01:28:35 +01:00
/*----------
* The SQL standard defines the interval literal
8.4 pgindent run, with new combined Linux/FreeBSD/MinGW typedef list provided by Andrew.
2009-06-11 16:49:15 +02:00
* '-1 1:00:00'
Add a new GUC variable called "IntervalStyle" that decouples interval output from DateStyle, and create a new interval style that produces output matching the SQL standard (at least for interval values that fall within the standard's restrictions). IntervalStyle is also used to resolve the conflict between the standard and traditional Postgres rules for interpreting negative interval input. Ron Mayer
2008-11-09 01:28:35 +01:00
* to mean "negative 1 days and negative 1 hours", while Postgres
* traditionally treats this as meaning "negative 1 days and positive
* 1 hours". In SQL_STANDARD intervalstyle, we apply the leading sign
* to all fields if there are no other explicit signs.
*
* We leave the signs alone if there are additional explicit signs.
* This protects us against misinterpreting postgres-style dump output,
* since the postgres-style output code has always put an explicit sign on
* all fields following a negative field. But note that SQL-spec output
8.4 pgindent run, with new combined Linux/FreeBSD/MinGW typedef list provided by Andrew.
2009-06-11 16:49:15 +02:00
* is ambiguous and can be misinterpreted on load! (So it's best practice
Add a new GUC variable called "IntervalStyle" that decouples interval output from DateStyle, and create a new interval style that produces output matching the SQL standard (at least for interval values that fall within the standard's restrictions). IntervalStyle is also used to resolve the conflict between the standard and traditional Postgres rules for interpreting negative interval input. Ron Mayer
2008-11-09 01:28:35 +01:00
* to dump in postgres style, not SQL style.)
*----------
*/
if (IntervalStyle == INTSTYLE_SQL_STANDARD && *field[0] == '-')
{
/* Check for additional explicit signs */
8.4 pgindent run, with new combined Linux/FreeBSD/MinGW typedef list provided by Andrew.
2009-06-11 16:49:15 +02:00
bool more_signs = false;
Add a new GUC variable called "IntervalStyle" that decouples interval output from DateStyle, and create a new interval style that produces output matching the SQL standard (at least for interval values that fall within the standard's restrictions). IntervalStyle is also used to resolve the conflict between the standard and traditional Postgres rules for interpreting negative interval input. Ron Mayer
2008-11-09 01:28:35 +01:00
for (i = 1; i < nf; i++)
{
if (*field[i] == '-' || *field[i] == '+')
{
more_signs = true;
break;
}
}
if (!more_signs)
{
/*
8.4 pgindent run, with new combined Linux/FreeBSD/MinGW typedef list provided by Andrew.
2009-06-11 16:49:15 +02:00
* Rather than re-determining which field was field[0], just force
* 'em all negative.
Add a new GUC variable called "IntervalStyle" that decouples interval output from DateStyle, and create a new interval style that produces output matching the SQL standard (at least for interval values that fall within the standard's restrictions). IntervalStyle is also used to resolve the conflict between the standard and traditional Postgres rules for interpreting negative interval input. Ron Mayer
2008-11-09 01:28:35 +01:00
*/
if (*fsec > 0)
*fsec = -(*fsec);
if (tm->tm_sec > 0)
tm->tm_sec = -tm->tm_sec;
if (tm->tm_min > 0)
tm->tm_min = -tm->tm_min;
if (tm->tm_hour > 0)
tm->tm_hour = -tm->tm_hour;
if (tm->tm_mday > 0)
tm->tm_mday = -tm->tm_mday;
if (tm->tm_mon > 0)
tm->tm_mon = -tm->tm_mon;
if (tm->tm_year > 0)
tm->tm_year = -tm->tm_year;
}
}
/* finally, AGO negates everything */
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
if (is_before)
{
*fsec = -(*fsec);
Remove more extraneous parentheses in date/time functions.
2005-05-24 04:09:45 +02:00
tm->tm_sec = -tm->tm_sec;
tm->tm_min = -tm->tm_min;
tm->tm_hour = -tm->tm_hour;
tm->tm_mday = -tm->tm_mday;
tm->tm_mon = -tm->tm_mon;
tm->tm_year = -tm->tm_year;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
}
Date/Time updates from Thomas...
1997-03-15 00:21:12 +01:00
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
return 0;
}
Date/Time updates from Thomas...
1997-03-15 00:21:12 +01:00
Add support for input and output of interval values formatted per ISO 8601; specifically, we can input either the "format with designators" or the "alternative format", and we can output the former when IntervalStyle is set to iso_8601. Ron Mayer
2008-11-11 03:42:33 +01:00
/*
* Helper functions to avoid duplicated code in DecodeISO8601Interval.
*
* Parse a decimal value and break it into integer and fractional parts.
* Returns 0 or DTERR code.
*/
static int
ParseISO8601Number(char *str, char **endptr, int *ipart, double *fpart)
{
double val;
if (!(isdigit((unsigned char) *str) || *str == '-' || *str == '.'))
return DTERR_BAD_FORMAT;
errno = 0;
val = strtod(str, endptr);
/* did we not see anything that looks like a double? */
if (*endptr == str || errno != 0)
return DTERR_BAD_FORMAT;
/* watch out for overflow */
if (val < INT_MIN || val > INT_MAX)
return DTERR_FIELD_OVERFLOW;
/* be very sure we truncate towards zero (cf dtrunc()) */
if (val >= 0)
*ipart = (int) floor(val);
else
*ipart = (int) -floor(-val);
*fpart = val - *ipart;
return 0;
}
/*
* Determine number of integral digits in a valid ISO 8601 number field
* (we should ignore sign and any fraction part)
*/
static int
ISO8601IntegerWidth(char *fieldstart)
{
/* We might have had a leading '-' */
if (*fieldstart == '-')
fieldstart++;
return strspn(fieldstart, "0123456789");
}
/* DecodeISO8601Interval()
8.4 pgindent run, with new combined Linux/FreeBSD/MinGW typedef list provided by Andrew.
2009-06-11 16:49:15 +02:00
* Decode an ISO 8601 time interval of the "format with designators"
* (section 4.4.3.2) or "alternative format" (section 4.4.3.3)
* Examples: P1D for 1 day
* PT1H for 1 hour
* P2Y6M7DT1H30M for 2 years, 6 months, 7 days 1 hour 30 min
* P0002-06-07T01:30:00 the same value in alternative format
Add support for input and output of interval values formatted per ISO 8601; specifically, we can input either the "format with designators" or the "alternative format", and we can output the former when IntervalStyle is set to iso_8601. Ron Mayer
2008-11-11 03:42:33 +01:00
*
* Returns 0 if successful, DTERR code if bogus input detected.
* Note: error code should be DTERR_BAD_FORMAT if input doesn't look like
* ISO8601, otherwise this could cause unexpected error messages.
* dtype, tm, fsec are output parameters.
*
8.4 pgindent run, with new combined Linux/FreeBSD/MinGW typedef list provided by Andrew.
2009-06-11 16:49:15 +02:00
* A couple exceptions from the spec:
* - a week field ('W') may coexist with other units
* - allows decimals in fields other than the least significant unit.
Add support for input and output of interval values formatted per ISO 8601; specifically, we can input either the "format with designators" or the "alternative format", and we can output the former when IntervalStyle is set to iso_8601. Ron Mayer
2008-11-11 03:42:33 +01:00
*/
int
DecodeISO8601Interval(char *str,
Initial pgindent run with pg_bsd_indent version 2.0. The new indent version includes numerous fixes thanks to Piotr Stefaniak. The main changes visible in this commit are: * Nicer formatting of function-pointer declarations. * No longer unexpectedly removes spaces in expressions using casts, sizeof, or offsetof. * No longer wants to add a space in "struct structname *varname", as well as some similar cases for const- or volatile-qualified pointers. * Declarations using PG_USED_FOR_ASSERTS_ONLY are formatted more nicely. * Fixes bug where comments following declarations were sometimes placed with no space separating them from the code. * Fixes some odd decisions for comments following case labels. * Fixes some cases where comments following code were indented to less than the expected column 33. On the less good side, it now tends to put more whitespace around typedef names that are not listed in typedefs.list. This might encourage us to put more effort into typedef name collection; it's not really a bug in indent itself. There are more changes coming after this round, having to do with comment indentation and alignment of lines appearing within parentheses. I wanted to limit the size of the diffs to something that could be reviewed without one's eyes completely glazing over, so it seemed better to split up the changes as much as practical. Discussion: https://postgr.es/m/E1dAmxK-0006EE-1r@gemulon.postgresql.org Discussion: https://postgr.es/m/30527.1495162840@sss.pgh.pa.us
2017-06-21 20:39:04 +02:00
int *dtype, struct pg_tm *tm, fsec_t *fsec)
Add support for input and output of interval values formatted per ISO 8601; specifically, we can input either the "format with designators" or the "alternative format", and we can output the former when IntervalStyle is set to iso_8601. Ron Mayer
2008-11-11 03:42:33 +01:00
{
8.4 pgindent run, with new combined Linux/FreeBSD/MinGW typedef list provided by Andrew.
2009-06-11 16:49:15 +02:00
bool datepart = true;
bool havefield = false;
Add support for input and output of interval values formatted per ISO 8601; specifically, we can input either the "format with designators" or the "alternative format", and we can output the former when IntervalStyle is set to iso_8601. Ron Mayer
2008-11-11 03:42:33 +01:00
*dtype = DTK_DELTA;
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
ClearPgTm(tm, fsec);
Add support for input and output of interval values formatted per ISO 8601; specifically, we can input either the "format with designators" or the "alternative format", and we can output the former when IntervalStyle is set to iso_8601. Ron Mayer
2008-11-11 03:42:33 +01:00
if (strlen(str) < 2 || str[0] != 'P')
return DTERR_BAD_FORMAT;
str++;
while (*str)
{
8.4 pgindent run, with new combined Linux/FreeBSD/MinGW typedef list provided by Andrew.
2009-06-11 16:49:15 +02:00
char *fieldstart;
int val;
double fval;
char unit;
int dterr;
Add support for input and output of interval values formatted per ISO 8601; specifically, we can input either the "format with designators" or the "alternative format", and we can output the former when IntervalStyle is set to iso_8601. Ron Mayer
2008-11-11 03:42:33 +01:00
8.4 pgindent run, with new combined Linux/FreeBSD/MinGW typedef list provided by Andrew.
2009-06-11 16:49:15 +02:00
if (*str == 'T') /* T indicates the beginning of the time part */
Add support for input and output of interval values formatted per ISO 8601; specifically, we can input either the "format with designators" or the "alternative format", and we can output the former when IntervalStyle is set to iso_8601. Ron Mayer
2008-11-11 03:42:33 +01:00
{
datepart = false;
havefield = false;
str++;
continue;
}
fieldstart = str;
dterr = ParseISO8601Number(str, &str, &val, &fval);
if (dterr)
return dterr;
/*
pgindent run for 9.4 This includes removing tabs after periods in C comments, which was applied to back branches, so this change should not effect backpatching.
2014-05-06 18:12:18 +02:00
* Note: we could step off the end of the string here. Code below
Add support for input and output of interval values formatted per ISO 8601; specifically, we can input either the "format with designators" or the "alternative format", and we can output the former when IntervalStyle is set to iso_8601. Ron Mayer
2008-11-11 03:42:33 +01:00
* *must* exit the loop if unit == '\0'.
*/
unit = *str++;
if (datepart)
{
8.4 pgindent run, with new combined Linux/FreeBSD/MinGW typedef list provided by Andrew.
2009-06-11 16:49:15 +02:00
switch (unit) /* before T: Y M W D */
Add support for input and output of interval values formatted per ISO 8601; specifically, we can input either the "format with designators" or the "alternative format", and we can output the former when IntervalStyle is set to iso_8601. Ron Mayer
2008-11-11 03:42:33 +01:00
{
case 'Y':
tm->tm_year += val;
Use macros for time-based constants, rather than constants.
2011-03-12 15:31:18 +01:00
tm->tm_mon += (fval * MONTHS_PER_YEAR);
Add support for input and output of interval values formatted per ISO 8601; specifically, we can input either the "format with designators" or the "alternative format", and we can output the former when IntervalStyle is set to iso_8601. Ron Mayer
2008-11-11 03:42:33 +01:00
break;
case 'M':
tm->tm_mon += val;
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
AdjustFractDays(fval, tm, fsec, DAYS_PER_MONTH);
Add support for input and output of interval values formatted per ISO 8601; specifically, we can input either the "format with designators" or the "alternative format", and we can output the former when IntervalStyle is set to iso_8601. Ron Mayer
2008-11-11 03:42:33 +01:00
break;
case 'W':
tm->tm_mday += val * 7;
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
AdjustFractDays(fval, tm, fsec, 7);
Add support for input and output of interval values formatted per ISO 8601; specifically, we can input either the "format with designators" or the "alternative format", and we can output the former when IntervalStyle is set to iso_8601. Ron Mayer
2008-11-11 03:42:33 +01:00
break;
case 'D':
tm->tm_mday += val;
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
AdjustFractSeconds(fval, tm, fsec, SECS_PER_DAY);
Add support for input and output of interval values formatted per ISO 8601; specifically, we can input either the "format with designators" or the "alternative format", and we can output the former when IntervalStyle is set to iso_8601. Ron Mayer
2008-11-11 03:42:33 +01:00
break;
8.4 pgindent run, with new combined Linux/FreeBSD/MinGW typedef list provided by Andrew.
2009-06-11 16:49:15 +02:00
case 'T': /* ISO 8601 4.4.3.3 Alternative Format / Basic */
Add support for input and output of interval values formatted per ISO 8601; specifically, we can input either the "format with designators" or the "alternative format", and we can output the former when IntervalStyle is set to iso_8601. Ron Mayer
2008-11-11 03:42:33 +01:00
case '\0':
if (ISO8601IntegerWidth(fieldstart) == 8 && !havefield)
{
tm->tm_year += val / 10000;
8.4 pgindent run, with new combined Linux/FreeBSD/MinGW typedef list provided by Andrew.
2009-06-11 16:49:15 +02:00
tm->tm_mon += (val / 100) % 100;
Add support for input and output of interval values formatted per ISO 8601; specifically, we can input either the "format with designators" or the "alternative format", and we can output the former when IntervalStyle is set to iso_8601. Ron Mayer
2008-11-11 03:42:33 +01:00
tm->tm_mday += val % 100;
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
AdjustFractSeconds(fval, tm, fsec, SECS_PER_DAY);
Add support for input and output of interval values formatted per ISO 8601; specifically, we can input either the "format with designators" or the "alternative format", and we can output the former when IntervalStyle is set to iso_8601. Ron Mayer
2008-11-11 03:42:33 +01:00
if (unit == '\0')
return 0;
datepart = false;
havefield = false;
continue;
}
/* Else fall through to extended alternative format */
Clean up warnings from -Wimplicit-fallthrough. Recent gcc can warn about switch-case fall throughs that are not explicitly labeled as intentional. This seems like a good thing, so clean up the warnings exposed thereby by labeling all such cases with comments that gcc will recognize. In files that already had one or more suitable comments, I generally matched the existing style of those. Otherwise I went with /* FALLTHROUGH */, which is one of the spellings approved at the more-restrictive-than-default level -Wimplicit-fallthrough=4. (At the default level you can also spell it /* FALL ?THRU */, and it's not picky about case. What you can't do is include additional text in the same comment, so some existing comments containing versions of this aren't good enough.) Testing with gcc 8.0.1 (Fedora 28's current version), I found that I also had to put explicit "break"s after elog(ERROR) or ereport(ERROR); apparently, for this purpose gcc doesn't recognize that those don't return. That seems like possibly a gcc bug, but it's fine because in most places we did that anyway; so this amounts to a visit from the style police. Discussion: https://postgr.es/m/15083.1525207729@sss.pgh.pa.us
2018-05-02 01:35:08 +02:00
/* FALLTHROUGH */
8.4 pgindent run, with new combined Linux/FreeBSD/MinGW typedef list provided by Andrew.
2009-06-11 16:49:15 +02:00
case '-': /* ISO 8601 4.4.3.3 Alternative Format,
* Extended */
Add support for input and output of interval values formatted per ISO 8601; specifically, we can input either the "format with designators" or the "alternative format", and we can output the former when IntervalStyle is set to iso_8601. Ron Mayer
2008-11-11 03:42:33 +01:00
if (havefield)
return DTERR_BAD_FORMAT;
tm->tm_year += val;
Use macros for time-based constants, rather than constants.
2011-03-12 15:31:18 +01:00
tm->tm_mon += (fval * MONTHS_PER_YEAR);
Add support for input and output of interval values formatted per ISO 8601; specifically, we can input either the "format with designators" or the "alternative format", and we can output the former when IntervalStyle is set to iso_8601. Ron Mayer
2008-11-11 03:42:33 +01:00
if (unit == '\0')
return 0;
if (unit == 'T')
{
datepart = false;
havefield = false;
continue;
}
dterr = ParseISO8601Number(str, &str, &val, &fval);
if (dterr)
return dterr;
8.4 pgindent run, with new combined Linux/FreeBSD/MinGW typedef list provided by Andrew.
2009-06-11 16:49:15 +02:00
tm->tm_mon += val;
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
AdjustFractDays(fval, tm, fsec, DAYS_PER_MONTH);
Add support for input and output of interval values formatted per ISO 8601; specifically, we can input either the "format with designators" or the "alternative format", and we can output the former when IntervalStyle is set to iso_8601. Ron Mayer
2008-11-11 03:42:33 +01:00
if (*str == '\0')
return 0;
if (*str == 'T')
{
datepart = false;
havefield = false;
continue;
}
if (*str != '-')
return DTERR_BAD_FORMAT;
str++;
8.4 pgindent run, with new combined Linux/FreeBSD/MinGW typedef list provided by Andrew.
2009-06-11 16:49:15 +02:00
Add support for input and output of interval values formatted per ISO 8601; specifically, we can input either the "format with designators" or the "alternative format", and we can output the former when IntervalStyle is set to iso_8601. Ron Mayer
2008-11-11 03:42:33 +01:00
dterr = ParseISO8601Number(str, &str, &val, &fval);
if (dterr)
return dterr;
tm->tm_mday += val;
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
AdjustFractSeconds(fval, tm, fsec, SECS_PER_DAY);
Add support for input and output of interval values formatted per ISO 8601; specifically, we can input either the "format with designators" or the "alternative format", and we can output the former when IntervalStyle is set to iso_8601. Ron Mayer
2008-11-11 03:42:33 +01:00
if (*str == '\0')
return 0;
if (*str == 'T')
{
datepart = false;
havefield = false;
continue;
}
return DTERR_BAD_FORMAT;
default:
/* not a valid date unit suffix */
return DTERR_BAD_FORMAT;
}
}
else
{
8.4 pgindent run, with new combined Linux/FreeBSD/MinGW typedef list provided by Andrew.
2009-06-11 16:49:15 +02:00
switch (unit) /* after T: H M S */
Add support for input and output of interval values formatted per ISO 8601; specifically, we can input either the "format with designators" or the "alternative format", and we can output the former when IntervalStyle is set to iso_8601. Ron Mayer
2008-11-11 03:42:33 +01:00
{
case 'H':
tm->tm_hour += val;
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
AdjustFractSeconds(fval, tm, fsec, SECS_PER_HOUR);
Add support for input and output of interval values formatted per ISO 8601; specifically, we can input either the "format with designators" or the "alternative format", and we can output the former when IntervalStyle is set to iso_8601. Ron Mayer
2008-11-11 03:42:33 +01:00
break;
case 'M':
tm->tm_min += val;
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
AdjustFractSeconds(fval, tm, fsec, SECS_PER_MINUTE);
Add support for input and output of interval values formatted per ISO 8601; specifically, we can input either the "format with designators" or the "alternative format", and we can output the former when IntervalStyle is set to iso_8601. Ron Mayer
2008-11-11 03:42:33 +01:00
break;
case 'S':
tm->tm_sec += val;
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
AdjustFractSeconds(fval, tm, fsec, 1);
Add support for input and output of interval values formatted per ISO 8601; specifically, we can input either the "format with designators" or the "alternative format", and we can output the former when IntervalStyle is set to iso_8601. Ron Mayer
2008-11-11 03:42:33 +01:00
break;
8.4 pgindent run, with new combined Linux/FreeBSD/MinGW typedef list provided by Andrew.
2009-06-11 16:49:15 +02:00
case '\0': /* ISO 8601 4.4.3.3 Alternative Format */
if (ISO8601IntegerWidth(fieldstart) == 6 && !havefield)
Add support for input and output of interval values formatted per ISO 8601; specifically, we can input either the "format with designators" or the "alternative format", and we can output the former when IntervalStyle is set to iso_8601. Ron Mayer
2008-11-11 03:42:33 +01:00
{
tm->tm_hour += val / 10000;
8.4 pgindent run, with new combined Linux/FreeBSD/MinGW typedef list provided by Andrew.
2009-06-11 16:49:15 +02:00
tm->tm_min += (val / 100) % 100;
tm->tm_sec += val % 100;
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
AdjustFractSeconds(fval, tm, fsec, 1);
Add support for input and output of interval values formatted per ISO 8601; specifically, we can input either the "format with designators" or the "alternative format", and we can output the former when IntervalStyle is set to iso_8601. Ron Mayer
2008-11-11 03:42:33 +01:00
return 0;
}
/* Else fall through to extended alternative format */
Clean up warnings from -Wimplicit-fallthrough. Recent gcc can warn about switch-case fall throughs that are not explicitly labeled as intentional. This seems like a good thing, so clean up the warnings exposed thereby by labeling all such cases with comments that gcc will recognize. In files that already had one or more suitable comments, I generally matched the existing style of those. Otherwise I went with /* FALLTHROUGH */, which is one of the spellings approved at the more-restrictive-than-default level -Wimplicit-fallthrough=4. (At the default level you can also spell it /* FALL ?THRU */, and it's not picky about case. What you can't do is include additional text in the same comment, so some existing comments containing versions of this aren't good enough.) Testing with gcc 8.0.1 (Fedora 28's current version), I found that I also had to put explicit "break"s after elog(ERROR) or ereport(ERROR); apparently, for this purpose gcc doesn't recognize that those don't return. That seems like possibly a gcc bug, but it's fine because in most places we did that anyway; so this amounts to a visit from the style police. Discussion: https://postgr.es/m/15083.1525207729@sss.pgh.pa.us
2018-05-02 01:35:08 +02:00
/* FALLTHROUGH */
8.4 pgindent run, with new combined Linux/FreeBSD/MinGW typedef list provided by Andrew.
2009-06-11 16:49:15 +02:00
case ':': /* ISO 8601 4.4.3.3 Alternative Format,
* Extended */
Add support for input and output of interval values formatted per ISO 8601; specifically, we can input either the "format with designators" or the "alternative format", and we can output the former when IntervalStyle is set to iso_8601. Ron Mayer
2008-11-11 03:42:33 +01:00
if (havefield)
return DTERR_BAD_FORMAT;
tm->tm_hour += val;
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
AdjustFractSeconds(fval, tm, fsec, SECS_PER_HOUR);
Add support for input and output of interval values formatted per ISO 8601; specifically, we can input either the "format with designators" or the "alternative format", and we can output the former when IntervalStyle is set to iso_8601. Ron Mayer
2008-11-11 03:42:33 +01:00
if (unit == '\0')
return 0;
8.4 pgindent run, with new combined Linux/FreeBSD/MinGW typedef list provided by Andrew.
2009-06-11 16:49:15 +02:00
Add support for input and output of interval values formatted per ISO 8601; specifically, we can input either the "format with designators" or the "alternative format", and we can output the former when IntervalStyle is set to iso_8601. Ron Mayer
2008-11-11 03:42:33 +01:00
dterr = ParseISO8601Number(str, &str, &val, &fval);
if (dterr)
return dterr;
8.4 pgindent run, with new combined Linux/FreeBSD/MinGW typedef list provided by Andrew.
2009-06-11 16:49:15 +02:00
tm->tm_min += val;
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
AdjustFractSeconds(fval, tm, fsec, SECS_PER_MINUTE);
Add support for input and output of interval values formatted per ISO 8601; specifically, we can input either the "format with designators" or the "alternative format", and we can output the former when IntervalStyle is set to iso_8601. Ron Mayer
2008-11-11 03:42:33 +01:00
if (*str == '\0')
return 0;
if (*str != ':')
return DTERR_BAD_FORMAT;
str++;
8.4 pgindent run, with new combined Linux/FreeBSD/MinGW typedef list provided by Andrew.
2009-06-11 16:49:15 +02:00
Add support for input and output of interval values formatted per ISO 8601; specifically, we can input either the "format with designators" or the "alternative format", and we can output the former when IntervalStyle is set to iso_8601. Ron Mayer
2008-11-11 03:42:33 +01:00
dterr = ParseISO8601Number(str, &str, &val, &fval);
if (dterr)
return dterr;
8.4 pgindent run, with new combined Linux/FreeBSD/MinGW typedef list provided by Andrew.
2009-06-11 16:49:15 +02:00
tm->tm_sec += val;
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
AdjustFractSeconds(fval, tm, fsec, 1);
Add support for input and output of interval values formatted per ISO 8601; specifically, we can input either the "format with designators" or the "alternative format", and we can output the former when IntervalStyle is set to iso_8601. Ron Mayer
2008-11-11 03:42:33 +01:00
if (*str == '\0')
return 0;
return DTERR_BAD_FORMAT;
default:
/* not a valid time unit suffix */
return DTERR_BAD_FORMAT;
}
}
havefield = true;
}
return 0;
}
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
/* DecodeUnits()
* Decode text string using lookup table.
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
*
* This routine recognizes keywords associated with time interval units.
*
* Given string must be lowercased already.
*
* Implement a cache lookup since it is likely that dates
* will be related in format.
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
*/
int
DecodeUnits(int field, char *lowtoken, int *val)
{
int type;
Remove hard-wired lists of timezone abbreviations in favor of providing configuration files that can be altered by a DBA. The australian_timezones GUC setting disappears, replaced by a timezone_abbreviations setting (set this to 'Australia' to get the effect of australian_timezones). The list of zone names defined by default has undergone a bit of cleanup, too. Documentation still needs some work --- in particular, should we fix Table B-4, or just get rid of it? Joachim Wieland, with some editorializing by moi.
2006-07-25 05:51:23 +02:00
const datetkn *tp;
Date/Time updates from Thomas...
1997-03-15 00:21:12 +01:00
Remove hard-wired lists of timezone abbreviations in favor of providing configuration files that can be altered by a DBA. The australian_timezones GUC setting disappears, replaced by a timezone_abbreviations setting (set this to 'Australia' to get the effect of australian_timezones). The list of zone names defined by default has undergone a bit of cleanup, too. Documentation still needs some work --- in particular, should we fix Table B-4, or just get rid of it? Joachim Wieland, with some editorializing by moi.
2006-07-25 05:51:23 +02:00
tp = deltacache[field];
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
/* use strncmp so that we match truncated tokens */
Remove hard-wired lists of timezone abbreviations in favor of providing configuration files that can be altered by a DBA. The australian_timezones GUC setting disappears, replaced by a timezone_abbreviations setting (set this to 'Australia' to get the effect of australian_timezones). The list of zone names defined by default has undergone a bit of cleanup, too. Documentation still needs some work --- in particular, should we fix Table B-4, or just get rid of it? Joachim Wieland, with some editorializing by moi.
2006-07-25 05:51:23 +02:00
if (tp == NULL || strncmp(lowtoken, tp->token, TOKMAXLEN) != 0)
{
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
tp = datebsearch(lowtoken, deltatktbl, szdeltatktbl);
Remove hard-wired lists of timezone abbreviations in favor of providing configuration files that can be altered by a DBA. The australian_timezones GUC setting disappears, replaced by a timezone_abbreviations setting (set this to 'Australia' to get the effect of australian_timezones). The list of zone names defined by default has undergone a bit of cleanup, too. Documentation still needs some work --- in particular, should we fix Table B-4, or just get rid of it? Joachim Wieland, with some editorializing by moi.
2006-07-25 05:51:23 +02:00
}
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
if (tp == NULL)
{
Measure the current transaction time to milliseconds. Define a new function, GetCurrentTransactionStartTimeUsec() to get the time to this precision. Allow now() and timestamp 'now' to use this higher precision result so we now have fractional seconds in this "constant". Add timestamp without time zone type. Move previous timestamp type to timestamp with time zone. Accept another ISO variant for date/time values: yyyy-mm-ddThh:mm:ss (note the "T" separating the day from hours information). Remove 'current' from date/time types; convert to 'now' in input. Separate time and timetz regression tests. Separate timestamp and timestamptz regression test.
2001-09-28 10:09:14 +02:00
type = UNKNOWN_FIELD;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
*val = 0;
}
else
{
Remove hard-wired lists of timezone abbreviations in favor of providing configuration files that can be altered by a DBA. The australian_timezones GUC setting disappears, replaced by a timezone_abbreviations setting (set this to 'Australia' to get the effect of australian_timezones). The list of zone names defined by default has undergone a bit of cleanup, too. Documentation still needs some work --- in particular, should we fix Table B-4, or just get rid of it? Joachim Wieland, with some editorializing by moi.
2006-07-25 05:51:23 +02:00
deltacache[field] = tp;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
type = tp->type;
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
*val = tp->value;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
}
return type;
Initial pgindent run with pg_bsd_indent version 2.0. The new indent version includes numerous fixes thanks to Piotr Stefaniak. The main changes visible in this commit are: * Nicer formatting of function-pointer declarations. * No longer unexpectedly removes spaces in expressions using casts, sizeof, or offsetof. * No longer wants to add a space in "struct structname *varname", as well as some similar cases for const- or volatile-qualified pointers. * Declarations using PG_USED_FOR_ASSERTS_ONLY are formatted more nicely. * Fixes bug where comments following declarations were sometimes placed with no space separating them from the code. * Fixes some odd decisions for comments following case labels. * Fixes some cases where comments following code were indented to less than the expected column 33. On the less good side, it now tends to put more whitespace around typedef names that are not listed in typedefs.list. This might encourage us to put more effort into typedef name collection; it's not really a bug in indent itself. There are more changes coming after this round, having to do with comment indentation and alignment of lines appearing within parentheses. I wanted to limit the size of the diffs to something that could be reviewed without one's eyes completely glazing over, so it seemed better to split up the changes as much as practical. Discussion: https://postgr.es/m/E1dAmxK-0006EE-1r@gemulon.postgresql.org Discussion: https://postgr.es/m/30527.1495162840@sss.pgh.pa.us
2017-06-21 20:39:04 +02:00
} /* DecodeUnits() */
Date/Time updates from Thomas...
1997-03-15 00:21:12 +01:00
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
/*
* Report an error detected by one of the datetime input processing routines.
*
* dterr is the error code, str is the original input string, datatype is
* the name of the datatype we were trying to accept.
*
* Note: it might seem useless to distinguish DTERR_INTERVAL_OVERFLOW and
* DTERR_TZDISP_OVERFLOW from DTERR_FIELD_OVERFLOW, but SQL99 mandates three
* separate SQLSTATE codes, so ...
*/
void
DateTimeParseError(int dterr, const char *str, const char *datatype)
{
switch (dterr)
{
case DTERR_FIELD_OVERFLOW:
ereport(ERROR,
(errcode(ERRCODE_DATETIME_FIELD_OVERFLOW),
errmsg("date/time field value out of range: \"%s\"",
str)));
break;
case DTERR_MD_FIELD_OVERFLOW:
/* <nanny>same as above, but add hint about DateStyle</nanny> */
ereport(ERROR,
(errcode(ERRCODE_DATETIME_FIELD_OVERFLOW),
errmsg("date/time field value out of range: \"%s\"",
str),
Phase 3 of pgindent updates. Don't move parenthesized lines to the left, even if that means they flow past the right margin. By default, BSD indent lines up statement continuation lines that are within parentheses so that they start just to the right of the preceding left parenthesis. However, traditionally, if that resulted in the continuation line extending to the right of the desired right margin, then indent would push it left just far enough to not overrun the margin, if it could do so without making the continuation line start to the left of the current statement indent. That makes for a weird mix of indentations unless one has been completely rigid about never violating the 80-column limit. This behavior has been pretty universally panned by Postgres developers. Hence, disable it with indent's new -lpl switch, so that parenthesized lines are always lined up with the preceding left paren. This patch is much less interesting than the first round of indent changes, but also bulkier, so I thought it best to separate the effects. Discussion: https://postgr.es/m/E1dAmxK-0006EE-1r@gemulon.postgresql.org Discussion: https://postgr.es/m/30527.1495162840@sss.pgh.pa.us
2017-06-21 21:35:54 +02:00
errhint("Perhaps you need a different \"datestyle\" setting.")));
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
break;
case DTERR_INTERVAL_OVERFLOW:
ereport(ERROR,
(errcode(ERRCODE_INTERVAL_FIELD_OVERFLOW),
errmsg("interval field value out of range: \"%s\"",
str)));
break;
case DTERR_TZDISP_OVERFLOW:
ereport(ERROR,
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
(errcode(ERRCODE_INVALID_TIME_ZONE_DISPLACEMENT_VALUE),
errmsg("time zone displacement out of range: \"%s\"",
str)));
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
break;
case DTERR_BAD_FORMAT:
default:
ereport(ERROR,
(errcode(ERRCODE_INVALID_DATETIME_FORMAT),
Message editing: remove gratuitous variations in message wording, standardize terms, add some clarifications, fix some untranslatable attempts at dynamic message building.
2003-09-25 08:58:07 +02:00
errmsg("invalid input syntax for type %s: \"%s\"",
Adjust date/time input parsing code to correctly distinguish the four SQLSTATE error codes required by SQL99 (invalid format, datetime field overflow, interval field overflow, invalid time zone displacement value). Also emit a HINT about DateStyle in cases where it seems appropriate. Per recent gripes.
2003-08-28 01:29:29 +02:00
datatype, str)));
break;
}
}
Date/Time updates from Thomas...
1997-03-15 00:21:12 +01:00
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
/* datebsearch()
* Binary search -- from Knuth (6.2.1) Algorithm B. Special case like this
* is WAY faster than the generic bsearch().
*/
Remove hard-wired lists of timezone abbreviations in favor of providing configuration files that can be altered by a DBA. The australian_timezones GUC setting disappears, replaced by a timezone_abbreviations setting (set this to 'Australia' to get the effect of australian_timezones). The list of zone names defined by default has undergone a bit of cleanup, too. Documentation still needs some work --- in particular, should we fix Table B-4, or just get rid of it? Joachim Wieland, with some editorializing by moi.
2006-07-25 05:51:23 +02:00
static const datetkn *
datebsearch(const char *key, const datetkn *base, int nel)
Date/Time updates from Thomas...
1997-03-15 00:21:12 +01:00
{
Prevent datebsearch() from crashing on base == NULL && nel == 0. Normally nel == 0 works okay because the initial value of "last" will be less than "base"; but if "base" is zero then the calculation wraps around and we have a very large (unsigned) value for "last", so that the loop can be entered and we get a SIGSEGV on a bogus pointer. This is certainly the proximate cause of the recent reports of Windows builds crashing on 'infinity'::timestamp --- evidently, they're either not setting an active timezonetktbl, or setting an empty one. It's not yet clear to me why it's only happening on Windows and not happening on any buildfarm member. But even if that's due to some bug elsewhere, it seems wise for this function to not choke on the powerup values of timezonetktbl/sztimezonetktbl. I also changed the copy of this code in ecpglib, although I am not sure whether it's exposed to a similar hazard. Per report and stack trace from Richard Broersma.
2011-05-11 02:36:22 +02:00
if (nel > 0)
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
{
Prevent datebsearch() from crashing on base == NULL && nel == 0. Normally nel == 0 works okay because the initial value of "last" will be less than "base"; but if "base" is zero then the calculation wraps around and we have a very large (unsigned) value for "last", so that the loop can be entered and we get a SIGSEGV on a bogus pointer. This is certainly the proximate cause of the recent reports of Windows builds crashing on 'infinity'::timestamp --- evidently, they're either not setting an active timezonetktbl, or setting an empty one. It's not yet clear to me why it's only happening on Windows and not happening on any buildfarm member. But even if that's due to some bug elsewhere, it seems wise for this function to not choke on the powerup values of timezonetktbl/sztimezonetktbl. I also changed the copy of this code in ecpglib, although I am not sure whether it's exposed to a similar hazard. Per report and stack trace from Richard Broersma.
2011-05-11 02:36:22 +02:00
const datetkn *last = base + nel - 1,
*position;
int result;
while (last >= base)
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
{
Prevent datebsearch() from crashing on base == NULL && nel == 0. Normally nel == 0 works okay because the initial value of "last" will be less than "base"; but if "base" is zero then the calculation wraps around and we have a very large (unsigned) value for "last", so that the loop can be entered and we get a SIGSEGV on a bogus pointer. This is certainly the proximate cause of the recent reports of Windows builds crashing on 'infinity'::timestamp --- evidently, they're either not setting an active timezonetktbl, or setting an empty one. It's not yet clear to me why it's only happening on Windows and not happening on any buildfarm member. But even if that's due to some bug elsewhere, it seems wise for this function to not choke on the powerup values of timezonetktbl/sztimezonetktbl. I also changed the copy of this code in ecpglib, although I am not sure whether it's exposed to a similar hazard. Per report and stack trace from Richard Broersma.
2011-05-11 02:36:22 +02:00
position = base + ((last - base) >> 1);
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
/* precheck the first character for a bit of extra speed */
result = (int) key[0] - (int) position->token[0];
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
if (result == 0)
Prevent datebsearch() from crashing on base == NULL && nel == 0. Normally nel == 0 works okay because the initial value of "last" will be less than "base"; but if "base" is zero then the calculation wraps around and we have a very large (unsigned) value for "last", so that the loop can be entered and we get a SIGSEGV on a bogus pointer. This is certainly the proximate cause of the recent reports of Windows builds crashing on 'infinity'::timestamp --- evidently, they're either not setting an active timezonetktbl, or setting an empty one. It's not yet clear to me why it's only happening on Windows and not happening on any buildfarm member. But even if that's due to some bug elsewhere, it seems wise for this function to not choke on the powerup values of timezonetktbl/sztimezonetktbl. I also changed the copy of this code in ecpglib, although I am not sure whether it's exposed to a similar hazard. Per report and stack trace from Richard Broersma.
2011-05-11 02:36:22 +02:00
{
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
/* use strncmp so that we match truncated tokens */
Prevent datebsearch() from crashing on base == NULL && nel == 0. Normally nel == 0 works okay because the initial value of "last" will be less than "base"; but if "base" is zero then the calculation wraps around and we have a very large (unsigned) value for "last", so that the loop can be entered and we get a SIGSEGV on a bogus pointer. This is certainly the proximate cause of the recent reports of Windows builds crashing on 'infinity'::timestamp --- evidently, they're either not setting an active timezonetktbl, or setting an empty one. It's not yet clear to me why it's only happening on Windows and not happening on any buildfarm member. But even if that's due to some bug elsewhere, it seems wise for this function to not choke on the powerup values of timezonetktbl/sztimezonetktbl. I also changed the copy of this code in ecpglib, although I am not sure whether it's exposed to a similar hazard. Per report and stack trace from Richard Broersma.
2011-05-11 02:36:22 +02:00
result = strncmp(key, position->token, TOKMAXLEN);
if (result == 0)
return position;
}
if (result < 0)
last = position - 1;
else
base = position + 1;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
}
}
return NULL;
}
Date/Time updates from Thomas...
1997-03-15 00:21:12 +01:00
Fix up some problems in handling of zic-style time zone names in datetime input routines. Remove the former "DecodePosixTimezone" function in favor of letting the zic code handle POSIX-style zone specs (see tzparse()). In particular this means that "PST+3" now means the same as "-03", whereas it used to mean "-11" --- the zone abbreviation is effectively just a noise word in this syntax. Make sure that all named and POSIX-style zone names will be parsed as a single token. Fix long-standing bogosities in printing and input of fractional-hour timezone offsets (since the tzparse() code will accept these, we'd better make 'em work). Also correct an error in the original coding of the zic-zone-name patch: in "timestamp without time zone" input, zone names are supposed to be allowed but ignored, but the coding was such that the zone changed the interpretation anyway.
2006-10-17 23:03:21 +02:00
/* EncodeTimezone()
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
* Copies representation of a numeric timezone offset to str.
*
* Returns a pointer to the new end of string. No NUL terminator is put
* there; callers are responsible for NUL terminating str themselves.
Fix up some problems in handling of zic-style time zone names in datetime input routines. Remove the former "DecodePosixTimezone" function in favor of letting the zic code handle POSIX-style zone specs (see tzparse()). In particular this means that "PST+3" now means the same as "-03", whereas it used to mean "-11" --- the zone abbreviation is effectively just a noise word in this syntax. Make sure that all named and POSIX-style zone names will be parsed as a single token. Fix long-standing bogosities in printing and input of fractional-hour timezone offsets (since the tzparse() code will accept these, we'd better make 'em work). Also correct an error in the original coding of the zic-zone-name patch: in "timestamp without time zone" input, zone names are supposed to be allowed but ignored, but the coding was such that the zone changed the interpretation anyway.
2006-10-17 23:03:21 +02:00
*/
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
static char *
Fix date/time formats for XML Schema output. Pavel Stehule
2007-03-01 15:52:04 +01:00
EncodeTimezone(char *str, int tz, int style)
Fix up some problems in handling of zic-style time zone names in datetime input routines. Remove the former "DecodePosixTimezone" function in favor of letting the zic code handle POSIX-style zone specs (see tzparse()). In particular this means that "PST+3" now means the same as "-03", whereas it used to mean "-11" --- the zone abbreviation is effectively just a noise word in this syntax. Make sure that all named and POSIX-style zone names will be parsed as a single token. Fix long-standing bogosities in printing and input of fractional-hour timezone offsets (since the tzparse() code will accept these, we'd better make 'em work). Also correct an error in the original coding of the zic-zone-name patch: in "timestamp without time zone" input, zone names are supposed to be allowed but ignored, but the coding was such that the zone changed the interpretation anyway.
2006-10-17 23:03:21 +02:00
{
int hour,
min,
sec;
sec = abs(tz);
min = sec / SECS_PER_MINUTE;
sec -= min * SECS_PER_MINUTE;
hour = min / MINS_PER_HOUR;
min -= hour * MINS_PER_HOUR;
/* TZ is negated compared to sign we wish to display ... */
*str++ = (tz <= 0 ? '+' : '-');
if (sec != 0)
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
{
Optimizations for integer to decimal output. Using a lookup table of digit pairs reduces the number of divisions needed, and calculating the length upfront saves some work; these ideas are taken from the code previously committed for floats. David Fetter, reviewed by Kyotaro Horiguchi, Tels, and me. Discussion: https://postgr.es/m/20190924052620.GP31596%40fetter.org
2020-02-01 22:57:14 +01:00
str = pg_ultostr_zeropad(str, hour, 2);
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
*str++ = ':';
Optimizations for integer to decimal output. Using a lookup table of digit pairs reduces the number of divisions needed, and calculating the length upfront saves some work; these ideas are taken from the code previously committed for floats. David Fetter, reviewed by Kyotaro Horiguchi, Tels, and me. Discussion: https://postgr.es/m/20190924052620.GP31596%40fetter.org
2020-02-01 22:57:14 +01:00
str = pg_ultostr_zeropad(str, min, 2);
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
*str++ = ':';
Optimizations for integer to decimal output. Using a lookup table of digit pairs reduces the number of divisions needed, and calculating the length upfront saves some work; these ideas are taken from the code previously committed for floats. David Fetter, reviewed by Kyotaro Horiguchi, Tels, and me. Discussion: https://postgr.es/m/20190924052620.GP31596%40fetter.org
2020-02-01 22:57:14 +01:00
str = pg_ultostr_zeropad(str, sec, 2);
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
}
Fix date/time formats for XML Schema output. Pavel Stehule
2007-03-01 15:52:04 +01:00
else if (min != 0 || style == USE_XSD_DATES)
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
{
Optimizations for integer to decimal output. Using a lookup table of digit pairs reduces the number of divisions needed, and calculating the length upfront saves some work; these ideas are taken from the code previously committed for floats. David Fetter, reviewed by Kyotaro Horiguchi, Tels, and me. Discussion: https://postgr.es/m/20190924052620.GP31596%40fetter.org
2020-02-01 22:57:14 +01:00
str = pg_ultostr_zeropad(str, hour, 2);
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
*str++ = ':';
Optimizations for integer to decimal output. Using a lookup table of digit pairs reduces the number of divisions needed, and calculating the length upfront saves some work; these ideas are taken from the code previously committed for floats. David Fetter, reviewed by Kyotaro Horiguchi, Tels, and me. Discussion: https://postgr.es/m/20190924052620.GP31596%40fetter.org
2020-02-01 22:57:14 +01:00
str = pg_ultostr_zeropad(str, min, 2);
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
}
Fix up some problems in handling of zic-style time zone names in datetime input routines. Remove the former "DecodePosixTimezone" function in favor of letting the zic code handle POSIX-style zone specs (see tzparse()). In particular this means that "PST+3" now means the same as "-03", whereas it used to mean "-11" --- the zone abbreviation is effectively just a noise word in this syntax. Make sure that all named and POSIX-style zone names will be parsed as a single token. Fix long-standing bogosities in printing and input of fractional-hour timezone offsets (since the tzparse() code will accept these, we'd better make 'em work). Also correct an error in the original coding of the zic-zone-name patch: in "timestamp without time zone" input, zone names are supposed to be allowed but ignored, but the coding was such that the zone changed the interpretation anyway.
2006-10-17 23:03:21 +02:00
else
Optimizations for integer to decimal output. Using a lookup table of digit pairs reduces the number of divisions needed, and calculating the length upfront saves some work; these ideas are taken from the code previously committed for floats. David Fetter, reviewed by Kyotaro Horiguchi, Tels, and me. Discussion: https://postgr.es/m/20190924052620.GP31596%40fetter.org
2020-02-01 22:57:14 +01:00
str = pg_ultostr_zeropad(str, hour, 2);
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
return str;
Fix up some problems in handling of zic-style time zone names in datetime input routines. Remove the former "DecodePosixTimezone" function in favor of letting the zic code handle POSIX-style zone specs (see tzparse()). In particular this means that "PST+3" now means the same as "-03", whereas it used to mean "-11" --- the zone abbreviation is effectively just a noise word in this syntax. Make sure that all named and POSIX-style zone names will be parsed as a single token. Fix long-standing bogosities in printing and input of fractional-hour timezone offsets (since the tzparse() code will accept these, we'd better make 'em work). Also correct an error in the original coding of the zic-zone-name patch: in "timestamp without time zone" input, zone names are supposed to be allowed but ignored, but the coding was such that the zone changed the interpretation anyway.
2006-10-17 23:03:21 +02:00
}
Date/Time updates from Thomas...
1997-03-15 00:21:12 +01:00
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
/* EncodeDateOnly()
* Encode date as local time.
*/
Remove the useless and rather inconsistent return values of EncodeDateOnly, EncodeTimeOnly, EncodeDateTime, EncodeInterval. These don't have any good reason to fail, and their callers were mostly not checking anyway.
2009-05-26 04:17:50 +02:00
void
Initial pgindent run with pg_bsd_indent version 2.0. The new indent version includes numerous fixes thanks to Piotr Stefaniak. The main changes visible in this commit are: * Nicer formatting of function-pointer declarations. * No longer unexpectedly removes spaces in expressions using casts, sizeof, or offsetof. * No longer wants to add a space in "struct structname *varname", as well as some similar cases for const- or volatile-qualified pointers. * Declarations using PG_USED_FOR_ASSERTS_ONLY are formatted more nicely. * Fixes bug where comments following declarations were sometimes placed with no space separating them from the code. * Fixes some odd decisions for comments following case labels. * Fixes some cases where comments following code were indented to less than the expected column 33. On the less good side, it now tends to put more whitespace around typedef names that are not listed in typedefs.list. This might encourage us to put more effort into typedef name collection; it's not really a bug in indent itself. There are more changes coming after this round, having to do with comment indentation and alignment of lines appearing within parentheses. I wanted to limit the size of the diffs to something that could be reviewed without one's eyes completely glazing over, so it seemed better to split up the changes as much as practical. Discussion: https://postgr.es/m/E1dAmxK-0006EE-1r@gemulon.postgresql.org Discussion: https://postgr.es/m/30527.1495162840@sss.pgh.pa.us
2017-06-21 20:39:04 +02:00
EncodeDateOnly(struct pg_tm *tm, int style, char *str)
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
{
Remove the useless and rather inconsistent return values of EncodeDateOnly, EncodeTimeOnly, EncodeDateTime, EncodeInterval. These don't have any good reason to fail, and their callers were mostly not checking anyway.
2009-05-26 04:17:50 +02:00
Assert(tm->tm_mon >= 1 && tm->tm_mon <= MONTHS_PER_YEAR);
Fix datetime and abstime conversions to and from date. Bring optional new-storage date and time up to date and test. This new storage format should fix the "Sparc gcc -O2 bug". (Enable new code with USE_NEW_DATE and USE_NEW_TIME in dt.h)
1997-05-16 09:19:50 +02:00
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
switch (style)
{
case USE_ISO_DATES:
Fix date/time formats for XML Schema output. Pavel Stehule
2007-03-01 15:52:04 +01:00
case USE_XSD_DATES:
Measure the current transaction time to milliseconds. Define a new function, GetCurrentTransactionStartTimeUsec() to get the time to this precision. Allow now() and timestamp 'now' to use this higher precision result so we now have fractional seconds in this "constant". Add timestamp without time zone type. Move previous timestamp type to timestamp with time zone. Accept another ISO variant for date/time values: yyyy-mm-ddThh:mm:ss (note the "T" separating the day from hours information). Remove 'current' from date/time types; convert to 'now' in input. Separate time and timetz regression tests. Separate timestamp and timestamptz regression test.
2001-09-28 10:09:14 +02:00
/* compatible with ISO date formats */
Optimizations for integer to decimal output. Using a lookup table of digit pairs reduces the number of divisions needed, and calculating the length upfront saves some work; these ideas are taken from the code previously committed for floats. David Fetter, reviewed by Kyotaro Horiguchi, Tels, and me. Discussion: https://postgr.es/m/20190924052620.GP31596%40fetter.org
2020-02-01 22:57:14 +01:00
str = pg_ultostr_zeropad(str,
Phase 3 of pgindent updates. Don't move parenthesized lines to the left, even if that means they flow past the right margin. By default, BSD indent lines up statement continuation lines that are within parentheses so that they start just to the right of the preceding left parenthesis. However, traditionally, if that resulted in the continuation line extending to the right of the desired right margin, then indent would push it left just far enough to not overrun the margin, if it could do so without making the continuation line start to the left of the current statement indent. That makes for a weird mix of indentations unless one has been completely rigid about never violating the 80-column limit. This behavior has been pretty universally panned by Postgres developers. Hence, disable it with indent's new -lpl switch, so that parenthesized lines are always lined up with the preceding left paren. This patch is much less interesting than the first round of indent changes, but also bulkier, so I thought it best to separate the effects. Discussion: https://postgr.es/m/E1dAmxK-0006EE-1r@gemulon.postgresql.org Discussion: https://postgr.es/m/30527.1495162840@sss.pgh.pa.us
2017-06-21 21:35:54 +02:00
(tm->tm_year > 0) ? tm->tm_year : -(tm->tm_year - 1), 4);
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
*str++ = '-';
Optimizations for integer to decimal output. Using a lookup table of digit pairs reduces the number of divisions needed, and calculating the length upfront saves some work; these ideas are taken from the code previously committed for floats. David Fetter, reviewed by Kyotaro Horiguchi, Tels, and me. Discussion: https://postgr.es/m/20190924052620.GP31596%40fetter.org
2020-02-01 22:57:14 +01:00
str = pg_ultostr_zeropad(str, tm->tm_mon, 2);
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
*str++ = '-';
Optimizations for integer to decimal output. Using a lookup table of digit pairs reduces the number of divisions needed, and calculating the length upfront saves some work; these ideas are taken from the code previously committed for floats. David Fetter, reviewed by Kyotaro Horiguchi, Tels, and me. Discussion: https://postgr.es/m/20190924052620.GP31596%40fetter.org
2020-02-01 22:57:14 +01:00
str = pg_ultostr_zeropad(str, tm->tm_mday, 2);
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
break;
Fix datetime and abstime conversions to and from date. Bring optional new-storage date and time up to date and test. This new storage format should fix the "Sparc gcc -O2 bug". (Enable new code with USE_NEW_DATE and USE_NEW_TIME in dt.h)
1997-05-16 09:19:50 +02:00
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
case USE_SQL_DATES:
Measure the current transaction time to milliseconds. Define a new function, GetCurrentTransactionStartTimeUsec() to get the time to this precision. Allow now() and timestamp 'now' to use this higher precision result so we now have fractional seconds in this "constant". Add timestamp without time zone type. Move previous timestamp type to timestamp with time zone. Accept another ISO variant for date/time values: yyyy-mm-ddThh:mm:ss (note the "T" separating the day from hours information). Remove 'current' from date/time types; convert to 'now' in input. Separate time and timetz regression tests. Separate timestamp and timestamptz regression test.
2001-09-28 10:09:14 +02:00
/* compatible with Oracle/Ingres date formats */
Apply (a somewhat revised version of) Greg Mullane's patch to eliminate heuristic determination of day vs month in date/time input. Add the ability to specify that input is interpreted as yy-mm-dd order (which formerly worked, but only for yy greater than 31). DateStyle's input component now has the preferred spellings DMY, MDY, or YMD; the older keywords European and US are now aliases for the first two of these. Per recent discussions on pgsql-general.
2003-07-29 02:03:19 +02:00
if (DateOrder == DATEORDER_DMY)
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
{
Optimizations for integer to decimal output. Using a lookup table of digit pairs reduces the number of divisions needed, and calculating the length upfront saves some work; these ideas are taken from the code previously committed for floats. David Fetter, reviewed by Kyotaro Horiguchi, Tels, and me. Discussion: https://postgr.es/m/20190924052620.GP31596%40fetter.org
2020-02-01 22:57:14 +01:00
str = pg_ultostr_zeropad(str, tm->tm_mday, 2);
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
*str++ = '/';
Optimizations for integer to decimal output. Using a lookup table of digit pairs reduces the number of divisions needed, and calculating the length upfront saves some work; these ideas are taken from the code previously committed for floats. David Fetter, reviewed by Kyotaro Horiguchi, Tels, and me. Discussion: https://postgr.es/m/20190924052620.GP31596%40fetter.org
2020-02-01 22:57:14 +01:00
str = pg_ultostr_zeropad(str, tm->tm_mon, 2);
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
}
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
else
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
{
Optimizations for integer to decimal output. Using a lookup table of digit pairs reduces the number of divisions needed, and calculating the length upfront saves some work; these ideas are taken from the code previously committed for floats. David Fetter, reviewed by Kyotaro Horiguchi, Tels, and me. Discussion: https://postgr.es/m/20190924052620.GP31596%40fetter.org
2020-02-01 22:57:14 +01:00
str = pg_ultostr_zeropad(str, tm->tm_mon, 2);
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
*str++ = '/';
Optimizations for integer to decimal output. Using a lookup table of digit pairs reduces the number of divisions needed, and calculating the length upfront saves some work; these ideas are taken from the code previously committed for floats. David Fetter, reviewed by Kyotaro Horiguchi, Tels, and me. Discussion: https://postgr.es/m/20190924052620.GP31596%40fetter.org
2020-02-01 22:57:14 +01:00
str = pg_ultostr_zeropad(str, tm->tm_mday, 2);
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
}
*str++ = '/';
Optimizations for integer to decimal output. Using a lookup table of digit pairs reduces the number of divisions needed, and calculating the length upfront saves some work; these ideas are taken from the code previously committed for floats. David Fetter, reviewed by Kyotaro Horiguchi, Tels, and me. Discussion: https://postgr.es/m/20190924052620.GP31596%40fetter.org
2020-02-01 22:57:14 +01:00
str = pg_ultostr_zeropad(str,
Phase 3 of pgindent updates. Don't move parenthesized lines to the left, even if that means they flow past the right margin. By default, BSD indent lines up statement continuation lines that are within parentheses so that they start just to the right of the preceding left parenthesis. However, traditionally, if that resulted in the continuation line extending to the right of the desired right margin, then indent would push it left just far enough to not overrun the margin, if it could do so without making the continuation line start to the left of the current statement indent. That makes for a weird mix of indentations unless one has been completely rigid about never violating the 80-column limit. This behavior has been pretty universally panned by Postgres developers. Hence, disable it with indent's new -lpl switch, so that parenthesized lines are always lined up with the preceding left paren. This patch is much less interesting than the first round of indent changes, but also bulkier, so I thought it best to separate the effects. Discussion: https://postgr.es/m/E1dAmxK-0006EE-1r@gemulon.postgresql.org Discussion: https://postgr.es/m/30527.1495162840@sss.pgh.pa.us
2017-06-21 21:35:54 +02:00
(tm->tm_year > 0) ? tm->tm_year : -(tm->tm_year - 1), 4);
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
break;
Use common parser and encoder for timestamp data type. Remove older date and time code (retain NEW_DATE_CODE and NEW_TIME_CODE). Use common encoder for date and time. Fix datetime +/- timespan math bug.
1997-07-01 02:22:46 +02:00
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
case USE_GERMAN_DATES:
Measure the current transaction time to milliseconds. Define a new function, GetCurrentTransactionStartTimeUsec() to get the time to this precision. Allow now() and timestamp 'now' to use this higher precision result so we now have fractional seconds in this "constant". Add timestamp without time zone type. Move previous timestamp type to timestamp with time zone. Accept another ISO variant for date/time values: yyyy-mm-ddThh:mm:ss (note the "T" separating the day from hours information). Remove 'current' from date/time types; convert to 'now' in input. Separate time and timetz regression tests. Separate timestamp and timestamptz regression test.
2001-09-28 10:09:14 +02:00
/* German-style date format */
Optimizations for integer to decimal output. Using a lookup table of digit pairs reduces the number of divisions needed, and calculating the length upfront saves some work; these ideas are taken from the code previously committed for floats. David Fetter, reviewed by Kyotaro Horiguchi, Tels, and me. Discussion: https://postgr.es/m/20190924052620.GP31596%40fetter.org
2020-02-01 22:57:14 +01:00
str = pg_ultostr_zeropad(str, tm->tm_mday, 2);
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
*str++ = '.';
Optimizations for integer to decimal output. Using a lookup table of digit pairs reduces the number of divisions needed, and calculating the length upfront saves some work; these ideas are taken from the code previously committed for floats. David Fetter, reviewed by Kyotaro Horiguchi, Tels, and me. Discussion: https://postgr.es/m/20190924052620.GP31596%40fetter.org
2020-02-01 22:57:14 +01:00
str = pg_ultostr_zeropad(str, tm->tm_mon, 2);
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
*str++ = '.';
Optimizations for integer to decimal output. Using a lookup table of digit pairs reduces the number of divisions needed, and calculating the length upfront saves some work; these ideas are taken from the code previously committed for floats. David Fetter, reviewed by Kyotaro Horiguchi, Tels, and me. Discussion: https://postgr.es/m/20190924052620.GP31596%40fetter.org
2020-02-01 22:57:14 +01:00
str = pg_ultostr_zeropad(str,
Phase 3 of pgindent updates. Don't move parenthesized lines to the left, even if that means they flow past the right margin. By default, BSD indent lines up statement continuation lines that are within parentheses so that they start just to the right of the preceding left parenthesis. However, traditionally, if that resulted in the continuation line extending to the right of the desired right margin, then indent would push it left just far enough to not overrun the margin, if it could do so without making the continuation line start to the left of the current statement indent. That makes for a weird mix of indentations unless one has been completely rigid about never violating the 80-column limit. This behavior has been pretty universally panned by Postgres developers. Hence, disable it with indent's new -lpl switch, so that parenthesized lines are always lined up with the preceding left paren. This patch is much less interesting than the first round of indent changes, but also bulkier, so I thought it best to separate the effects. Discussion: https://postgr.es/m/E1dAmxK-0006EE-1r@gemulon.postgresql.org Discussion: https://postgr.es/m/30527.1495162840@sss.pgh.pa.us
2017-06-21 21:35:54 +02:00
(tm->tm_year > 0) ? tm->tm_year : -(tm->tm_year - 1), 4);
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
break;
Date/Time updates from Thomas...
1997-03-15 00:21:12 +01:00
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
case USE_POSTGRES_DATES:
default:
Measure the current transaction time to milliseconds. Define a new function, GetCurrentTransactionStartTimeUsec() to get the time to this precision. Allow now() and timestamp 'now' to use this higher precision result so we now have fractional seconds in this "constant". Add timestamp without time zone type. Move previous timestamp type to timestamp with time zone. Accept another ISO variant for date/time values: yyyy-mm-ddThh:mm:ss (note the "T" separating the day from hours information). Remove 'current' from date/time types; convert to 'now' in input. Separate time and timetz regression tests. Separate timestamp and timestamptz regression test.
2001-09-28 10:09:14 +02:00
/* traditional date-only style for Postgres */
Apply (a somewhat revised version of) Greg Mullane's patch to eliminate heuristic determination of day vs month in date/time input. Add the ability to specify that input is interpreted as yy-mm-dd order (which formerly worked, but only for yy greater than 31). DateStyle's input component now has the preferred spellings DMY, MDY, or YMD; the older keywords European and US are now aliases for the first two of these. Per recent discussions on pgsql-general.
2003-07-29 02:03:19 +02:00
if (DateOrder == DATEORDER_DMY)
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
{
Optimizations for integer to decimal output. Using a lookup table of digit pairs reduces the number of divisions needed, and calculating the length upfront saves some work; these ideas are taken from the code previously committed for floats. David Fetter, reviewed by Kyotaro Horiguchi, Tels, and me. Discussion: https://postgr.es/m/20190924052620.GP31596%40fetter.org
2020-02-01 22:57:14 +01:00
str = pg_ultostr_zeropad(str, tm->tm_mday, 2);
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
*str++ = '-';
Optimizations for integer to decimal output. Using a lookup table of digit pairs reduces the number of divisions needed, and calculating the length upfront saves some work; these ideas are taken from the code previously committed for floats. David Fetter, reviewed by Kyotaro Horiguchi, Tels, and me. Discussion: https://postgr.es/m/20190924052620.GP31596%40fetter.org
2020-02-01 22:57:14 +01:00
str = pg_ultostr_zeropad(str, tm->tm_mon, 2);
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
}
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
else
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
{
Optimizations for integer to decimal output. Using a lookup table of digit pairs reduces the number of divisions needed, and calculating the length upfront saves some work; these ideas are taken from the code previously committed for floats. David Fetter, reviewed by Kyotaro Horiguchi, Tels, and me. Discussion: https://postgr.es/m/20190924052620.GP31596%40fetter.org
2020-02-01 22:57:14 +01:00
str = pg_ultostr_zeropad(str, tm->tm_mon, 2);
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
*str++ = '-';
Optimizations for integer to decimal output. Using a lookup table of digit pairs reduces the number of divisions needed, and calculating the length upfront saves some work; these ideas are taken from the code previously committed for floats. David Fetter, reviewed by Kyotaro Horiguchi, Tels, and me. Discussion: https://postgr.es/m/20190924052620.GP31596%40fetter.org
2020-02-01 22:57:14 +01:00
str = pg_ultostr_zeropad(str, tm->tm_mday, 2);
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
}
*str++ = '-';
Optimizations for integer to decimal output. Using a lookup table of digit pairs reduces the number of divisions needed, and calculating the length upfront saves some work; these ideas are taken from the code previously committed for floats. David Fetter, reviewed by Kyotaro Horiguchi, Tels, and me. Discussion: https://postgr.es/m/20190924052620.GP31596%40fetter.org
2020-02-01 22:57:14 +01:00
str = pg_ultostr_zeropad(str,
Phase 3 of pgindent updates. Don't move parenthesized lines to the left, even if that means they flow past the right margin. By default, BSD indent lines up statement continuation lines that are within parentheses so that they start just to the right of the preceding left parenthesis. However, traditionally, if that resulted in the continuation line extending to the right of the desired right margin, then indent would push it left just far enough to not overrun the margin, if it could do so without making the continuation line start to the left of the current statement indent. That makes for a weird mix of indentations unless one has been completely rigid about never violating the 80-column limit. This behavior has been pretty universally panned by Postgres developers. Hence, disable it with indent's new -lpl switch, so that parenthesized lines are always lined up with the preceding left paren. This patch is much less interesting than the first round of indent changes, but also bulkier, so I thought it best to separate the effects. Discussion: https://postgr.es/m/E1dAmxK-0006EE-1r@gemulon.postgresql.org Discussion: https://postgr.es/m/30527.1495162840@sss.pgh.pa.us
2017-06-21 21:35:54 +02:00
(tm->tm_year > 0) ? tm->tm_year : -(tm->tm_year - 1), 4);
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
break;
}
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
if (tm->tm_year <= 0)
{
memcpy(str, " BC", 3); /* Don't copy NUL */
str += 3;
}
*str = '\0';
Remove the useless and rather inconsistent return values of EncodeDateOnly, EncodeTimeOnly, EncodeDateTime, EncodeInterval. These don't have any good reason to fail, and their callers were mostly not checking anyway.
2009-05-26 04:17:50 +02:00
}
Date/Time updates from Thomas...
1997-03-15 00:21:12 +01:00
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
/* EncodeTimeOnly()
* Encode time fields only.
Improve EncodeDateTime and EncodeTimeOnly APIs Use an explicit argument to tell whether to include the time zone in the output, rather than using some undocumented pointer magic.
2012-03-14 22:03:34 +01:00
*
* tm and fsec are the value to encode, print_tz determines whether to include
* a time zone (the difference between time and timetz types), tz is the
* numeric time zone offset, style is the date style, str is where to write the
* output.
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
*/
Remove the useless and rather inconsistent return values of EncodeDateOnly, EncodeTimeOnly, EncodeDateTime, EncodeInterval. These don't have any good reason to fail, and their callers were mostly not checking anyway.
2009-05-26 04:17:50 +02:00
void
Initial pgindent run with pg_bsd_indent version 2.0. The new indent version includes numerous fixes thanks to Piotr Stefaniak. The main changes visible in this commit are: * Nicer formatting of function-pointer declarations. * No longer unexpectedly removes spaces in expressions using casts, sizeof, or offsetof. * No longer wants to add a space in "struct structname *varname", as well as some similar cases for const- or volatile-qualified pointers. * Declarations using PG_USED_FOR_ASSERTS_ONLY are formatted more nicely. * Fixes bug where comments following declarations were sometimes placed with no space separating them from the code. * Fixes some odd decisions for comments following case labels. * Fixes some cases where comments following code were indented to less than the expected column 33. On the less good side, it now tends to put more whitespace around typedef names that are not listed in typedefs.list. This might encourage us to put more effort into typedef name collection; it's not really a bug in indent itself. There are more changes coming after this round, having to do with comment indentation and alignment of lines appearing within parentheses. I wanted to limit the size of the diffs to something that could be reviewed without one's eyes completely glazing over, so it seemed better to split up the changes as much as practical. Discussion: https://postgr.es/m/E1dAmxK-0006EE-1r@gemulon.postgresql.org Discussion: https://postgr.es/m/30527.1495162840@sss.pgh.pa.us
2017-06-21 20:39:04 +02:00
EncodeTimeOnly(struct pg_tm *tm, fsec_t fsec, bool print_tz, int tz, int style, char *str)
Date/Time updates from Thomas...
1997-03-15 00:21:12 +01:00
{
Optimizations for integer to decimal output. Using a lookup table of digit pairs reduces the number of divisions needed, and calculating the length upfront saves some work; these ideas are taken from the code previously committed for floats. David Fetter, reviewed by Kyotaro Horiguchi, Tels, and me. Discussion: https://postgr.es/m/20190924052620.GP31596%40fetter.org
2020-02-01 22:57:14 +01:00
str = pg_ultostr_zeropad(str, tm->tm_hour, 2);
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
*str++ = ':';
Optimizations for integer to decimal output. Using a lookup table of digit pairs reduces the number of divisions needed, and calculating the length upfront saves some work; these ideas are taken from the code previously committed for floats. David Fetter, reviewed by Kyotaro Horiguchi, Tels, and me. Discussion: https://postgr.es/m/20190924052620.GP31596%40fetter.org
2020-02-01 22:57:14 +01:00
str = pg_ultostr_zeropad(str, tm->tm_min, 2);
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
*str++ = ':';
str = AppendSeconds(str, tm->tm_sec, fsec, MAX_TIME_PRECISION, true);
Improve EncodeDateTime and EncodeTimeOnly APIs Use an explicit argument to tell whether to include the time zone in the output, rather than using some undocumented pointer magic.
2012-03-14 22:03:34 +01:00
if (print_tz)
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
str = EncodeTimezone(str, tz, style);
*str = '\0';
Remove the useless and rather inconsistent return values of EncodeDateOnly, EncodeTimeOnly, EncodeDateTime, EncodeInterval. These don't have any good reason to fail, and their callers were mostly not checking anyway.
2009-05-26 04:17:50 +02:00
}
Date/Time updates from Thomas...
1997-03-15 00:21:12 +01:00
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
/* EncodeDateTime()
* Encode date and time interpreted as local time.
Improve EncodeDateTime and EncodeTimeOnly APIs Use an explicit argument to tell whether to include the time zone in the output, rather than using some undocumented pointer magic.
2012-03-14 22:03:34 +01:00
*
* tm and fsec are the value to encode, print_tz determines whether to include
* a time zone (the difference between timestamp and timestamptz types), tz is
* the numeric time zone offset, tzn is the textual time zone, which if
* specified will be used instead of tz by some styles, style is the date
* style, str is where to write the output.
*
* Supported date styles:
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
* Postgres - day mon hh:mm:ss yyyy tz
* SQL - mm/dd/yyyy hh:mm:ss.ss tz
* ISO - yyyy-mm-dd hh:mm:ss+/-tz
Rework the date/time parsing to tighten up some cases and to enable other cases which should have worked but did not. Now supports julian day (J2452271), ISO time labels (T040506) and various combinations of spaces and run-togethers of dates, times, and time zones. All regression tests pass, and I have more tests to add after the 7.2 release (don't want to require changes to the ancillary horology result files until after then).
2001-12-29 19:31:48 +01:00
* German - dd.mm.yyyy hh:mm:ss tz
Fix date/time formats for XML Schema output. Pavel Stehule
2007-03-01 15:52:04 +01:00
* XSD - yyyy-mm-ddThh:mm:ss.ss+/-tz
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
*/
Remove the useless and rather inconsistent return values of EncodeDateOnly, EncodeTimeOnly, EncodeDateTime, EncodeInterval. These don't have any good reason to fail, and their callers were mostly not checking anyway.
2009-05-26 04:17:50 +02:00
void
Initial pgindent run with pg_bsd_indent version 2.0. The new indent version includes numerous fixes thanks to Piotr Stefaniak. The main changes visible in this commit are: * Nicer formatting of function-pointer declarations. * No longer unexpectedly removes spaces in expressions using casts, sizeof, or offsetof. * No longer wants to add a space in "struct structname *varname", as well as some similar cases for const- or volatile-qualified pointers. * Declarations using PG_USED_FOR_ASSERTS_ONLY are formatted more nicely. * Fixes bug where comments following declarations were sometimes placed with no space separating them from the code. * Fixes some odd decisions for comments following case labels. * Fixes some cases where comments following code were indented to less than the expected column 33. On the less good side, it now tends to put more whitespace around typedef names that are not listed in typedefs.list. This might encourage us to put more effort into typedef name collection; it's not really a bug in indent itself. There are more changes coming after this round, having to do with comment indentation and alignment of lines appearing within parentheses. I wanted to limit the size of the diffs to something that could be reviewed without one's eyes completely glazing over, so it seemed better to split up the changes as much as practical. Discussion: https://postgr.es/m/E1dAmxK-0006EE-1r@gemulon.postgresql.org Discussion: https://postgr.es/m/30527.1495162840@sss.pgh.pa.us
2017-06-21 20:39:04 +02:00
EncodeDateTime(struct pg_tm *tm, fsec_t fsec, bool print_tz, int tz, const char *tzn, int style, char *str)
Date/Time updates from Thomas...
1997-03-15 00:21:12 +01:00
{
Fix up some problems in handling of zic-style time zone names in datetime input routines. Remove the former "DecodePosixTimezone" function in favor of letting the zic code handle POSIX-style zone specs (see tzparse()). In particular this means that "PST+3" now means the same as "-03", whereas it used to mean "-11" --- the zone abbreviation is effectively just a noise word in this syntax. Make sure that all named and POSIX-style zone names will be parsed as a single token. Fix long-standing bogosities in printing and input of fractional-hour timezone offsets (since the tzparse() code will accept these, we'd better make 'em work). Also correct an error in the original coding of the zic-zone-name patch: in "timestamp without time zone" input, zone names are supposed to be allowed but ignored, but the coding was such that the zone changed the interpretation anyway.
2006-10-17 23:03:21 +02:00
int day;
pgindent run.
2002-09-04 22:31:48 +02:00
Add time/date macros for code clarity: #define DAYS_PER_YEAR 365.25 #define MONTHS_PER_YEAR 12 #define DAYS_PER_MONTH 30 #define HOURS_PER_DAY 24
2005-07-21 05:56:25 +02:00
Assert(tm->tm_mon >= 1 && tm->tm_mon <= MONTHS_PER_YEAR);
Date/Time updates from Thomas...
1997-03-15 00:21:12 +01:00
Improve EncodeDateTime and EncodeTimeOnly APIs Use an explicit argument to tell whether to include the time zone in the output, rather than using some undocumented pointer magic.
2012-03-14 22:03:34 +01:00
/*
* Negative tm_isdst means we have no valid time zone translation.
*/
if (tm->tm_isdst < 0)
print_tz = false;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
switch (style)
{
case USE_ISO_DATES:
Fix date/time formats for XML Schema output. Pavel Stehule
2007-03-01 15:52:04 +01:00
case USE_XSD_DATES:
Support alternate storage scheme of 64-bit integer for date/time types. Use "--enable-integer-datetimes" in configuration to use this rather than the original float8 storage. I would recommend the integer-based storage for any platform on which it is available. We perhaps should make this the default for the production release. Change timezone(timestamptz) results to return timestamp rather than a character string. Formerly, we didn't have a way to represent timestamps with an explicit time zone other than freezing the info into a string. Now, we can reasonably omit the explicit time zone from the result and return a timestamp with values appropriate for the specified time zone. Much cleaner, and if you need the time zone in the result you can put it into a character string pretty easily anyway. Allow fractional seconds in date/time types even for dates prior to 1BC. Limit timestamp data types to 6 decimal places of precision. Just right for a micro-second storage of int8 date/time types, and reduces the number of places ad-hoc rounding was occuring for the float8-based types. Use lookup tables for precision/rounding calculations for timestamp and interval types. Formerly used pow() to calculate the desired value but with a more limited range there is no reason to not type in a lookup table. Should be *much* better performance, though formerly there were some optimizations to help minimize the number of times pow() was called. Define a HAVE_INT64_TIMESTAMP variable. Based on the configure option "--enable-integer-datetimes" and the existing internal INT64_IS_BUSTED. Add explicit date/interval operators and functions for addition and subtraction. Formerly relied on implicit type promotion from date to timestamp with time zone. Change timezone conversion functions for the timetz type from "timetz()" to "timezone()". This is consistant with other time zone coersion functions for other types. Bump the catalog version to 200204201. Fix up regression tests to reflect changes in fractional seconds representation for date/times in BC eras. All regression tests pass on my Linux box.
2002-04-21 21:52:18 +02:00
/* Compatible with ISO-8601 date formats */
Optimizations for integer to decimal output. Using a lookup table of digit pairs reduces the number of divisions needed, and calculating the length upfront saves some work; these ideas are taken from the code previously committed for floats. David Fetter, reviewed by Kyotaro Horiguchi, Tels, and me. Discussion: https://postgr.es/m/20190924052620.GP31596%40fetter.org
2020-02-01 22:57:14 +01:00
str = pg_ultostr_zeropad(str,
Phase 3 of pgindent updates. Don't move parenthesized lines to the left, even if that means they flow past the right margin. By default, BSD indent lines up statement continuation lines that are within parentheses so that they start just to the right of the preceding left parenthesis. However, traditionally, if that resulted in the continuation line extending to the right of the desired right margin, then indent would push it left just far enough to not overrun the margin, if it could do so without making the continuation line start to the left of the current statement indent. That makes for a weird mix of indentations unless one has been completely rigid about never violating the 80-column limit. This behavior has been pretty universally panned by Postgres developers. Hence, disable it with indent's new -lpl switch, so that parenthesized lines are always lined up with the preceding left paren. This patch is much less interesting than the first round of indent changes, but also bulkier, so I thought it best to separate the effects. Discussion: https://postgr.es/m/E1dAmxK-0006EE-1r@gemulon.postgresql.org Discussion: https://postgr.es/m/30527.1495162840@sss.pgh.pa.us
2017-06-21 21:35:54 +02:00
(tm->tm_year > 0) ? tm->tm_year : -(tm->tm_year - 1), 4);
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
*str++ = '-';
Optimizations for integer to decimal output. Using a lookup table of digit pairs reduces the number of divisions needed, and calculating the length upfront saves some work; these ideas are taken from the code previously committed for floats. David Fetter, reviewed by Kyotaro Horiguchi, Tels, and me. Discussion: https://postgr.es/m/20190924052620.GP31596%40fetter.org
2020-02-01 22:57:14 +01:00
str = pg_ultostr_zeropad(str, tm->tm_mon, 2);
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
*str++ = '-';
Optimizations for integer to decimal output. Using a lookup table of digit pairs reduces the number of divisions needed, and calculating the length upfront saves some work; these ideas are taken from the code previously committed for floats. David Fetter, reviewed by Kyotaro Horiguchi, Tels, and me. Discussion: https://postgr.es/m/20190924052620.GP31596%40fetter.org
2020-02-01 22:57:14 +01:00
str = pg_ultostr_zeropad(str, tm->tm_mday, 2);
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
*str++ = (style == USE_ISO_DATES) ? ' ' : 'T';
Optimizations for integer to decimal output. Using a lookup table of digit pairs reduces the number of divisions needed, and calculating the length upfront saves some work; these ideas are taken from the code previously committed for floats. David Fetter, reviewed by Kyotaro Horiguchi, Tels, and me. Discussion: https://postgr.es/m/20190924052620.GP31596%40fetter.org
2020-02-01 22:57:14 +01:00
str = pg_ultostr_zeropad(str, tm->tm_hour, 2);
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
*str++ = ':';
Optimizations for integer to decimal output. Using a lookup table of digit pairs reduces the number of divisions needed, and calculating the length upfront saves some work; these ideas are taken from the code previously committed for floats. David Fetter, reviewed by Kyotaro Horiguchi, Tels, and me. Discussion: https://postgr.es/m/20190924052620.GP31596%40fetter.org
2020-02-01 22:57:14 +01:00
str = pg_ultostr_zeropad(str, tm->tm_min, 2);
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
*str++ = ':';
str = AppendTimestampSeconds(str, tm, fsec);
Improve EncodeDateTime and EncodeTimeOnly APIs Use an explicit argument to tell whether to include the time zone in the output, rather than using some undocumented pointer magic.
2012-03-14 22:03:34 +01:00
if (print_tz)
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
str = EncodeTimezone(str, tz, style);
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
break;
Date/Time updates from Thomas...
1997-03-15 00:21:12 +01:00
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
case USE_SQL_DATES:
Support alternate storage scheme of 64-bit integer for date/time types. Use "--enable-integer-datetimes" in configuration to use this rather than the original float8 storage. I would recommend the integer-based storage for any platform on which it is available. We perhaps should make this the default for the production release. Change timezone(timestamptz) results to return timestamp rather than a character string. Formerly, we didn't have a way to represent timestamps with an explicit time zone other than freezing the info into a string. Now, we can reasonably omit the explicit time zone from the result and return a timestamp with values appropriate for the specified time zone. Much cleaner, and if you need the time zone in the result you can put it into a character string pretty easily anyway. Allow fractional seconds in date/time types even for dates prior to 1BC. Limit timestamp data types to 6 decimal places of precision. Just right for a micro-second storage of int8 date/time types, and reduces the number of places ad-hoc rounding was occuring for the float8-based types. Use lookup tables for precision/rounding calculations for timestamp and interval types. Formerly used pow() to calculate the desired value but with a more limited range there is no reason to not type in a lookup table. Should be *much* better performance, though formerly there were some optimizations to help minimize the number of times pow() was called. Define a HAVE_INT64_TIMESTAMP variable. Based on the configure option "--enable-integer-datetimes" and the existing internal INT64_IS_BUSTED. Add explicit date/interval operators and functions for addition and subtraction. Formerly relied on implicit type promotion from date to timestamp with time zone. Change timezone conversion functions for the timetz type from "timetz()" to "timezone()". This is consistant with other time zone coersion functions for other types. Bump the catalog version to 200204201. Fix up regression tests to reflect changes in fractional seconds representation for date/times in BC eras. All regression tests pass on my Linux box.
2002-04-21 21:52:18 +02:00
/* Compatible with Oracle/Ingres date formats */
Apply (a somewhat revised version of) Greg Mullane's patch to eliminate heuristic determination of day vs month in date/time input. Add the ability to specify that input is interpreted as yy-mm-dd order (which formerly worked, but only for yy greater than 31). DateStyle's input component now has the preferred spellings DMY, MDY, or YMD; the older keywords European and US are now aliases for the first two of these. Per recent discussions on pgsql-general.
2003-07-29 02:03:19 +02:00
if (DateOrder == DATEORDER_DMY)
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
{
Optimizations for integer to decimal output. Using a lookup table of digit pairs reduces the number of divisions needed, and calculating the length upfront saves some work; these ideas are taken from the code previously committed for floats. David Fetter, reviewed by Kyotaro Horiguchi, Tels, and me. Discussion: https://postgr.es/m/20190924052620.GP31596%40fetter.org
2020-02-01 22:57:14 +01:00
str = pg_ultostr_zeropad(str, tm->tm_mday, 2);
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
*str++ = '/';
Optimizations for integer to decimal output. Using a lookup table of digit pairs reduces the number of divisions needed, and calculating the length upfront saves some work; these ideas are taken from the code previously committed for floats. David Fetter, reviewed by Kyotaro Horiguchi, Tels, and me. Discussion: https://postgr.es/m/20190924052620.GP31596%40fetter.org
2020-02-01 22:57:14 +01:00
str = pg_ultostr_zeropad(str, tm->tm_mon, 2);
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
}
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
else
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
{
Optimizations for integer to decimal output. Using a lookup table of digit pairs reduces the number of divisions needed, and calculating the length upfront saves some work; these ideas are taken from the code previously committed for floats. David Fetter, reviewed by Kyotaro Horiguchi, Tels, and me. Discussion: https://postgr.es/m/20190924052620.GP31596%40fetter.org
2020-02-01 22:57:14 +01:00
str = pg_ultostr_zeropad(str, tm->tm_mon, 2);
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
*str++ = '/';
Optimizations for integer to decimal output. Using a lookup table of digit pairs reduces the number of divisions needed, and calculating the length upfront saves some work; these ideas are taken from the code previously committed for floats. David Fetter, reviewed by Kyotaro Horiguchi, Tels, and me. Discussion: https://postgr.es/m/20190924052620.GP31596%40fetter.org
2020-02-01 22:57:14 +01:00
str = pg_ultostr_zeropad(str, tm->tm_mday, 2);
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
}
*str++ = '/';
Optimizations for integer to decimal output. Using a lookup table of digit pairs reduces the number of divisions needed, and calculating the length upfront saves some work; these ideas are taken from the code previously committed for floats. David Fetter, reviewed by Kyotaro Horiguchi, Tels, and me. Discussion: https://postgr.es/m/20190924052620.GP31596%40fetter.org
2020-02-01 22:57:14 +01:00
str = pg_ultostr_zeropad(str,
Phase 3 of pgindent updates. Don't move parenthesized lines to the left, even if that means they flow past the right margin. By default, BSD indent lines up statement continuation lines that are within parentheses so that they start just to the right of the preceding left parenthesis. However, traditionally, if that resulted in the continuation line extending to the right of the desired right margin, then indent would push it left just far enough to not overrun the margin, if it could do so without making the continuation line start to the left of the current statement indent. That makes for a weird mix of indentations unless one has been completely rigid about never violating the 80-column limit. This behavior has been pretty universally panned by Postgres developers. Hence, disable it with indent's new -lpl switch, so that parenthesized lines are always lined up with the preceding left paren. This patch is much less interesting than the first round of indent changes, but also bulkier, so I thought it best to separate the effects. Discussion: https://postgr.es/m/E1dAmxK-0006EE-1r@gemulon.postgresql.org Discussion: https://postgr.es/m/30527.1495162840@sss.pgh.pa.us
2017-06-21 21:35:54 +02:00
(tm->tm_year > 0) ? tm->tm_year : -(tm->tm_year - 1), 4);
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
*str++ = ' ';
Optimizations for integer to decimal output. Using a lookup table of digit pairs reduces the number of divisions needed, and calculating the length upfront saves some work; these ideas are taken from the code previously committed for floats. David Fetter, reviewed by Kyotaro Horiguchi, Tels, and me. Discussion: https://postgr.es/m/20190924052620.GP31596%40fetter.org
2020-02-01 22:57:14 +01:00
str = pg_ultostr_zeropad(str, tm->tm_hour, 2);
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
*str++ = ':';
Optimizations for integer to decimal output. Using a lookup table of digit pairs reduces the number of divisions needed, and calculating the length upfront saves some work; these ideas are taken from the code previously committed for floats. David Fetter, reviewed by Kyotaro Horiguchi, Tels, and me. Discussion: https://postgr.es/m/20190924052620.GP31596%40fetter.org
2020-02-01 22:57:14 +01:00
str = pg_ultostr_zeropad(str, tm->tm_min, 2);
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
*str++ = ':';
str = AppendTimestampSeconds(str, tm, fsec);
Implement precision support for timestamp and time, both with and without time zones. SQL99 spec requires a default of zero (round to seconds) which is set in gram.y as typmod is set in the parse tree. We *could* change to a default of either 6 (for internal compatibility with previous versions) or 2 (for external compatibility with previous versions). Evaluate entries in pg_proc wrt the iscachable attribute for timestamp and other date/time types. Try to recognize cases where side effects like the current time zone setting may have an effect on results to decide whether something is cachable or not.
2001-10-03 07:29:27 +02:00
Work around a subtle portability problem in use of printf %s format. Depending on which spec you read, field widths and precisions in %s may be counted either in bytes or characters. Our code was assuming bytes, which is wrong at least for glibc's implementation, and in any case libc might have a different idea of the prevailing encoding than we do. Hence, for portable results we must avoid using anything more complex than just "%s" unless the string to be printed is known to be all-ASCII. This patch fixes the cases I could find, including the psql formatting failure reported by Hernan Gonzalez. In HEAD only, I also added comments to some places where it appears safe to continue using "%.*s".
2010-05-08 18:39:53 +02:00
/*
Adjust comments about avoiding use of printf's %.*s. My initial impression that glibc was measuring the precision in characters (which is what the Linux man page says it does) was incorrect. It does take the precision to be in bytes, but it also tries to truncate the string at a character boundary. The bottom line remains the same: it will mess up if the string is not in the encoding it expects, so we need to avoid %.*s anytime there's a significant risk of that. Previous code changes are still good, but adjust the comments to reflect this knowledge. Per research by Hernan Gonzalez.
2010-05-09 04:16:00 +02:00
* Note: the uses of %.*s in this function would be risky if the
* timezone names ever contain non-ASCII characters. However, all
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
* TZ abbreviations in the IANA database are plain ASCII.
Work around a subtle portability problem in use of printf %s format. Depending on which spec you read, field widths and precisions in %s may be counted either in bytes or characters. Our code was assuming bytes, which is wrong at least for glibc's implementation, and in any case libc might have a different idea of the prevailing encoding than we do. Hence, for portable results we must avoid using anything more complex than just "%s" unless the string to be printed is known to be all-ASCII. This patch fixes the cases I could find, including the psql formatting failure reported by Hernan Gonzalez. In HEAD only, I also added comments to some places where it appears safe to continue using "%.*s".
2010-05-08 18:39:53 +02:00
*/
Improve EncodeDateTime and EncodeTimeOnly APIs Use an explicit argument to tell whether to include the time zone in the output, rather than using some undocumented pointer magic.
2012-03-14 22:03:34 +01:00
if (print_tz)
Support alternate storage scheme of 64-bit integer for date/time types. Use "--enable-integer-datetimes" in configuration to use this rather than the original float8 storage. I would recommend the integer-based storage for any platform on which it is available. We perhaps should make this the default for the production release. Change timezone(timestamptz) results to return timestamp rather than a character string. Formerly, we didn't have a way to represent timestamps with an explicit time zone other than freezing the info into a string. Now, we can reasonably omit the explicit time zone from the result and return a timestamp with values appropriate for the specified time zone. Much cleaner, and if you need the time zone in the result you can put it into a character string pretty easily anyway. Allow fractional seconds in date/time types even for dates prior to 1BC. Limit timestamp data types to 6 decimal places of precision. Just right for a micro-second storage of int8 date/time types, and reduces the number of places ad-hoc rounding was occuring for the float8-based types. Use lookup tables for precision/rounding calculations for timestamp and interval types. Formerly used pow() to calculate the desired value but with a more limited range there is no reason to not type in a lookup table. Should be *much* better performance, though formerly there were some optimizations to help minimize the number of times pow() was called. Define a HAVE_INT64_TIMESTAMP variable. Based on the configure option "--enable-integer-datetimes" and the existing internal INT64_IS_BUSTED. Add explicit date/interval operators and functions for addition and subtraction. Formerly relied on implicit type promotion from date to timestamp with time zone. Change timezone conversion functions for the timetz type from "timetz()" to "timezone()". This is consistant with other time zone coersion functions for other types. Bump the catalog version to 200204201. Fix up regression tests to reflect changes in fractional seconds representation for date/times in BC eras. All regression tests pass on my Linux box.
2002-04-21 21:52:18 +02:00
{
Improve EncodeDateTime and EncodeTimeOnly APIs Use an explicit argument to tell whether to include the time zone in the output, rather than using some undocumented pointer magic.
2012-03-14 22:03:34 +01:00
if (tzn)
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
{
sprintf(str, " %.*s", MAXTZLEN, tzn);
str += strlen(str);
}
Support alternate storage scheme of 64-bit integer for date/time types. Use "--enable-integer-datetimes" in configuration to use this rather than the original float8 storage. I would recommend the integer-based storage for any platform on which it is available. We perhaps should make this the default for the production release. Change timezone(timestamptz) results to return timestamp rather than a character string. Formerly, we didn't have a way to represent timestamps with an explicit time zone other than freezing the info into a string. Now, we can reasonably omit the explicit time zone from the result and return a timestamp with values appropriate for the specified time zone. Much cleaner, and if you need the time zone in the result you can put it into a character string pretty easily anyway. Allow fractional seconds in date/time types even for dates prior to 1BC. Limit timestamp data types to 6 decimal places of precision. Just right for a micro-second storage of int8 date/time types, and reduces the number of places ad-hoc rounding was occuring for the float8-based types. Use lookup tables for precision/rounding calculations for timestamp and interval types. Formerly used pow() to calculate the desired value but with a more limited range there is no reason to not type in a lookup table. Should be *much* better performance, though formerly there were some optimizations to help minimize the number of times pow() was called. Define a HAVE_INT64_TIMESTAMP variable. Based on the configure option "--enable-integer-datetimes" and the existing internal INT64_IS_BUSTED. Add explicit date/interval operators and functions for addition and subtraction. Formerly relied on implicit type promotion from date to timestamp with time zone. Change timezone conversion functions for the timetz type from "timetz()" to "timezone()". This is consistant with other time zone coersion functions for other types. Bump the catalog version to 200204201. Fix up regression tests to reflect changes in fractional seconds representation for date/times in BC eras. All regression tests pass on my Linux box.
2002-04-21 21:52:18 +02:00
else
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
str = EncodeTimezone(str, tz, style);
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
}
break;
Date/Time updates from Thomas...
1997-03-15 00:21:12 +01:00
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
case USE_GERMAN_DATES:
Support alternate storage scheme of 64-bit integer for date/time types. Use "--enable-integer-datetimes" in configuration to use this rather than the original float8 storage. I would recommend the integer-based storage for any platform on which it is available. We perhaps should make this the default for the production release. Change timezone(timestamptz) results to return timestamp rather than a character string. Formerly, we didn't have a way to represent timestamps with an explicit time zone other than freezing the info into a string. Now, we can reasonably omit the explicit time zone from the result and return a timestamp with values appropriate for the specified time zone. Much cleaner, and if you need the time zone in the result you can put it into a character string pretty easily anyway. Allow fractional seconds in date/time types even for dates prior to 1BC. Limit timestamp data types to 6 decimal places of precision. Just right for a micro-second storage of int8 date/time types, and reduces the number of places ad-hoc rounding was occuring for the float8-based types. Use lookup tables for precision/rounding calculations for timestamp and interval types. Formerly used pow() to calculate the desired value but with a more limited range there is no reason to not type in a lookup table. Should be *much* better performance, though formerly there were some optimizations to help minimize the number of times pow() was called. Define a HAVE_INT64_TIMESTAMP variable. Based on the configure option "--enable-integer-datetimes" and the existing internal INT64_IS_BUSTED. Add explicit date/interval operators and functions for addition and subtraction. Formerly relied on implicit type promotion from date to timestamp with time zone. Change timezone conversion functions for the timetz type from "timetz()" to "timezone()". This is consistant with other time zone coersion functions for other types. Bump the catalog version to 200204201. Fix up regression tests to reflect changes in fractional seconds representation for date/times in BC eras. All regression tests pass on my Linux box.
2002-04-21 21:52:18 +02:00
/* German variant on European style */
Optimizations for integer to decimal output. Using a lookup table of digit pairs reduces the number of divisions needed, and calculating the length upfront saves some work; these ideas are taken from the code previously committed for floats. David Fetter, reviewed by Kyotaro Horiguchi, Tels, and me. Discussion: https://postgr.es/m/20190924052620.GP31596%40fetter.org
2020-02-01 22:57:14 +01:00
str = pg_ultostr_zeropad(str, tm->tm_mday, 2);
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
*str++ = '.';
Optimizations for integer to decimal output. Using a lookup table of digit pairs reduces the number of divisions needed, and calculating the length upfront saves some work; these ideas are taken from the code previously committed for floats. David Fetter, reviewed by Kyotaro Horiguchi, Tels, and me. Discussion: https://postgr.es/m/20190924052620.GP31596%40fetter.org
2020-02-01 22:57:14 +01:00
str = pg_ultostr_zeropad(str, tm->tm_mon, 2);
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
*str++ = '.';
Optimizations for integer to decimal output. Using a lookup table of digit pairs reduces the number of divisions needed, and calculating the length upfront saves some work; these ideas are taken from the code previously committed for floats. David Fetter, reviewed by Kyotaro Horiguchi, Tels, and me. Discussion: https://postgr.es/m/20190924052620.GP31596%40fetter.org
2020-02-01 22:57:14 +01:00
str = pg_ultostr_zeropad(str,
Phase 3 of pgindent updates. Don't move parenthesized lines to the left, even if that means they flow past the right margin. By default, BSD indent lines up statement continuation lines that are within parentheses so that they start just to the right of the preceding left parenthesis. However, traditionally, if that resulted in the continuation line extending to the right of the desired right margin, then indent would push it left just far enough to not overrun the margin, if it could do so without making the continuation line start to the left of the current statement indent. That makes for a weird mix of indentations unless one has been completely rigid about never violating the 80-column limit. This behavior has been pretty universally panned by Postgres developers. Hence, disable it with indent's new -lpl switch, so that parenthesized lines are always lined up with the preceding left paren. This patch is much less interesting than the first round of indent changes, but also bulkier, so I thought it best to separate the effects. Discussion: https://postgr.es/m/E1dAmxK-0006EE-1r@gemulon.postgresql.org Discussion: https://postgr.es/m/30527.1495162840@sss.pgh.pa.us
2017-06-21 21:35:54 +02:00
(tm->tm_year > 0) ? tm->tm_year : -(tm->tm_year - 1), 4);
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
*str++ = ' ';
Optimizations for integer to decimal output. Using a lookup table of digit pairs reduces the number of divisions needed, and calculating the length upfront saves some work; these ideas are taken from the code previously committed for floats. David Fetter, reviewed by Kyotaro Horiguchi, Tels, and me. Discussion: https://postgr.es/m/20190924052620.GP31596%40fetter.org
2020-02-01 22:57:14 +01:00
str = pg_ultostr_zeropad(str, tm->tm_hour, 2);
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
*str++ = ':';
Optimizations for integer to decimal output. Using a lookup table of digit pairs reduces the number of divisions needed, and calculating the length upfront saves some work; these ideas are taken from the code previously committed for floats. David Fetter, reviewed by Kyotaro Horiguchi, Tels, and me. Discussion: https://postgr.es/m/20190924052620.GP31596%40fetter.org
2020-02-01 22:57:14 +01:00
str = pg_ultostr_zeropad(str, tm->tm_min, 2);
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
*str++ = ':';
str = AppendTimestampSeconds(str, tm, fsec);
Implement precision support for timestamp and time, both with and without time zones. SQL99 spec requires a default of zero (round to seconds) which is set in gram.y as typmod is set in the parse tree. We *could* change to a default of either 6 (for internal compatibility with previous versions) or 2 (for external compatibility with previous versions). Evaluate entries in pg_proc wrt the iscachable attribute for timestamp and other date/time types. Try to recognize cases where side effects like the current time zone setting may have an effect on results to decide whether something is cachable or not.
2001-10-03 07:29:27 +02:00
Improve EncodeDateTime and EncodeTimeOnly APIs Use an explicit argument to tell whether to include the time zone in the output, rather than using some undocumented pointer magic.
2012-03-14 22:03:34 +01:00
if (print_tz)
Support alternate storage scheme of 64-bit integer for date/time types. Use "--enable-integer-datetimes" in configuration to use this rather than the original float8 storage. I would recommend the integer-based storage for any platform on which it is available. We perhaps should make this the default for the production release. Change timezone(timestamptz) results to return timestamp rather than a character string. Formerly, we didn't have a way to represent timestamps with an explicit time zone other than freezing the info into a string. Now, we can reasonably omit the explicit time zone from the result and return a timestamp with values appropriate for the specified time zone. Much cleaner, and if you need the time zone in the result you can put it into a character string pretty easily anyway. Allow fractional seconds in date/time types even for dates prior to 1BC. Limit timestamp data types to 6 decimal places of precision. Just right for a micro-second storage of int8 date/time types, and reduces the number of places ad-hoc rounding was occuring for the float8-based types. Use lookup tables for precision/rounding calculations for timestamp and interval types. Formerly used pow() to calculate the desired value but with a more limited range there is no reason to not type in a lookup table. Should be *much* better performance, though formerly there were some optimizations to help minimize the number of times pow() was called. Define a HAVE_INT64_TIMESTAMP variable. Based on the configure option "--enable-integer-datetimes" and the existing internal INT64_IS_BUSTED. Add explicit date/interval operators and functions for addition and subtraction. Formerly relied on implicit type promotion from date to timestamp with time zone. Change timezone conversion functions for the timetz type from "timetz()" to "timezone()". This is consistant with other time zone coersion functions for other types. Bump the catalog version to 200204201. Fix up regression tests to reflect changes in fractional seconds representation for date/times in BC eras. All regression tests pass on my Linux box.
2002-04-21 21:52:18 +02:00
{
Improve EncodeDateTime and EncodeTimeOnly APIs Use an explicit argument to tell whether to include the time zone in the output, rather than using some undocumented pointer magic.
2012-03-14 22:03:34 +01:00
if (tzn)
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
{
sprintf(str, " %.*s", MAXTZLEN, tzn);
str += strlen(str);
}
Support alternate storage scheme of 64-bit integer for date/time types. Use "--enable-integer-datetimes" in configuration to use this rather than the original float8 storage. I would recommend the integer-based storage for any platform on which it is available. We perhaps should make this the default for the production release. Change timezone(timestamptz) results to return timestamp rather than a character string. Formerly, we didn't have a way to represent timestamps with an explicit time zone other than freezing the info into a string. Now, we can reasonably omit the explicit time zone from the result and return a timestamp with values appropriate for the specified time zone. Much cleaner, and if you need the time zone in the result you can put it into a character string pretty easily anyway. Allow fractional seconds in date/time types even for dates prior to 1BC. Limit timestamp data types to 6 decimal places of precision. Just right for a micro-second storage of int8 date/time types, and reduces the number of places ad-hoc rounding was occuring for the float8-based types. Use lookup tables for precision/rounding calculations for timestamp and interval types. Formerly used pow() to calculate the desired value but with a more limited range there is no reason to not type in a lookup table. Should be *much* better performance, though formerly there were some optimizations to help minimize the number of times pow() was called. Define a HAVE_INT64_TIMESTAMP variable. Based on the configure option "--enable-integer-datetimes" and the existing internal INT64_IS_BUSTED. Add explicit date/interval operators and functions for addition and subtraction. Formerly relied on implicit type promotion from date to timestamp with time zone. Change timezone conversion functions for the timetz type from "timetz()" to "timezone()". This is consistant with other time zone coersion functions for other types. Bump the catalog version to 200204201. Fix up regression tests to reflect changes in fractional seconds representation for date/times in BC eras. All regression tests pass on my Linux box.
2002-04-21 21:52:18 +02:00
else
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
str = EncodeTimezone(str, tz, style);
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
}
break;
Date/Time updates from Thomas...
1997-03-15 00:21:12 +01:00
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
case USE_POSTGRES_DATES:
default:
Support alternate storage scheme of 64-bit integer for date/time types. Use "--enable-integer-datetimes" in configuration to use this rather than the original float8 storage. I would recommend the integer-based storage for any platform on which it is available. We perhaps should make this the default for the production release. Change timezone(timestamptz) results to return timestamp rather than a character string. Formerly, we didn't have a way to represent timestamps with an explicit time zone other than freezing the info into a string. Now, we can reasonably omit the explicit time zone from the result and return a timestamp with values appropriate for the specified time zone. Much cleaner, and if you need the time zone in the result you can put it into a character string pretty easily anyway. Allow fractional seconds in date/time types even for dates prior to 1BC. Limit timestamp data types to 6 decimal places of precision. Just right for a micro-second storage of int8 date/time types, and reduces the number of places ad-hoc rounding was occuring for the float8-based types. Use lookup tables for precision/rounding calculations for timestamp and interval types. Formerly used pow() to calculate the desired value but with a more limited range there is no reason to not type in a lookup table. Should be *much* better performance, though formerly there were some optimizations to help minimize the number of times pow() was called. Define a HAVE_INT64_TIMESTAMP variable. Based on the configure option "--enable-integer-datetimes" and the existing internal INT64_IS_BUSTED. Add explicit date/interval operators and functions for addition and subtraction. Formerly relied on implicit type promotion from date to timestamp with time zone. Change timezone conversion functions for the timetz type from "timetz()" to "timezone()". This is consistant with other time zone coersion functions for other types. Bump the catalog version to 200204201. Fix up regression tests to reflect changes in fractional seconds representation for date/times in BC eras. All regression tests pass on my Linux box.
2002-04-21 21:52:18 +02:00
/* Backward-compatible with traditional Postgres abstime dates */
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
day = date2j(tm->tm_year, tm->tm_mon, tm->tm_mday);
tm->tm_wday = j2day(day);
Replace a bunch more uses of strncpy() with safer coding. strncpy() has a well-deserved reputation for being unsafe, so make an effort to get rid of nearly all occurrences in HEAD. A large fraction of the remaining uses were passing length less than or equal to the known strlen() of the source, in which case no null-padding can occur and the behavior is equivalent to memcpy(), though doubtless slower and certainly harder to reason about. So just use memcpy() in these cases. In other cases, use either StrNCpy() or strlcpy() as appropriate (depending on whether padding to the full length of the destination buffer seems useful). I left a few strncpy() calls alone in the src/timezone/ code, to keep it in sync with upstream (the IANA tzcode distribution). There are also a few such calls in ecpg that could possibly do with more analysis. AFAICT, none of these changes are more than cosmetic, except for the four occurrences in fe-secure-openssl.c, which are in fact buggy: an overlength source leads to a non-null-terminated destination buffer and ensuing misbehavior. These don't seem like security issues, first because no stack clobber is possible and second because if your values of sslcert etc are coming from untrusted sources then you've got problems way worse than this. Still, it's undesirable to have unpredictable behavior for overlength inputs, so back-patch those four changes to all active branches.
2015-01-24 19:05:42 +01:00
memcpy(str, days[tm->tm_wday], 3);
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
str += 3;
*str++ = ' ';
Apply (a somewhat revised version of) Greg Mullane's patch to eliminate heuristic determination of day vs month in date/time input. Add the ability to specify that input is interpreted as yy-mm-dd order (which formerly worked, but only for yy greater than 31). DateStyle's input component now has the preferred spellings DMY, MDY, or YMD; the older keywords European and US are now aliases for the first two of these. Per recent discussions on pgsql-general.
2003-07-29 02:03:19 +02:00
if (DateOrder == DATEORDER_DMY)
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
{
Optimizations for integer to decimal output. Using a lookup table of digit pairs reduces the number of divisions needed, and calculating the length upfront saves some work; these ideas are taken from the code previously committed for floats. David Fetter, reviewed by Kyotaro Horiguchi, Tels, and me. Discussion: https://postgr.es/m/20190924052620.GP31596%40fetter.org
2020-02-01 22:57:14 +01:00
str = pg_ultostr_zeropad(str, tm->tm_mday, 2);
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
*str++ = ' ';
memcpy(str, months[tm->tm_mon - 1], 3);
str += 3;
}
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
else
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
{
memcpy(str, months[tm->tm_mon - 1], 3);
str += 3;
*str++ = ' ';
Optimizations for integer to decimal output. Using a lookup table of digit pairs reduces the number of divisions needed, and calculating the length upfront saves some work; these ideas are taken from the code previously committed for floats. David Fetter, reviewed by Kyotaro Horiguchi, Tels, and me. Discussion: https://postgr.es/m/20190924052620.GP31596%40fetter.org
2020-02-01 22:57:14 +01:00
str = pg_ultostr_zeropad(str, tm->tm_mday, 2);
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
}
*str++ = ' ';
Optimizations for integer to decimal output. Using a lookup table of digit pairs reduces the number of divisions needed, and calculating the length upfront saves some work; these ideas are taken from the code previously committed for floats. David Fetter, reviewed by Kyotaro Horiguchi, Tels, and me. Discussion: https://postgr.es/m/20190924052620.GP31596%40fetter.org
2020-02-01 22:57:14 +01:00
str = pg_ultostr_zeropad(str, tm->tm_hour, 2);
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
*str++ = ':';
Optimizations for integer to decimal output. Using a lookup table of digit pairs reduces the number of divisions needed, and calculating the length upfront saves some work; these ideas are taken from the code previously committed for floats. David Fetter, reviewed by Kyotaro Horiguchi, Tels, and me. Discussion: https://postgr.es/m/20190924052620.GP31596%40fetter.org
2020-02-01 22:57:14 +01:00
str = pg_ultostr_zeropad(str, tm->tm_min, 2);
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
*str++ = ':';
str = AppendTimestampSeconds(str, tm, fsec);
*str++ = ' ';
Optimizations for integer to decimal output. Using a lookup table of digit pairs reduces the number of divisions needed, and calculating the length upfront saves some work; these ideas are taken from the code previously committed for floats. David Fetter, reviewed by Kyotaro Horiguchi, Tels, and me. Discussion: https://postgr.es/m/20190924052620.GP31596%40fetter.org
2020-02-01 22:57:14 +01:00
str = pg_ultostr_zeropad(str,
Phase 3 of pgindent updates. Don't move parenthesized lines to the left, even if that means they flow past the right margin. By default, BSD indent lines up statement continuation lines that are within parentheses so that they start just to the right of the preceding left parenthesis. However, traditionally, if that resulted in the continuation line extending to the right of the desired right margin, then indent would push it left just far enough to not overrun the margin, if it could do so without making the continuation line start to the left of the current statement indent. That makes for a weird mix of indentations unless one has been completely rigid about never violating the 80-column limit. This behavior has been pretty universally panned by Postgres developers. Hence, disable it with indent's new -lpl switch, so that parenthesized lines are always lined up with the preceding left paren. This patch is much less interesting than the first round of indent changes, but also bulkier, so I thought it best to separate the effects. Discussion: https://postgr.es/m/E1dAmxK-0006EE-1r@gemulon.postgresql.org Discussion: https://postgr.es/m/30527.1495162840@sss.pgh.pa.us
2017-06-21 21:35:54 +02:00
(tm->tm_year > 0) ? tm->tm_year : -(tm->tm_year - 1), 4);
Accept an INTERVAL argument for SET TIME ZONE per SQL99. Modified the parser and the SET handlers to use full Node structures rather than simply a character string argument. Implement INTERVAL() YEAR TO MONTH (etc) syntax per SQL99. Does not yet accept the goofy string format that goes along with, but this should be fairly straight forward to fix now as a bug or later as a feature. Implement precision for the INTERVAL() type. Use the typmod mechanism for both of INTERVAL features. Fix the INTERVAL syntax in the parser: opt_interval was in the wrong place. INTERVAL is now a reserved word, otherwise we get reduce/reduce errors. Implement an explicit date_part() function for TIMETZ. Should fix coersion problem with INTERVAL reported by Peter E. Fix up some error messages for date/time types. Use all caps for type names within message. Fix recently introduced side-effect bug disabling 'epoch' as a recognized field for date_part() etc. Reported by Peter E. (??) Bump catalog version number. Rename "microseconds" current transaction time field from ...Msec to ...Usec. Duh! date/time regression tests updated for reference platform, but a few changes will be necessary for others.
2001-10-18 19:30:21 +02:00
Improve EncodeDateTime and EncodeTimeOnly APIs Use an explicit argument to tell whether to include the time zone in the output, rather than using some undocumented pointer magic.
2012-03-14 22:03:34 +01:00
if (print_tz)
Support alternate storage scheme of 64-bit integer for date/time types. Use "--enable-integer-datetimes" in configuration to use this rather than the original float8 storage. I would recommend the integer-based storage for any platform on which it is available. We perhaps should make this the default for the production release. Change timezone(timestamptz) results to return timestamp rather than a character string. Formerly, we didn't have a way to represent timestamps with an explicit time zone other than freezing the info into a string. Now, we can reasonably omit the explicit time zone from the result and return a timestamp with values appropriate for the specified time zone. Much cleaner, and if you need the time zone in the result you can put it into a character string pretty easily anyway. Allow fractional seconds in date/time types even for dates prior to 1BC. Limit timestamp data types to 6 decimal places of precision. Just right for a micro-second storage of int8 date/time types, and reduces the number of places ad-hoc rounding was occuring for the float8-based types. Use lookup tables for precision/rounding calculations for timestamp and interval types. Formerly used pow() to calculate the desired value but with a more limited range there is no reason to not type in a lookup table. Should be *much* better performance, though formerly there were some optimizations to help minimize the number of times pow() was called. Define a HAVE_INT64_TIMESTAMP variable. Based on the configure option "--enable-integer-datetimes" and the existing internal INT64_IS_BUSTED. Add explicit date/interval operators and functions for addition and subtraction. Formerly relied on implicit type promotion from date to timestamp with time zone. Change timezone conversion functions for the timetz type from "timetz()" to "timezone()". This is consistant with other time zone coersion functions for other types. Bump the catalog version to 200204201. Fix up regression tests to reflect changes in fractional seconds representation for date/times in BC eras. All regression tests pass on my Linux box.
2002-04-21 21:52:18 +02:00
{
Improve EncodeDateTime and EncodeTimeOnly APIs Use an explicit argument to tell whether to include the time zone in the output, rather than using some undocumented pointer magic.
2012-03-14 22:03:34 +01:00
if (tzn)
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
{
sprintf(str, " %.*s", MAXTZLEN, tzn);
str += strlen(str);
}
Support alternate storage scheme of 64-bit integer for date/time types. Use "--enable-integer-datetimes" in configuration to use this rather than the original float8 storage. I would recommend the integer-based storage for any platform on which it is available. We perhaps should make this the default for the production release. Change timezone(timestamptz) results to return timestamp rather than a character string. Formerly, we didn't have a way to represent timestamps with an explicit time zone other than freezing the info into a string. Now, we can reasonably omit the explicit time zone from the result and return a timestamp with values appropriate for the specified time zone. Much cleaner, and if you need the time zone in the result you can put it into a character string pretty easily anyway. Allow fractional seconds in date/time types even for dates prior to 1BC. Limit timestamp data types to 6 decimal places of precision. Just right for a micro-second storage of int8 date/time types, and reduces the number of places ad-hoc rounding was occuring for the float8-based types. Use lookup tables for precision/rounding calculations for timestamp and interval types. Formerly used pow() to calculate the desired value but with a more limited range there is no reason to not type in a lookup table. Should be *much* better performance, though formerly there were some optimizations to help minimize the number of times pow() was called. Define a HAVE_INT64_TIMESTAMP variable. Based on the configure option "--enable-integer-datetimes" and the existing internal INT64_IS_BUSTED. Add explicit date/interval operators and functions for addition and subtraction. Formerly relied on implicit type promotion from date to timestamp with time zone. Change timezone conversion functions for the timetz type from "timetz()" to "timezone()". This is consistant with other time zone coersion functions for other types. Bump the catalog version to 200204201. Fix up regression tests to reflect changes in fractional seconds representation for date/times in BC eras. All regression tests pass on my Linux box.
2002-04-21 21:52:18 +02:00
else
Accept an INTERVAL argument for SET TIME ZONE per SQL99. Modified the parser and the SET handlers to use full Node structures rather than simply a character string argument. Implement INTERVAL() YEAR TO MONTH (etc) syntax per SQL99. Does not yet accept the goofy string format that goes along with, but this should be fairly straight forward to fix now as a bug or later as a feature. Implement precision for the INTERVAL() type. Use the typmod mechanism for both of INTERVAL features. Fix the INTERVAL syntax in the parser: opt_interval was in the wrong place. INTERVAL is now a reserved word, otherwise we get reduce/reduce errors. Implement an explicit date_part() function for TIMETZ. Should fix coersion problem with INTERVAL reported by Peter E. Fix up some error messages for date/time types. Use all caps for type names within message. Fix recently introduced side-effect bug disabling 'epoch' as a recognized field for date_part() etc. Reported by Peter E. (??) Bump catalog version number. Rename "microseconds" current transaction time field from ...Msec to ...Usec. Duh! date/time regression tests updated for reference platform, but a few changes will be necessary for others.
2001-10-18 19:30:21 +02:00
{
Support alternate storage scheme of 64-bit integer for date/time types. Use "--enable-integer-datetimes" in configuration to use this rather than the original float8 storage. I would recommend the integer-based storage for any platform on which it is available. We perhaps should make this the default for the production release. Change timezone(timestamptz) results to return timestamp rather than a character string. Formerly, we didn't have a way to represent timestamps with an explicit time zone other than freezing the info into a string. Now, we can reasonably omit the explicit time zone from the result and return a timestamp with values appropriate for the specified time zone. Much cleaner, and if you need the time zone in the result you can put it into a character string pretty easily anyway. Allow fractional seconds in date/time types even for dates prior to 1BC. Limit timestamp data types to 6 decimal places of precision. Just right for a micro-second storage of int8 date/time types, and reduces the number of places ad-hoc rounding was occuring for the float8-based types. Use lookup tables for precision/rounding calculations for timestamp and interval types. Formerly used pow() to calculate the desired value but with a more limited range there is no reason to not type in a lookup table. Should be *much* better performance, though formerly there were some optimizations to help minimize the number of times pow() was called. Define a HAVE_INT64_TIMESTAMP variable. Based on the configure option "--enable-integer-datetimes" and the existing internal INT64_IS_BUSTED. Add explicit date/interval operators and functions for addition and subtraction. Formerly relied on implicit type promotion from date to timestamp with time zone. Change timezone conversion functions for the timetz type from "timetz()" to "timezone()". This is consistant with other time zone coersion functions for other types. Bump the catalog version to 200204201. Fix up regression tests to reflect changes in fractional seconds representation for date/times in BC eras. All regression tests pass on my Linux box.
2002-04-21 21:52:18 +02:00
/*
pgindent run.
2002-09-04 22:31:48 +02:00
* We have a time zone, but no string version. Use the
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* numeric form, but be sure to include a leading space to
* avoid formatting something which would be rejected by
* the date/time parser later. - thomas 2001-10-19
Support alternate storage scheme of 64-bit integer for date/time types. Use "--enable-integer-datetimes" in configuration to use this rather than the original float8 storage. I would recommend the integer-based storage for any platform on which it is available. We perhaps should make this the default for the production release. Change timezone(timestamptz) results to return timestamp rather than a character string. Formerly, we didn't have a way to represent timestamps with an explicit time zone other than freezing the info into a string. Now, we can reasonably omit the explicit time zone from the result and return a timestamp with values appropriate for the specified time zone. Much cleaner, and if you need the time zone in the result you can put it into a character string pretty easily anyway. Allow fractional seconds in date/time types even for dates prior to 1BC. Limit timestamp data types to 6 decimal places of precision. Just right for a micro-second storage of int8 date/time types, and reduces the number of places ad-hoc rounding was occuring for the float8-based types. Use lookup tables for precision/rounding calculations for timestamp and interval types. Formerly used pow() to calculate the desired value but with a more limited range there is no reason to not type in a lookup table. Should be *much* better performance, though formerly there were some optimizations to help minimize the number of times pow() was called. Define a HAVE_INT64_TIMESTAMP variable. Based on the configure option "--enable-integer-datetimes" and the existing internal INT64_IS_BUSTED. Add explicit date/interval operators and functions for addition and subtraction. Formerly relied on implicit type promotion from date to timestamp with time zone. Change timezone conversion functions for the timetz type from "timetz()" to "timezone()". This is consistant with other time zone coersion functions for other types. Bump the catalog version to 200204201. Fix up regression tests to reflect changes in fractional seconds representation for date/times in BC eras. All regression tests pass on my Linux box.
2002-04-21 21:52:18 +02:00
*/
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
*str++ = ' ';
str = EncodeTimezone(str, tz, style);
Accept an INTERVAL argument for SET TIME ZONE per SQL99. Modified the parser and the SET handlers to use full Node structures rather than simply a character string argument. Implement INTERVAL() YEAR TO MONTH (etc) syntax per SQL99. Does not yet accept the goofy string format that goes along with, but this should be fairly straight forward to fix now as a bug or later as a feature. Implement precision for the INTERVAL() type. Use the typmod mechanism for both of INTERVAL features. Fix the INTERVAL syntax in the parser: opt_interval was in the wrong place. INTERVAL is now a reserved word, otherwise we get reduce/reduce errors. Implement an explicit date_part() function for TIMETZ. Should fix coersion problem with INTERVAL reported by Peter E. Fix up some error messages for date/time types. Use all caps for type names within message. Fix recently introduced side-effect bug disabling 'epoch' as a recognized field for date_part() etc. Reported by Peter E. (??) Bump catalog version number. Rename "microseconds" current transaction time field from ...Msec to ...Usec. Duh! date/time regression tests updated for reference platform, but a few changes will be necessary for others.
2001-10-18 19:30:21 +02:00
}
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
}
break;
}
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
if (tm->tm_year <= 0)
{
memcpy(str, " BC", 3); /* Don't copy NUL */
str += 3;
}
*str = '\0';
Cause the format of BC timestamptz output to be 'datetime zone BC' rather than 'datetime BC zone', because the former is accepted by the timestamptz input converter while the latter may not be depending on spacing. This is not a loss of compatibility w.r.t. 7.4 and before, because until very recently there was never a case where we'd output both zone and 'BC'.
2004-07-11 06:57:20 +02:00
}
From: Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov> Subject: [HACKERS] More date time functions Here are some additional patches mostly related to the date and time data types. It includes some type conversion routines to move between the different date types and some other date manipulation routines such as date_part(units,datetime). I noticed Edmund Mergl et al's neat trick for getting function overloading for builtin functions, so started to use that for the date and time stuff. Later, if someone figures out how to get function overloading directly for internal C code, then we can move to that technique. These patches include documentation updates (don't faint!) for the built-in man page. Doesn't yet include mention of timestamp, since I don't know much about it and since it may change a bit to become a _real_ ANSI timestamp which would include parser support for the declaration syntax (what do you think, Dan?). The patches were developed on the 970330 release, but have been rebuilt off of the 970402 release. The first patch below is to get libpq to compile, on my Linux box, but is not related to the rest of the patches and you can choose not to apply that one at this time. Thanks in advance, scrappy!
1997-04-02 20:36:24 +02:00
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
Add a new GUC variable called "IntervalStyle" that decouples interval output from DateStyle, and create a new interval style that produces output matching the SQL standard (at least for interval values that fall within the standard's restrictions). IntervalStyle is also used to resolve the conflict between the standard and traditional Postgres rules for interpreting negative interval input. Ron Mayer
2008-11-09 01:28:35 +01:00
/*
Add support for input and output of interval values formatted per ISO 8601; specifically, we can input either the "format with designators" or the "alternative format", and we can output the former when IntervalStyle is set to iso_8601. Ron Mayer
2008-11-11 03:42:33 +01:00
* Helper functions to avoid duplicated code in EncodeInterval.
Add a new GUC variable called "IntervalStyle" that decouples interval output from DateStyle, and create a new interval style that produces output matching the SQL standard (at least for interval values that fall within the standard's restrictions). IntervalStyle is also used to resolve the conflict between the standard and traditional Postgres rules for interpreting negative interval input. Ron Mayer
2008-11-09 01:28:35 +01:00
*/
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
/* Append an ISO-8601-style interval field, but only if value isn't zero */
static char *
AddISO8601IntPart(char *cp, int value, char units)
Add a new GUC variable called "IntervalStyle" that decouples interval output from DateStyle, and create a new interval style that produces output matching the SQL standard (at least for interval values that fall within the standard's restrictions). IntervalStyle is also used to resolve the conflict between the standard and traditional Postgres rules for interpreting negative interval input. Ron Mayer
2008-11-09 01:28:35 +01:00
{
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
if (value == 0)
return cp;
sprintf(cp, "%d%c", value, units);
return cp + strlen(cp);
Add a new GUC variable called "IntervalStyle" that decouples interval output from DateStyle, and create a new interval style that produces output matching the SQL standard (at least for interval values that fall within the standard's restrictions). IntervalStyle is also used to resolve the conflict between the standard and traditional Postgres rules for interpreting negative interval input. Ron Mayer
2008-11-09 01:28:35 +01:00
}
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
/* Append a postgres-style interval field, but only if value isn't zero */
Add support for input and output of interval values formatted per ISO 8601; specifically, we can input either the "format with designators" or the "alternative format", and we can output the former when IntervalStyle is set to iso_8601. Ron Mayer
2008-11-11 03:42:33 +01:00
static char *
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
AddPostgresIntPart(char *cp, int value, const char *units,
bool *is_zero, bool *is_before)
Add support for input and output of interval values formatted per ISO 8601; specifically, we can input either the "format with designators" or the "alternative format", and we can output the former when IntervalStyle is set to iso_8601. Ron Mayer
2008-11-11 03:42:33 +01:00
{
if (value == 0)
return cp;
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
sprintf(cp, "%s%s%d %s%s",
(!*is_zero) ? " " : "",
(*is_before && value > 0) ? "+" : "",
value,
units,
(value != 1) ? "s" : "");
8.4 pgindent run, with new combined Linux/FreeBSD/MinGW typedef list provided by Andrew.
2009-06-11 16:49:15 +02:00
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
/*
8.4 pgindent run, with new combined Linux/FreeBSD/MinGW typedef list provided by Andrew.
2009-06-11 16:49:15 +02:00
* Each nonzero field sets is_before for (only) the next one. This is a
* tad bizarre but it's how it worked before...
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
*/
*is_before = (value < 0);
Change TRUE/FALSE to true/false The lower case spellings are C and C++ standard and are used in most parts of the PostgreSQL sources. The upper case spellings are only used in some files/modules. So standardize on the standard spellings. The APIs for ICU, Perl, and Windows define their own TRUE and FALSE, so those are left as is when using those APIs. In code comments, we use the lower-case spelling for the C concepts and keep the upper-case spelling for the SQL concepts. Reviewed-by: Michael Paquier <michael.paquier@gmail.com>
2017-08-16 06:22:32 +02:00
*is_zero = false;
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
return cp + strlen(cp);
}
/* Append a verbose-style interval field, but only if value isn't zero */
static char *
AddVerboseIntPart(char *cp, int value, const char *units,
bool *is_zero, bool *is_before)
{
if (value == 0)
return cp;
/* first nonzero value sets is_before */
if (*is_zero)
{
*is_before = (value < 0);
value = abs(value);
}
else if (*is_before)
value = -value;
sprintf(cp, " %d %s%s", value, units, (value == 1) ? "" : "s");
Change TRUE/FALSE to true/false The lower case spellings are C and C++ standard and are used in most parts of the PostgreSQL sources. The upper case spellings are only used in some files/modules. So standardize on the standard spellings. The APIs for ICU, Perl, and Windows define their own TRUE and FALSE, so those are left as is when using those APIs. In code comments, we use the lower-case spelling for the C concepts and keep the upper-case spelling for the SQL concepts. Reviewed-by: Michael Paquier <michael.paquier@gmail.com>
2017-08-16 06:22:32 +02:00
*is_zero = false;
Add support for input and output of interval values formatted per ISO 8601; specifically, we can input either the "format with designators" or the "alternative format", and we can output the former when IntervalStyle is set to iso_8601. Ron Mayer
2008-11-11 03:42:33 +01:00
return cp + strlen(cp);
}
Add a new GUC variable called "IntervalStyle" that decouples interval output from DateStyle, and create a new interval style that produces output matching the SQL standard (at least for interval values that fall within the standard's restrictions). IntervalStyle is also used to resolve the conflict between the standard and traditional Postgres rules for interpreting negative interval input. Ron Mayer
2008-11-09 01:28:35 +01:00
Support alternate storage scheme of 64-bit integer for date/time types. Use "--enable-integer-datetimes" in configuration to use this rather than the original float8 storage. I would recommend the integer-based storage for any platform on which it is available. We perhaps should make this the default for the production release. Change timezone(timestamptz) results to return timestamp rather than a character string. Formerly, we didn't have a way to represent timestamps with an explicit time zone other than freezing the info into a string. Now, we can reasonably omit the explicit time zone from the result and return a timestamp with values appropriate for the specified time zone. Much cleaner, and if you need the time zone in the result you can put it into a character string pretty easily anyway. Allow fractional seconds in date/time types even for dates prior to 1BC. Limit timestamp data types to 6 decimal places of precision. Just right for a micro-second storage of int8 date/time types, and reduces the number of places ad-hoc rounding was occuring for the float8-based types. Use lookup tables for precision/rounding calculations for timestamp and interval types. Formerly used pow() to calculate the desired value but with a more limited range there is no reason to not type in a lookup table. Should be *much* better performance, though formerly there were some optimizations to help minimize the number of times pow() was called. Define a HAVE_INT64_TIMESTAMP variable. Based on the configure option "--enable-integer-datetimes" and the existing internal INT64_IS_BUSTED. Add explicit date/interval operators and functions for addition and subtraction. Formerly relied on implicit type promotion from date to timestamp with time zone. Change timezone conversion functions for the timetz type from "timetz()" to "timezone()". This is consistant with other time zone coersion functions for other types. Bump the catalog version to 200204201. Fix up regression tests to reflect changes in fractional seconds representation for date/times in BC eras. All regression tests pass on my Linux box.
2002-04-21 21:52:18 +02:00
/* EncodeInterval()
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
* Interpret time structure as a delta time and convert to string.
*
* Support "traditional Postgres" and ISO-8601 styles.
* Actually, afaik ISO does not address time interval formatting,
* but this looks similar to the spec for absolute date/time.
* - thomas 1998-04-30
Add a new GUC variable called "IntervalStyle" that decouples interval output from DateStyle, and create a new interval style that produces output matching the SQL standard (at least for interval values that fall within the standard's restrictions). IntervalStyle is also used to resolve the conflict between the standard and traditional Postgres rules for interpreting negative interval input. Ron Mayer
2008-11-09 01:28:35 +01:00
*
* Actually, afaik, ISO 8601 does specify formats for "time
* intervals...[of the]...format with time-unit designators", which
* are pretty ugly. The format looks something like
8.4 pgindent run, with new combined Linux/FreeBSD/MinGW typedef list provided by Andrew.
2009-06-11 16:49:15 +02:00
* P1Y1M1DT1H1M1.12345S
Add a new GUC variable called "IntervalStyle" that decouples interval output from DateStyle, and create a new interval style that produces output matching the SQL standard (at least for interval values that fall within the standard's restrictions). IntervalStyle is also used to resolve the conflict between the standard and traditional Postgres rules for interpreting negative interval input. Ron Mayer
2008-11-09 01:28:35 +01:00
* but useful for exchanging data with computers instead of humans.
* - ron 2003-07-14
*
* And ISO's SQL 2008 standard specifies standards for
* "year-month literal"s (that look like '2-3') and
* "day-time literal"s (that look like ('4 5:6:7')
Add conversion from datetime to time data type. Rename date+time conversion to datetime to ensure less than 16 characters in routine name (required to fit in pg_proc table).
1997-10-25 07:16:09 +02:00
*/
Remove the useless and rather inconsistent return values of EncodeDateOnly, EncodeTimeOnly, EncodeDateTime, EncodeInterval. These don't have any good reason to fail, and their callers were mostly not checking anyway.
2009-05-26 04:17:50 +02:00
void
Initial pgindent run with pg_bsd_indent version 2.0. The new indent version includes numerous fixes thanks to Piotr Stefaniak. The main changes visible in this commit are: * Nicer formatting of function-pointer declarations. * No longer unexpectedly removes spaces in expressions using casts, sizeof, or offsetof. * No longer wants to add a space in "struct structname *varname", as well as some similar cases for const- or volatile-qualified pointers. * Declarations using PG_USED_FOR_ASSERTS_ONLY are formatted more nicely. * Fixes bug where comments following declarations were sometimes placed with no space separating them from the code. * Fixes some odd decisions for comments following case labels. * Fixes some cases where comments following code were indented to less than the expected column 33. On the less good side, it now tends to put more whitespace around typedef names that are not listed in typedefs.list. This might encourage us to put more effort into typedef name collection; it's not really a bug in indent itself. There are more changes coming after this round, having to do with comment indentation and alignment of lines appearing within parentheses. I wanted to limit the size of the diffs to something that could be reviewed without one's eyes completely glazing over, so it seemed better to split up the changes as much as practical. Discussion: https://postgr.es/m/E1dAmxK-0006EE-1r@gemulon.postgresql.org Discussion: https://postgr.es/m/30527.1495162840@sss.pgh.pa.us
2017-06-21 20:39:04 +02:00
EncodeInterval(struct pg_tm *tm, fsec_t fsec, int style, char *str)
Add conversion from datetime to time data type. Rename date+time conversion to datetime to ensure less than 16 characters in routine name (required to fit in pg_proc table).
1997-10-25 07:16:09 +02:00
{
Add a new GUC variable called "IntervalStyle" that decouples interval output from DateStyle, and create a new interval style that produces output matching the SQL standard (at least for interval values that fall within the standard's restrictions). IntervalStyle is also used to resolve the conflict between the standard and traditional Postgres rules for interpreting negative interval input. Ron Mayer
2008-11-09 01:28:35 +01:00
char *cp = str;
int year = tm->tm_year;
8.4 pgindent run, with new combined Linux/FreeBSD/MinGW typedef list provided by Andrew.
2009-06-11 16:49:15 +02:00
int mon = tm->tm_mon;
Add a new GUC variable called "IntervalStyle" that decouples interval output from DateStyle, and create a new interval style that produces output matching the SQL standard (at least for interval values that fall within the standard's restrictions). IntervalStyle is also used to resolve the conflict between the standard and traditional Postgres rules for interpreting negative interval input. Ron Mayer
2008-11-09 01:28:35 +01:00
int mday = tm->tm_mday;
int hour = tm->tm_hour;
8.4 pgindent run, with new combined Linux/FreeBSD/MinGW typedef list provided by Andrew.
2009-06-11 16:49:15 +02:00
int min = tm->tm_min;
int sec = tm->tm_sec;
Change TRUE/FALSE to true/false The lower case spellings are C and C++ standard and are used in most parts of the PostgreSQL sources. The upper case spellings are only used in some files/modules. So standardize on the standard spellings. The APIs for ICU, Perl, and Windows define their own TRUE and FALSE, so those are left as is when using those APIs. In code comments, we use the lower-case spelling for the C concepts and keep the upper-case spelling for the SQL concepts. Reviewed-by: Michael Paquier <michael.paquier@gmail.com>
2017-08-16 06:22:32 +02:00
bool is_before = false;
bool is_zero = true;
Add conversion from datetime to time data type. Rename date+time conversion to datetime to ensure less than 16 characters in routine name (required to fit in pg_proc table).
1997-10-25 07:16:09 +02:00
pgindent run. Make it all clean.
2001-03-22 05:01:46 +01:00
/*
* The sign of year and month are guaranteed to match, since they are
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* stored internally as "month". But we'll need to check for is_before and
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
* is_zero when determining the signs of day and hour/minute/seconds
Add a new GUC variable called "IntervalStyle" that decouples interval output from DateStyle, and create a new interval style that produces output matching the SQL standard (at least for interval values that fall within the standard's restrictions). IntervalStyle is also used to resolve the conflict between the standard and traditional Postgres rules for interpreting negative interval input. Ron Mayer
2008-11-09 01:28:35 +01:00
* fields.
Fix INTERVAL output when year/month has different sign as day/hour etc. Previously, all fields were unsigned, with only a trailing "ago" to indicate negative intervals. Now, ISO format does not use "ago", and and the traditional PostgreSQL format has the first numeric field unsigned with "ago" supporting that field. So "1 month - 2 days ago" is two days less than a month in the past. Fix interval arithmetic across daylight savings time boundaries. Previously, most math across boundaries introduced a one hour offset. Allow some date/time functions to return NULL if called with NULL args. Implement functions for AT TIME ZONE support. Support "SAT" as an Australian time zone if USE_AUSTRALIAN_RULES is defined.
2000-11-06 16:57:00 +01:00
*/
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
switch (style)
{
8.4 pgindent run, with new combined Linux/FreeBSD/MinGW typedef list provided by Andrew.
2009-06-11 16:49:15 +02:00
/* SQL Standard interval format */
Add a new GUC variable called "IntervalStyle" that decouples interval output from DateStyle, and create a new interval style that produces output matching the SQL standard (at least for interval values that fall within the standard's restrictions). IntervalStyle is also used to resolve the conflict between the standard and traditional Postgres rules for interpreting negative interval input. Ron Mayer
2008-11-09 01:28:35 +01:00
case INTSTYLE_SQL_STANDARD:
{
8.4 pgindent run, with new combined Linux/FreeBSD/MinGW typedef list provided by Andrew.
2009-06-11 16:49:15 +02:00
bool has_negative = year < 0 || mon < 0 ||
mday < 0 || hour < 0 ||
min < 0 || sec < 0 || fsec < 0;
bool has_positive = year > 0 || mon > 0 ||
mday > 0 || hour > 0 ||
min > 0 || sec > 0 || fsec > 0;
bool has_year_month = year != 0 || mon != 0;
bool has_day_time = mday != 0 || hour != 0 ||
min != 0 || sec != 0 || fsec != 0;
bool has_day = mday != 0;
bool sql_standard_value = !(has_negative && has_positive) &&
!(has_year_month && has_day_time);
Add a new GUC variable called "IntervalStyle" that decouples interval output from DateStyle, and create a new interval style that produces output matching the SQL standard (at least for interval values that fall within the standard's restrictions). IntervalStyle is also used to resolve the conflict between the standard and traditional Postgres rules for interpreting negative interval input. Ron Mayer
2008-11-09 01:28:35 +01:00
/*
* SQL Standard wants only 1 "<sign>" preceding the whole
* interval ... but can't do that if mixed signs.
*/
if (has_negative && sql_standard_value)
{
*cp++ = '-';
year = -year;
8.4 pgindent run, with new combined Linux/FreeBSD/MinGW typedef list provided by Andrew.
2009-06-11 16:49:15 +02:00
mon = -mon;
Add a new GUC variable called "IntervalStyle" that decouples interval output from DateStyle, and create a new interval style that produces output matching the SQL standard (at least for interval values that fall within the standard's restrictions). IntervalStyle is also used to resolve the conflict between the standard and traditional Postgres rules for interpreting negative interval input. Ron Mayer
2008-11-09 01:28:35 +01:00
mday = -mday;
hour = -hour;
8.4 pgindent run, with new combined Linux/FreeBSD/MinGW typedef list provided by Andrew.
2009-06-11 16:49:15 +02:00
min = -min;
sec = -sec;
Add a new GUC variable called "IntervalStyle" that decouples interval output from DateStyle, and create a new interval style that produces output matching the SQL standard (at least for interval values that fall within the standard's restrictions). IntervalStyle is also used to resolve the conflict between the standard and traditional Postgres rules for interpreting negative interval input. Ron Mayer
2008-11-09 01:28:35 +01:00
fsec = -fsec;
}
if (!has_negative && !has_positive)
{
sprintf(cp, "0");
}
else if (!sql_standard_value)
{
/*
8.4 pgindent run, with new combined Linux/FreeBSD/MinGW typedef list provided by Andrew.
2009-06-11 16:49:15 +02:00
* For non sql-standard interval values, force outputting
* the signs to avoid ambiguities with intervals with
* mixed sign components.
Add a new GUC variable called "IntervalStyle" that decouples interval output from DateStyle, and create a new interval style that produces output matching the SQL standard (at least for interval values that fall within the standard's restrictions). IntervalStyle is also used to resolve the conflict between the standard and traditional Postgres rules for interpreting negative interval input. Ron Mayer
2008-11-09 01:28:35 +01:00
*/
8.4 pgindent run, with new combined Linux/FreeBSD/MinGW typedef list provided by Andrew.
2009-06-11 16:49:15 +02:00
char year_sign = (year < 0 || mon < 0) ? '-' : '+';
char day_sign = (mday < 0) ? '-' : '+';
char sec_sign = (hour < 0 || min < 0 ||
sec < 0 || fsec < 0) ? '-' : '+';
Add a new GUC variable called "IntervalStyle" that decouples interval output from DateStyle, and create a new interval style that produces output matching the SQL standard (at least for interval values that fall within the standard's restrictions). IntervalStyle is also used to resolve the conflict between the standard and traditional Postgres rules for interpreting negative interval input. Ron Mayer
2008-11-09 01:28:35 +01:00
Add support for input and output of interval values formatted per ISO 8601; specifically, we can input either the "format with designators" or the "alternative format", and we can output the former when IntervalStyle is set to iso_8601. Ron Mayer
2008-11-11 03:42:33 +01:00
sprintf(cp, "%c%d-%d %c%d %c%d:%02d:",
Add a new GUC variable called "IntervalStyle" that decouples interval output from DateStyle, and create a new interval style that produces output matching the SQL standard (at least for interval values that fall within the standard's restrictions). IntervalStyle is also used to resolve the conflict between the standard and traditional Postgres rules for interpreting negative interval input. Ron Mayer
2008-11-09 01:28:35 +01:00
year_sign, abs(year), abs(mon),
day_sign, abs(mday),
sec_sign, abs(hour), abs(min));
cp += strlen(cp);
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
cp = AppendSeconds(cp, sec, fsec, MAX_INTERVAL_PRECISION, true);
*cp = '\0';
Add a new GUC variable called "IntervalStyle" that decouples interval output from DateStyle, and create a new interval style that produces output matching the SQL standard (at least for interval values that fall within the standard's restrictions). IntervalStyle is also used to resolve the conflict between the standard and traditional Postgres rules for interpreting negative interval input. Ron Mayer
2008-11-09 01:28:35 +01:00
}
else if (has_year_month)
{
sprintf(cp, "%d-%d", year, mon);
}
else if (has_day)
{
Add support for input and output of interval values formatted per ISO 8601; specifically, we can input either the "format with designators" or the "alternative format", and we can output the former when IntervalStyle is set to iso_8601. Ron Mayer
2008-11-11 03:42:33 +01:00
sprintf(cp, "%d %d:%02d:", mday, hour, min);
Add a new GUC variable called "IntervalStyle" that decouples interval output from DateStyle, and create a new interval style that produces output matching the SQL standard (at least for interval values that fall within the standard's restrictions). IntervalStyle is also used to resolve the conflict between the standard and traditional Postgres rules for interpreting negative interval input. Ron Mayer
2008-11-09 01:28:35 +01:00
cp += strlen(cp);
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
cp = AppendSeconds(cp, sec, fsec, MAX_INTERVAL_PRECISION, true);
*cp = '\0';
Add a new GUC variable called "IntervalStyle" that decouples interval output from DateStyle, and create a new interval style that produces output matching the SQL standard (at least for interval values that fall within the standard's restrictions). IntervalStyle is also used to resolve the conflict between the standard and traditional Postgres rules for interpreting negative interval input. Ron Mayer
2008-11-09 01:28:35 +01:00
}
else
{
Add support for input and output of interval values formatted per ISO 8601; specifically, we can input either the "format with designators" or the "alternative format", and we can output the former when IntervalStyle is set to iso_8601. Ron Mayer
2008-11-11 03:42:33 +01:00
sprintf(cp, "%d:%02d:", hour, min);
Add a new GUC variable called "IntervalStyle" that decouples interval output from DateStyle, and create a new interval style that produces output matching the SQL standard (at least for interval values that fall within the standard's restrictions). IntervalStyle is also used to resolve the conflict between the standard and traditional Postgres rules for interpreting negative interval input. Ron Mayer
2008-11-09 01:28:35 +01:00
cp += strlen(cp);
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
cp = AppendSeconds(cp, sec, fsec, MAX_INTERVAL_PRECISION, true);
*cp = '\0';
Add a new GUC variable called "IntervalStyle" that decouples interval output from DateStyle, and create a new interval style that produces output matching the SQL standard (at least for interval values that fall within the standard's restrictions). IntervalStyle is also used to resolve the conflict between the standard and traditional Postgres rules for interpreting negative interval input. Ron Mayer
2008-11-09 01:28:35 +01:00
}
}
break;
8.4 pgindent run, with new combined Linux/FreeBSD/MinGW typedef list provided by Andrew.
2009-06-11 16:49:15 +02:00
/* ISO 8601 "time-intervals by duration only" */
Add support for input and output of interval values formatted per ISO 8601; specifically, we can input either the "format with designators" or the "alternative format", and we can output the former when IntervalStyle is set to iso_8601. Ron Mayer
2008-11-11 03:42:33 +01:00
case INTSTYLE_ISO_8601:
/* special-case zero to avoid printing nothing */
if (year == 0 && mon == 0 && mday == 0 &&
8.4 pgindent run, with new combined Linux/FreeBSD/MinGW typedef list provided by Andrew.
2009-06-11 16:49:15 +02:00
hour == 0 && min == 0 && sec == 0 && fsec == 0)
Add support for input and output of interval values formatted per ISO 8601; specifically, we can input either the "format with designators" or the "alternative format", and we can output the former when IntervalStyle is set to iso_8601. Ron Mayer
2008-11-11 03:42:33 +01:00
{
sprintf(cp, "PT0S");
break;
}
*cp++ = 'P';
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
cp = AddISO8601IntPart(cp, year, 'Y');
8.4 pgindent run, with new combined Linux/FreeBSD/MinGW typedef list provided by Andrew.
2009-06-11 16:49:15 +02:00
cp = AddISO8601IntPart(cp, mon, 'M');
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
cp = AddISO8601IntPart(cp, mday, 'D');
Add support for input and output of interval values formatted per ISO 8601; specifically, we can input either the "format with designators" or the "alternative format", and we can output the former when IntervalStyle is set to iso_8601. Ron Mayer
2008-11-11 03:42:33 +01:00
if (hour != 0 || min != 0 || sec != 0 || fsec != 0)
*cp++ = 'T';
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
cp = AddISO8601IntPart(cp, hour, 'H');
8.4 pgindent run, with new combined Linux/FreeBSD/MinGW typedef list provided by Andrew.
2009-06-11 16:49:15 +02:00
cp = AddISO8601IntPart(cp, min, 'M');
Add support for input and output of interval values formatted per ISO 8601; specifically, we can input either the "format with designators" or the "alternative format", and we can output the former when IntervalStyle is set to iso_8601. Ron Mayer
2008-11-11 03:42:33 +01:00
if (sec != 0 || fsec != 0)
{
if (sec < 0 || fsec < 0)
*cp++ = '-';
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
cp = AppendSeconds(cp, sec, fsec, MAX_INTERVAL_PRECISION, false);
Add support for input and output of interval values formatted per ISO 8601; specifically, we can input either the "format with designators" or the "alternative format", and we can output the former when IntervalStyle is set to iso_8601. Ron Mayer
2008-11-11 03:42:33 +01:00
*cp++ = 'S';
*cp++ = '\0';
}
break;
8.4 pgindent run, with new combined Linux/FreeBSD/MinGW typedef list provided by Andrew.
2009-06-11 16:49:15 +02:00
/* Compatible with postgresql < 8.4 when DateStyle = 'iso' */
Add a new GUC variable called "IntervalStyle" that decouples interval output from DateStyle, and create a new interval style that produces output matching the SQL standard (at least for interval values that fall within the standard's restrictions). IntervalStyle is also used to resolve the conflict between the standard and traditional Postgres rules for interpreting negative interval input. Ron Mayer
2008-11-09 01:28:35 +01:00
case INTSTYLE_POSTGRES:
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
cp = AddPostgresIntPart(cp, year, "year", &is_zero, &is_before);
Pgindent run before 9.1 beta2.
2011-06-09 20:32:50 +02:00
Add C comment about why we don't spell out "month" in interval values.
2011-05-25 05:24:44 +02:00
/*
Pgindent run before 9.1 beta2.
2011-06-09 20:32:50 +02:00
* Ideally we should spell out "month" like we do for "year" and
* "day". However, for backward compatibility, we can't easily
* fix this. bjm 2011-05-24
Add C comment about why we don't spell out "month" in interval values.
2011-05-25 05:24:44 +02:00
*/
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
cp = AddPostgresIntPart(cp, mon, "mon", &is_zero, &is_before);
cp = AddPostgresIntPart(cp, mday, "day", &is_zero, &is_before);
if (is_zero || hour != 0 || min != 0 || sec != 0 || fsec != 0)
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
{
8.4 pgindent run, with new combined Linux/FreeBSD/MinGW typedef list provided by Andrew.
2009-06-11 16:49:15 +02:00
bool minus = (hour < 0 || min < 0 || sec < 0 || fsec < 0);
Fix INTERVAL output when year/month has different sign as day/hour etc. Previously, all fields were unsigned, with only a trailing "ago" to indicate negative intervals. Now, ISO format does not use "ago", and and the traditional PostgreSQL format has the first numeric field unsigned with "ago" supporting that field. So "1 month - 2 days ago" is two days less than a month in the past. Fix interval arithmetic across daylight savings time boundaries. Previously, most math across boundaries introduced a one hour offset. Allow some date/time functions to return NULL if called with NULL args. Implement functions for AT TIME ZONE support. Support "SAT" as an Australian time zone if USE_AUSTRALIAN_RULES is defined.
2000-11-06 16:57:00 +01:00
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
sprintf(cp, "%s%s%02d:%02d:",
is_zero ? "" : " ",
Fix up "Postgres-style" time interval representation when fields have mixed-signs. Previous effort left way too many minus signs, and was at least as broken as the one before that :( Clean up "ISO-style" time interval representation to omit zero fields if there is at least one non-zero field. Supress some leading plus signs when not necessary for clarity. Replace every #ifdef __CYGWIN__ block with a cleaner TIMEZONE_GLOBAL macro defined in datetime.h.
2001-01-18 08:22:43 +01:00
(minus ? "-" : (is_before ? "+" : "")),
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
abs(hour), abs(min));
Add a new GUC variable called "IntervalStyle" that decouples interval output from DateStyle, and create a new interval style that produces output matching the SQL standard (at least for interval values that fall within the standard's restrictions). IntervalStyle is also used to resolve the conflict between the standard and traditional Postgres rules for interpreting negative interval input. Ron Mayer
2008-11-09 01:28:35 +01:00
cp += strlen(cp);
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
cp = AppendSeconds(cp, sec, fsec, MAX_INTERVAL_PRECISION, true);
*cp = '\0';
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
}
break;
Add conversion from datetime to time data type. Rename date+time conversion to datetime to ensure less than 16 characters in routine name (required to fit in pg_proc table).
1997-10-25 07:16:09 +02:00
8.4 pgindent run, with new combined Linux/FreeBSD/MinGW typedef list provided by Andrew.
2009-06-11 16:49:15 +02:00
/* Compatible with postgresql < 8.4 when DateStyle != 'iso' */
Add a new GUC variable called "IntervalStyle" that decouples interval output from DateStyle, and create a new interval style that produces output matching the SQL standard (at least for interval values that fall within the standard's restrictions). IntervalStyle is also used to resolve the conflict between the standard and traditional Postgres rules for interpreting negative interval input. Ron Mayer
2008-11-09 01:28:35 +01:00
case INTSTYLE_POSTGRES_VERBOSE:
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
default:
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
strcpy(cp, "@");
cp++;
cp = AddVerboseIntPart(cp, year, "year", &is_zero, &is_before);
cp = AddVerboseIntPart(cp, mon, "mon", &is_zero, &is_before);
cp = AddVerboseIntPart(cp, mday, "day", &is_zero, &is_before);
cp = AddVerboseIntPart(cp, hour, "hour", &is_zero, &is_before);
cp = AddVerboseIntPart(cp, min, "min", &is_zero, &is_before);
if (sec != 0 || fsec != 0)
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
{
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
*cp++ = ' ';
if (sec < 0 || (sec == 0 && fsec < 0))
Fix mis-display of negative fractional seconds in interval values for --enable-integer-datetimes case. Per report from Oliver Siegmar.
2005-04-20 19:14:50 +02:00
{
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
if (is_zero)
Change TRUE/FALSE to true/false The lower case spellings are C and C++ standard and are used in most parts of the PostgreSQL sources. The upper case spellings are only used in some files/modules. So standardize on the standard spellings. The APIs for ICU, Perl, and Windows define their own TRUE and FALSE, so those are left as is when using those APIs. In code comments, we use the lower-case spelling for the C concepts and keep the upper-case spelling for the SQL concepts. Reviewed-by: Michael Paquier <michael.paquier@gmail.com>
2017-08-16 06:22:32 +02:00
is_before = true;
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
else if (!is_before)
*cp++ = '-';
Fix mis-display of negative fractional seconds in interval values for --enable-integer-datetimes case. Per report from Oliver Siegmar.
2005-04-20 19:14:50 +02:00
}
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
else if (is_before)
*cp++ = '-';
Improve speed of timestamp/time/date output functions. It seems that sprintf(), at least in glibc's version, is unreasonably slow compared to hand-rolled code for printing integers. Replacing most uses of sprintf() in the datetime.c output functions with special-purpose code turns out to give more than a 2X speedup in COPY of a table with a single timestamp column; which is pretty impressive considering all the other logic in that code path. David Rowley and Andres Freund, reviewed by Peter Geoghegan and myself
2016-02-07 05:11:28 +01:00
cp = AppendSeconds(cp, sec, fsec, MAX_INTERVAL_PRECISION, false);
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
sprintf(cp, " sec%s",
(abs(sec) != 1 || fsec != 0) ? "s" : "");
Change TRUE/FALSE to true/false The lower case spellings are C and C++ standard and are used in most parts of the PostgreSQL sources. The upper case spellings are only used in some files/modules. So standardize on the standard spellings. The APIs for ICU, Perl, and Windows define their own TRUE and FALSE, so those are left as is when using those APIs. In code comments, we use the lower-case spelling for the C concepts and keep the upper-case spelling for the SQL concepts. Reviewed-by: Michael Paquier <michael.paquier@gmail.com>
2017-08-16 06:22:32 +02:00
is_zero = false;
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
}
Add a new GUC variable called "IntervalStyle" that decouples interval output from DateStyle, and create a new interval style that produces output matching the SQL standard (at least for interval values that fall within the standard's restrictions). IntervalStyle is also used to resolve the conflict between the standard and traditional Postgres rules for interpreting negative interval input. Ron Mayer
2008-11-09 01:28:35 +01:00
/* identically zero? then put in a unitless zero... */
Clean up the ancient decision to show only two fractional-seconds digits in "postgres_verbose" intervalstyle, and the equally arbitrary decision to show at least two fractional-seconds digits in most other datetime display styles. This results in some minor changes in the expected regression test outputs. Also, coalesce a lot of repetitive code in datetime.c into subroutines, for clarity and ease of maintenance. In particular this roughly halves the number of #ifdef HAVE_INT64_TIMESTAMP segments. Ron Mayer, with some additional kibitzing from Tom Lane
2008-11-12 02:36:20 +01:00
if (is_zero)
strcat(cp, " 0");
Add a new GUC variable called "IntervalStyle" that decouples interval output from DateStyle, and create a new interval style that produces output matching the SQL standard (at least for interval values that fall within the standard's restrictions). IntervalStyle is also used to resolve the conflict between the standard and traditional Postgres rules for interpreting negative interval input. Ron Mayer
2008-11-09 01:28:35 +01:00
if (is_before)
strcat(cp, " ago");
All regression tests pass except for rules.sql (unrelated). Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 19:17:02 +01:00
break;
Add conversion from datetime to time data type. Rename date+time conversion to datetime to ensure less than 16 characters in routine name (required to fit in pg_proc table).
1997-10-25 07:16:09 +02:00
}
Remove the useless and rather inconsistent return values of EncodeDateOnly, EncodeTimeOnly, EncodeDateTime, EncodeInterval. These don't have any good reason to fail, and their callers were mostly not checking anyway.
2009-05-26 04:17:50 +02:00
}
Add GUC setting for Australian timezones. Uses new GUC boolean callback functions to clear date cache. Allow regression tests to pass when timezone set.
2001-06-18 18:14:44 +02:00
Repair an embarrassingly large number of alphabetization mistakes in the datetime token tables. Even more embarrassing, the regression tests revealed some of the problems --- but evidently the bogus output wasn't questioned. Add code to postmaster startup to directly check the tables for correct ordering, in hopes of not being embarrassed like this again.
2003-01-16 01:26:49 +01:00
/*
* We've been burnt by stupid errors in the ordering of the datetkn tables
pgindent run for 9.4 This includes removing tabs after periods in C comments, which was applied to back branches, so this change should not effect backpatching.
2014-05-06 18:12:18 +02:00
* once too often. Arrange to check them during postmaster start.
Repair an embarrassingly large number of alphabetization mistakes in the datetime token tables. Even more embarrassing, the regression tests revealed some of the problems --- but evidently the bogus output wasn't questioned. Add code to postmaster startup to directly check the tables for correct ordering, in hopes of not being embarrassed like this again.
2003-01-16 01:26:49 +01:00
*/
static bool
Remove hard-wired lists of timezone abbreviations in favor of providing configuration files that can be altered by a DBA. The australian_timezones GUC setting disappears, replaced by a timezone_abbreviations setting (set this to 'Australia' to get the effect of australian_timezones). The list of zone names defined by default has undergone a bit of cleanup, too. Documentation still needs some work --- in particular, should we fix Table B-4, or just get rid of it? Joachim Wieland, with some editorializing by moi.
2006-07-25 05:51:23 +02:00
CheckDateTokenTable(const char *tablename, const datetkn *base, int nel)
Repair an embarrassingly large number of alphabetization mistakes in the datetime token tables. Even more embarrassing, the regression tests revealed some of the problems --- but evidently the bogus output wasn't questioned. Add code to postmaster startup to directly check the tables for correct ordering, in hopes of not being embarrassed like this again.
2003-01-16 01:26:49 +01:00
{
bool ok = true;
Remove hard-wired lists of timezone abbreviations in favor of providing configuration files that can be altered by a DBA. The australian_timezones GUC setting disappears, replaced by a timezone_abbreviations setting (set this to 'Australia' to get the effect of australian_timezones). The list of zone names defined by default has undergone a bit of cleanup, too. Documentation still needs some work --- in particular, should we fix Table B-4, or just get rid of it? Joachim Wieland, with some editorializing by moi.
2006-07-25 05:51:23 +02:00
int i;
Repair an embarrassingly large number of alphabetization mistakes in the datetime token tables. Even more embarrassing, the regression tests revealed some of the problems --- but evidently the bogus output wasn't questioned. Add code to postmaster startup to directly check the tables for correct ordering, in hopes of not being embarrassed like this again.
2003-01-16 01:26:49 +01:00
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
for (i = 0; i < nel; i++)
Repair an embarrassingly large number of alphabetization mistakes in the datetime token tables. Even more embarrassing, the regression tests revealed some of the problems --- but evidently the bogus output wasn't questioned. Add code to postmaster startup to directly check the tables for correct ordering, in hopes of not being embarrassed like this again.
2003-01-16 01:26:49 +01:00
{
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
/* check for token strings that don't fit */
if (strlen(base[i].token) > TOKMAXLEN)
Repair an embarrassingly large number of alphabetization mistakes in the datetime token tables. Even more embarrassing, the regression tests revealed some of the problems --- but evidently the bogus output wasn't questioned. Add code to postmaster startup to directly check the tables for correct ordering, in hopes of not being embarrassed like this again.
2003-01-16 01:26:49 +01:00
{
Work around a subtle portability problem in use of printf %s format. Depending on which spec you read, field widths and precisions in %s may be counted either in bytes or characters. Our code was assuming bytes, which is wrong at least for glibc's implementation, and in any case libc might have a different idea of the prevailing encoding than we do. Hence, for portable results we must avoid using anything more complex than just "%s" unless the string to be printed is known to be all-ASCII. This patch fixes the cases I could find, including the psql formatting failure reported by Hernan Gonzalez. In HEAD only, I also added comments to some places where it appears safe to continue using "%.*s".
2010-05-08 18:39:53 +02:00
/* %.*s is safe since all our tokens are ASCII */
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
elog(LOG, "token too long in %s table: \"%.*s\"",
Repair an embarrassingly large number of alphabetization mistakes in the datetime token tables. Even more embarrassing, the regression tests revealed some of the problems --- but evidently the bogus output wasn't questioned. Add code to postmaster startup to directly check the tables for correct ordering, in hopes of not being embarrassed like this again.
2003-01-16 01:26:49 +01:00
tablename,
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
TOKMAXLEN + 1, base[i].token);
ok = false;
break; /* don't risk applying strcmp */
}
/* check for out of order */
if (i > 0 &&
strcmp(base[i - 1].token, base[i].token) >= 0)
{
elog(LOG, "ordering error in %s table: \"%s\" >= \"%s\"",
tablename,
base[i - 1].token,
base[i].token);
Repair an embarrassingly large number of alphabetization mistakes in the datetime token tables. Even more embarrassing, the regression tests revealed some of the problems --- but evidently the bogus output wasn't questioned. Add code to postmaster startup to directly check the tables for correct ordering, in hopes of not being embarrassed like this again.
2003-01-16 01:26:49 +01:00
ok = false;
}
}
return ok;
}
bool
CheckDateTokenTables(void)
{
bool ok = true;
Avoid repeated computation of the constants date2j(1970, 1, 1) and date2j(2000, 1, 1). Should make for some marginal speed improvement in date/time operations.
2003-04-04 06:50:44 +02:00
Assert(UNIX_EPOCH_JDATE == date2j(1970, 1, 1));
Assert(POSTGRES_EPOCH_JDATE == date2j(2000, 1, 1));
Repair an embarrassingly large number of alphabetization mistakes in the datetime token tables. Even more embarrassing, the regression tests revealed some of the problems --- but evidently the bogus output wasn't questioned. Add code to postmaster startup to directly check the tables for correct ordering, in hopes of not being embarrassed like this again.
2003-01-16 01:26:49 +01:00
ok &= CheckDateTokenTable("datetktbl", datetktbl, szdatetktbl);
ok &= CheckDateTokenTable("deltatktbl", deltatktbl, szdeltatktbl);
return ok;
}
Remove hard-wired lists of timezone abbreviations in favor of providing configuration files that can be altered by a DBA. The australian_timezones GUC setting disappears, replaced by a timezone_abbreviations setting (set this to 'Australia' to get the effect of australian_timezones). The list of zone names defined by default has undergone a bit of cleanup, too. Documentation still needs some work --- in particular, should we fix Table B-4, or just get rid of it? Joachim Wieland, with some editorializing by moi.
2006-07-25 05:51:23 +02:00
Add transform functions for various temporal typmod coercisions. This enables ALTER TABLE to skip table and index rebuilds in some cases. Noah Misch, with trivial changes by me.
2012-02-08 15:33:02 +01:00
/*
Create the infrastructure for planner support functions. Rename/repurpose pg_proc.protransform as "prosupport". The idea is still that it names an internal function that provides knowledge to the planner about the behavior of the function it's attached to; but redesign the API specification so that it's not limited to doing just one thing, but can support an extensible set of requests. The original purpose of simplifying a function call is handled by the first request type to be invented, SupportRequestSimplify. Adjust all the existing transform functions to handle this API, and rename them fron "xxx_transform" to "xxx_support" to reflect the potential generalization of what they do. (Since we never previously provided any way for extensions to add transform functions, this change doesn't create an API break for them.) Also add DDL and pg_dump support for attaching a support function to a user-defined function. Unfortunately, DDL access has to be restricted to superusers, at least for now; but seeing that support functions will pretty much have to be written in C, that limitation is just theoretical. (This support is untested in this patch, but a follow-on patch will add cases that exercise it.) Discussion: https://postgr.es/m/15193.1548028093@sss.pgh.pa.us
2019-02-10 00:08:48 +01:00
* Common code for temporal prosupport functions: simplify, if possible,
* a call to a temporal type's length-coercion function.
*
* Types time, timetz, timestamp and timestamptz each have a range of allowed
* precisions. An unspecified precision is rigorously equivalent to the
* highest specifiable precision. We can replace the function call with a
* no-op RelabelType if it is coercing to the same or higher precision as the
* input is known to have.
*
* The input Node is always a FuncExpr, but to reduce the #include footprint
* of datetime.h, we declare it as Node *.
Code review for protransform patches. Fix loss of previous expression-simplification work when a transform function fires: we must not simply revert to untransformed input tree. Instead build a dummy FuncExpr node to pass to the transform function. This has the additional advantage of providing a simpler, more uniform API for transform functions. Move documentation to a somewhat less buried spot, relocate some poorly-placed code, be more wary of null constants and invalid typmod values, add an opr_sanity check on protransform function signatures, and some other minor cosmetic adjustments. Note: although this patch touches pg_proc.h, no need for catversion bump, because the changes are cosmetic and don't actually change the intended catalog contents.
2012-03-23 22:29:57 +01:00
*
* Note: timestamp_scale throws an error when the typmod is out of range, but
* we can't get there from a cast: our typmodin will have caught it already.
Add transform functions for various temporal typmod coercisions. This enables ALTER TABLE to skip table and index rebuilds in some cases. Noah Misch, with trivial changes by me.
2012-02-08 15:33:02 +01:00
*/
Node *
Create the infrastructure for planner support functions. Rename/repurpose pg_proc.protransform as "prosupport". The idea is still that it names an internal function that provides knowledge to the planner about the behavior of the function it's attached to; but redesign the API specification so that it's not limited to doing just one thing, but can support an extensible set of requests. The original purpose of simplifying a function call is handled by the first request type to be invented, SupportRequestSimplify. Adjust all the existing transform functions to handle this API, and rename them fron "xxx_transform" to "xxx_support" to reflect the potential generalization of what they do. (Since we never previously provided any way for extensions to add transform functions, this change doesn't create an API break for them.) Also add DDL and pg_dump support for attaching a support function to a user-defined function. Unfortunately, DDL access has to be restricted to superusers, at least for now; but seeing that support functions will pretty much have to be written in C, that limitation is just theoretical. (This support is untested in this patch, but a follow-on patch will add cases that exercise it.) Discussion: https://postgr.es/m/15193.1548028093@sss.pgh.pa.us
2019-02-10 00:08:48 +01:00
TemporalSimplify(int32 max_precis, Node *node)
Add transform functions for various temporal typmod coercisions. This enables ALTER TABLE to skip table and index rebuilds in some cases. Noah Misch, with trivial changes by me.
2012-02-08 15:33:02 +01:00
{
Make more use of castNode()
2017-02-21 17:33:07 +01:00
FuncExpr *expr = castNode(FuncExpr, node);
Add transform functions for various temporal typmod coercisions. This enables ALTER TABLE to skip table and index rebuilds in some cases. Noah Misch, with trivial changes by me.
2012-02-08 15:33:02 +01:00
Node *ret = NULL;
Code review for protransform patches. Fix loss of previous expression-simplification work when a transform function fires: we must not simply revert to untransformed input tree. Instead build a dummy FuncExpr node to pass to the transform function. This has the additional advantage of providing a simpler, more uniform API for transform functions. Move documentation to a somewhat less buried spot, relocate some poorly-placed code, be more wary of null constants and invalid typmod values, add an opr_sanity check on protransform function signatures, and some other minor cosmetic adjustments. Note: although this patch touches pg_proc.h, no need for catversion bump, because the changes are cosmetic and don't actually change the intended catalog contents.
2012-03-23 22:29:57 +01:00
Node *typmod;
Add transform functions for various temporal typmod coercisions. This enables ALTER TABLE to skip table and index rebuilds in some cases. Noah Misch, with trivial changes by me.
2012-02-08 15:33:02 +01:00
Code review for protransform patches. Fix loss of previous expression-simplification work when a transform function fires: we must not simply revert to untransformed input tree. Instead build a dummy FuncExpr node to pass to the transform function. This has the additional advantage of providing a simpler, more uniform API for transform functions. Move documentation to a somewhat less buried spot, relocate some poorly-placed code, be more wary of null constants and invalid typmod values, add an opr_sanity check on protransform function signatures, and some other minor cosmetic adjustments. Note: although this patch touches pg_proc.h, no need for catversion bump, because the changes are cosmetic and don't actually change the intended catalog contents.
2012-03-23 22:29:57 +01:00
Assert(list_length(expr->args) >= 2);
Add transform functions for various temporal typmod coercisions. This enables ALTER TABLE to skip table and index rebuilds in some cases. Noah Misch, with trivial changes by me.
2012-02-08 15:33:02 +01:00
Code review for protransform patches. Fix loss of previous expression-simplification work when a transform function fires: we must not simply revert to untransformed input tree. Instead build a dummy FuncExpr node to pass to the transform function. This has the additional advantage of providing a simpler, more uniform API for transform functions. Move documentation to a somewhat less buried spot, relocate some poorly-placed code, be more wary of null constants and invalid typmod values, add an opr_sanity check on protransform function signatures, and some other minor cosmetic adjustments. Note: although this patch touches pg_proc.h, no need for catversion bump, because the changes are cosmetic and don't actually change the intended catalog contents.
2012-03-23 22:29:57 +01:00
typmod = (Node *) lsecond(expr->args);
Add transform functions for various temporal typmod coercisions. This enables ALTER TABLE to skip table and index rebuilds in some cases. Noah Misch, with trivial changes by me.
2012-02-08 15:33:02 +01:00
Run pgindent on 9.2 source tree in preparation for first 9.3 commit-fest.
2012-06-10 21:20:04 +02:00
if (IsA(typmod, Const) &&!((Const *) typmod)->constisnull)
Add transform functions for various temporal typmod coercisions. This enables ALTER TABLE to skip table and index rebuilds in some cases. Noah Misch, with trivial changes by me.
2012-02-08 15:33:02 +01:00
{
Code review for protransform patches. Fix loss of previous expression-simplification work when a transform function fires: we must not simply revert to untransformed input tree. Instead build a dummy FuncExpr node to pass to the transform function. This has the additional advantage of providing a simpler, more uniform API for transform functions. Move documentation to a somewhat less buried spot, relocate some poorly-placed code, be more wary of null constants and invalid typmod values, add an opr_sanity check on protransform function signatures, and some other minor cosmetic adjustments. Note: although this patch touches pg_proc.h, no need for catversion bump, because the changes are cosmetic and don't actually change the intended catalog contents.
2012-03-23 22:29:57 +01:00
Node *source = (Node *) linitial(expr->args);
Add transform functions for various temporal typmod coercisions. This enables ALTER TABLE to skip table and index rebuilds in some cases. Noah Misch, with trivial changes by me.
2012-02-08 15:33:02 +01:00
int32 old_precis = exprTypmod(source);
int32 new_precis = DatumGetInt32(((Const *) typmod)->constvalue);
Code review for protransform patches. Fix loss of previous expression-simplification work when a transform function fires: we must not simply revert to untransformed input tree. Instead build a dummy FuncExpr node to pass to the transform function. This has the additional advantage of providing a simpler, more uniform API for transform functions. Move documentation to a somewhat less buried spot, relocate some poorly-placed code, be more wary of null constants and invalid typmod values, add an opr_sanity check on protransform function signatures, and some other minor cosmetic adjustments. Note: although this patch touches pg_proc.h, no need for catversion bump, because the changes are cosmetic and don't actually change the intended catalog contents.
2012-03-23 22:29:57 +01:00
if (new_precis < 0 || new_precis == max_precis ||
(old_precis >= 0 && new_precis >= old_precis))
Add transform functions for various temporal typmod coercisions. This enables ALTER TABLE to skip table and index rebuilds in some cases. Noah Misch, with trivial changes by me.
2012-02-08 15:33:02 +01:00
ret = relabel_to_typmod(source, new_precis);
}
return ret;
}
Remove hard-wired lists of timezone abbreviations in favor of providing configuration files that can be altered by a DBA. The australian_timezones GUC setting disappears, replaced by a timezone_abbreviations setting (set this to 'Australia' to get the effect of australian_timezones). The list of zone names defined by default has undergone a bit of cleanup, too. Documentation still needs some work --- in particular, should we fix Table B-4, or just get rid of it? Joachim Wieland, with some editorializing by moi.
2006-07-25 05:51:23 +02:00
/*
* This function gets called during timezone config file load or reload
* to create the final array of timezone tokens. The argument array
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
* is already sorted in name order.
*
* The result is a TimeZoneAbbrevTable (which must be a single malloc'd chunk)
* or NULL on malloc failure. No other error conditions are defined.
Remove hard-wired lists of timezone abbreviations in favor of providing configuration files that can be altered by a DBA. The australian_timezones GUC setting disappears, replaced by a timezone_abbreviations setting (set this to 'Australia' to get the effect of australian_timezones). The list of zone names defined by default has undergone a bit of cleanup, too. Documentation still needs some work --- in particular, should we fix Table B-4, or just get rid of it? Joachim Wieland, with some editorializing by moi.
2006-07-25 05:51:23 +02:00
*/
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
TimeZoneAbbrevTable *
ConvertTimeZoneAbbrevs(struct tzEntry *abbrevs, int n)
Remove hard-wired lists of timezone abbreviations in favor of providing configuration files that can be altered by a DBA. The australian_timezones GUC setting disappears, replaced by a timezone_abbreviations setting (set this to 'Australia' to get the effect of australian_timezones). The list of zone names defined by default has undergone a bit of cleanup, too. Documentation still needs some work --- in particular, should we fix Table B-4, or just get rid of it? Joachim Wieland, with some editorializing by moi.
2006-07-25 05:51:23 +02:00
{
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
TimeZoneAbbrevTable *tbl;
Size tbl_size;
Remove hard-wired lists of timezone abbreviations in favor of providing configuration files that can be altered by a DBA. The australian_timezones GUC setting disappears, replaced by a timezone_abbreviations setting (set this to 'Australia' to get the effect of australian_timezones). The list of zone names defined by default has undergone a bit of cleanup, too. Documentation still needs some work --- in particular, should we fix Table B-4, or just get rid of it? Joachim Wieland, with some editorializing by moi.
2006-07-25 05:51:23 +02:00
int i;
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
/* Space for fixed fields and datetkn array */
tbl_size = offsetof(TimeZoneAbbrevTable, abbrevs) +
n * sizeof(datetkn);
tbl_size = MAXALIGN(tbl_size);
/* Count up space for dynamic abbreviations */
for (i = 0; i < n; i++)
{
struct tzEntry *abbr = abbrevs + i;
if (abbr->zone != NULL)
{
Size dsize;
dsize = offsetof(DynamicZoneAbbrev, zone) +
strlen(abbr->zone) + 1;
tbl_size += MAXALIGN(dsize);
}
}
/* Alloc the result ... */
tbl = malloc(tbl_size);
if (!tbl)
return NULL;
/* ... and fill it in */
tbl->tblsize = tbl_size;
Revise the API for GUC variable assign hooks. The previous functions of assign hooks are now split between check hooks and assign hooks, where the former can fail but the latter shouldn't. Aside from being conceptually clearer, this approach exposes the "canonicalized" form of the variable value to guc.c without having to do an actual assignment. And that lets us fix the problem recently noted by Bernd Helmle that the auto-tune patch for wal_buffers resulted in bogus log messages about "parameter "wal_buffers" cannot be changed without restarting the server". There may be some speed advantage too, because this design lets hook functions avoid re-parsing variable values when restoring a previous state after a rollback (they can store a pre-parsed representation of the value instead). This patch also resolves a longstanding annoyance about custom error messages from variable assign hooks: they should modify, not appear separately from, guc.c's own message about "invalid parameter value".
2011-04-07 06:11:01 +02:00
tbl->numabbrevs = n;
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
/* in this loop, tbl_size reprises the space calculation above */
tbl_size = offsetof(TimeZoneAbbrevTable, abbrevs) +
n * sizeof(datetkn);
tbl_size = MAXALIGN(tbl_size);
Remove hard-wired lists of timezone abbreviations in favor of providing configuration files that can be altered by a DBA. The australian_timezones GUC setting disappears, replaced by a timezone_abbreviations setting (set this to 'Australia' to get the effect of australian_timezones). The list of zone names defined by default has undergone a bit of cleanup, too. Documentation still needs some work --- in particular, should we fix Table B-4, or just get rid of it? Joachim Wieland, with some editorializing by moi.
2006-07-25 05:51:23 +02:00
for (i = 0; i < n; i++)
{
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
struct tzEntry *abbr = abbrevs + i;
datetkn *dtoken = tbl->abbrevs + i;
/* use strlcpy to truncate name if necessary */
strlcpy(dtoken->token, abbr->abbrev, TOKMAXLEN + 1);
if (abbr->zone != NULL)
{
/* Allocate a DynamicZoneAbbrev for this abbreviation */
DynamicZoneAbbrev *dtza;
Size dsize;
dtza = (DynamicZoneAbbrev *) ((char *) tbl + tbl_size);
dtza->tz = NULL;
strcpy(dtza->zone, abbr->zone);
dtoken->type = DYNTZ;
/* value is offset from table start to DynamicZoneAbbrev */
dtoken->value = (int32) tbl_size;
dsize = offsetof(DynamicZoneAbbrev, zone) +
strlen(abbr->zone) + 1;
tbl_size += MAXALIGN(dsize);
}
else
{
dtoken->type = abbr->is_dst ? DTZ : TZ;
dtoken->value = abbr->offset;
}
Remove hard-wired lists of timezone abbreviations in favor of providing configuration files that can be altered by a DBA. The australian_timezones GUC setting disappears, replaced by a timezone_abbreviations setting (set this to 'Australia' to get the effect of australian_timezones). The list of zone names defined by default has undergone a bit of cleanup, too. Documentation still needs some work --- in particular, should we fix Table B-4, or just get rid of it? Joachim Wieland, with some editorializing by moi.
2006-07-25 05:51:23 +02:00
}
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
/* Assert the two loops above agreed on size calculations */
Assert(tbl->tblsize == tbl_size);
Remove hard-wired lists of timezone abbreviations in favor of providing configuration files that can be altered by a DBA. The australian_timezones GUC setting disappears, replaced by a timezone_abbreviations setting (set this to 'Australia' to get the effect of australian_timezones). The list of zone names defined by default has undergone a bit of cleanup, too. Documentation still needs some work --- in particular, should we fix Table B-4, or just get rid of it? Joachim Wieland, with some editorializing by moi.
2006-07-25 05:51:23 +02:00
/* Check the ordering, if testing */
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
Assert(CheckDateTokenTable("timezone abbreviations", tbl->abbrevs, n));
return tbl;
Revise the API for GUC variable assign hooks. The previous functions of assign hooks are now split between check hooks and assign hooks, where the former can fail but the latter shouldn't. Aside from being conceptually clearer, this approach exposes the "canonicalized" form of the variable value to guc.c without having to do an actual assignment. And that lets us fix the problem recently noted by Bernd Helmle that the auto-tune patch for wal_buffers resulted in bogus log messages about "parameter "wal_buffers" cannot be changed without restarting the server". There may be some speed advantage too, because this design lets hook functions avoid re-parsing variable values when restoring a previous state after a rollback (they can store a pre-parsed representation of the value instead). This patch also resolves a longstanding annoyance about custom error messages from variable assign hooks: they should modify, not appear separately from, guc.c's own message about "invalid parameter value".
2011-04-07 06:11:01 +02:00
}
/*
* Install a TimeZoneAbbrevTable as the active table.
*
* Caller is responsible that the passed table doesn't go away while in use.
*/
void
InstallTimeZoneAbbrevs(TimeZoneAbbrevTable *tbl)
{
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
zoneabbrevtbl = tbl;
/* reset abbrevcache, which may contain pointers into old table */
memset(abbrevcache, 0, sizeof(abbrevcache));
}
Remove hard-wired lists of timezone abbreviations in favor of providing configuration files that can be altered by a DBA. The australian_timezones GUC setting disappears, replaced by a timezone_abbreviations setting (set this to 'Australia' to get the effect of australian_timezones). The list of zone names defined by default has undergone a bit of cleanup, too. Documentation still needs some work --- in particular, should we fix Table B-4, or just get rid of it? Joachim Wieland, with some editorializing by moi.
2006-07-25 05:51:23 +02:00
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
/*
* Helper subroutine to locate pg_tz timezone for a dynamic abbreviation.
*/
static pg_tz *
FetchDynamicTimeZone(TimeZoneAbbrevTable *tbl, const datetkn *tp)
{
DynamicZoneAbbrev *dtza;
Remove hard-wired lists of timezone abbreviations in favor of providing configuration files that can be altered by a DBA. The australian_timezones GUC setting disappears, replaced by a timezone_abbreviations setting (set this to 'Australia' to get the effect of australian_timezones). The list of zone names defined by default has undergone a bit of cleanup, too. Documentation still needs some work --- in particular, should we fix Table B-4, or just get rid of it? Joachim Wieland, with some editorializing by moi.
2006-07-25 05:51:23 +02:00
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
/* Just some sanity checks to prevent indexing off into nowhere */
Assert(tp->type == DYNTZ);
Assert(tp->value > 0 && tp->value < tbl->tblsize);
dtza = (DynamicZoneAbbrev *) ((char *) tbl + tp->value);
/* Look up the underlying zone if we haven't already */
if (dtza->tz == NULL)
{
dtza->tz = pg_tzset(dtza->zone);
/*
* Ideally we'd let the caller ereport instead of doing it here, but
* then there is no way to report the bad time zone name.
*/
if (dtza->tz == NULL)
ereport(ERROR,
(errcode(ERRCODE_CONFIG_FILE_ERROR),
errmsg("time zone \"%s\" not recognized",
dtza->zone),
errdetail("This time zone name appears in the configuration file for time zone abbreviation \"%s\".",
tp->token)));
}
return dtza->tz;
Remove hard-wired lists of timezone abbreviations in favor of providing configuration files that can be altered by a DBA. The australian_timezones GUC setting disappears, replaced by a timezone_abbreviations setting (set this to 'Australia' to get the effect of australian_timezones). The list of zone names defined by default has undergone a bit of cleanup, too. Documentation still needs some work --- in particular, should we fix Table B-4, or just get rid of it? Joachim Wieland, with some editorializing by moi.
2006-07-25 05:51:23 +02:00
}
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
Remove hard-wired lists of timezone abbreviations in favor of providing configuration files that can be altered by a DBA. The australian_timezones GUC setting disappears, replaced by a timezone_abbreviations setting (set this to 'Australia' to get the effect of australian_timezones). The list of zone names defined by default has undergone a bit of cleanup, too. Documentation still needs some work --- in particular, should we fix Table B-4, or just get rid of it? Joachim Wieland, with some editorializing by moi.
2006-07-25 05:51:23 +02:00
/*
* This set-returning function reads all the available time zone abbreviations
Rename the recently-added pg_timezonenames view to pg_timezone_abbrevs, and create a new view pg_timezone_names that provides information about the zones known in the 'zic' database. Magnus Hagander, with some additional work by Tom Lane.
2006-09-16 22:14:34 +02:00
* and returns a set of (abbrev, utc_offset, is_dst).
Remove hard-wired lists of timezone abbreviations in favor of providing configuration files that can be altered by a DBA. The australian_timezones GUC setting disappears, replaced by a timezone_abbreviations setting (set this to 'Australia' to get the effect of australian_timezones). The list of zone names defined by default has undergone a bit of cleanup, too. Documentation still needs some work --- in particular, should we fix Table B-4, or just get rid of it? Joachim Wieland, with some editorializing by moi.
2006-07-25 05:51:23 +02:00
*/
Datum
Rename the recently-added pg_timezonenames view to pg_timezone_abbrevs, and create a new view pg_timezone_names that provides information about the zones known in the 'zic' database. Magnus Hagander, with some additional work by Tom Lane.
2006-09-16 22:14:34 +02:00
pg_timezone_abbrevs(PG_FUNCTION_ARGS)
Remove hard-wired lists of timezone abbreviations in favor of providing configuration files that can be altered by a DBA. The australian_timezones GUC setting disappears, replaced by a timezone_abbreviations setting (set this to 'Australia' to get the effect of australian_timezones). The list of zone names defined by default has undergone a bit of cleanup, too. Documentation still needs some work --- in particular, should we fix Table B-4, or just get rid of it? Joachim Wieland, with some editorializing by moi.
2006-07-25 05:51:23 +02:00
{
pgindent run for 8.2.
2006-10-04 02:30:14 +02:00
FuncCallContext *funcctx;
int *pindex;
Datum result;
HeapTuple tuple;
Datum values[3];
bool nulls[3];
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
const datetkn *tp;
pgindent run for 8.2.
2006-10-04 02:30:14 +02:00
char buffer[TOKMAXLEN + 1];
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
int gmtoffset;
bool is_dst;
pgindent run for 8.2.
2006-10-04 02:30:14 +02:00
unsigned char *p;
struct pg_tm tm;
Interval *resInterval;
Remove hard-wired lists of timezone abbreviations in favor of providing configuration files that can be altered by a DBA. The australian_timezones GUC setting disappears, replaced by a timezone_abbreviations setting (set this to 'Australia' to get the effect of australian_timezones). The list of zone names defined by default has undergone a bit of cleanup, too. Documentation still needs some work --- in particular, should we fix Table B-4, or just get rid of it? Joachim Wieland, with some editorializing by moi.
2006-07-25 05:51:23 +02:00
/* stuff done only on the first call of the function */
if (SRF_IS_FIRSTCALL())
{
pgindent run for 8.2.
2006-10-04 02:30:14 +02:00
TupleDesc tupdesc;
MemoryContext oldcontext;
Remove hard-wired lists of timezone abbreviations in favor of providing configuration files that can be altered by a DBA. The australian_timezones GUC setting disappears, replaced by a timezone_abbreviations setting (set this to 'Australia' to get the effect of australian_timezones). The list of zone names defined by default has undergone a bit of cleanup, too. Documentation still needs some work --- in particular, should we fix Table B-4, or just get rid of it? Joachim Wieland, with some editorializing by moi.
2006-07-25 05:51:23 +02:00
/* create a function context for cross-call persistence */
funcctx = SRF_FIRSTCALL_INIT();
/*
pgindent run for 8.2.
2006-10-04 02:30:14 +02:00
* switch to memory context appropriate for multiple function calls
Remove hard-wired lists of timezone abbreviations in favor of providing configuration files that can be altered by a DBA. The australian_timezones GUC setting disappears, replaced by a timezone_abbreviations setting (set this to 'Australia' to get the effect of australian_timezones). The list of zone names defined by default has undergone a bit of cleanup, too. Documentation still needs some work --- in particular, should we fix Table B-4, or just get rid of it? Joachim Wieland, with some editorializing by moi.
2006-07-25 05:51:23 +02:00
*/
oldcontext = MemoryContextSwitchTo(funcctx->multi_call_memory_ctx);
/* allocate memory for user context */
pindex = (int *) palloc(sizeof(int));
*pindex = 0;
funcctx->user_fctx = (void *) pindex;
/*
Rename the recently-added pg_timezonenames view to pg_timezone_abbrevs, and create a new view pg_timezone_names that provides information about the zones known in the 'zic' database. Magnus Hagander, with some additional work by Tom Lane.
2006-09-16 22:14:34 +02:00
* build tupdesc for result tuples. This must match this function's
* pg_proc entry!
Remove hard-wired lists of timezone abbreviations in favor of providing configuration files that can be altered by a DBA. The australian_timezones GUC setting disappears, replaced by a timezone_abbreviations setting (set this to 'Australia' to get the effect of australian_timezones). The list of zone names defined by default has undergone a bit of cleanup, too. Documentation still needs some work --- in particular, should we fix Table B-4, or just get rid of it? Joachim Wieland, with some editorializing by moi.
2006-07-25 05:51:23 +02:00
*/
Remove WITH OIDS support, change oid catalog column visibility. Previously tables declared WITH OIDS, including a significant fraction of the catalog tables, stored the oid column not as a normal column, but as part of the tuple header. This special column was not shown by default, which was somewhat odd, as it's often (consider e.g. pg_class.oid) one of the more important parts of a row. Neither pg_dump nor COPY included the contents of the oid column by default. The fact that the oid column was not an ordinary column necessitated a significant amount of special case code to support oid columns. That already was painful for the existing, but upcoming work aiming to make table storage pluggable, would have required expanding and duplicating that "specialness" significantly. WITH OIDS has been deprecated since 2005 (commit ff02d0a05280e0). Remove it. Removing includes: - CREATE TABLE and ALTER TABLE syntax for declaring the table to be WITH OIDS has been removed (WITH (oids[ = true]) will error out) - pg_dump does not support dumping tables declared WITH OIDS and will issue a warning when dumping one (and ignore the oid column). - restoring an pg_dump archive with pg_restore will warn when restoring a table with oid contents (and ignore the oid column) - COPY will refuse to load binary dump that includes oids. - pg_upgrade will error out when encountering tables declared WITH OIDS, they have to be altered to remove the oid column first. - Functionality to access the oid of the last inserted row (like plpgsql's RESULT_OID, spi's SPI_lastoid, ...) has been removed. The syntax for declaring a table WITHOUT OIDS (or WITH (oids = false) for CREATE TABLE) is still supported. While that requires a bit of support code, it seems unnecessary to break applications / dumps that do not use oids, and are explicit about not using them. The biggest user of WITH OID columns was postgres' catalog. This commit changes all 'magic' oid columns to be columns that are normally declared and stored. To reduce unnecessary query breakage all the newly added columns are still named 'oid', even if a table's column naming scheme would indicate 'reloid' or such. This obviously requires adapting a lot code, mostly replacing oid access via HeapTupleGetOid() with access to the underlying Form_pg_*->oid column. The bootstrap process now assigns oids for all oid columns in genbki.pl that do not have an explicit value (starting at the largest oid previously used), only oids assigned later by oids will be above FirstBootstrapObjectId. As the oid column now is a normal column the special bootstrap syntax for oids has been removed. Oids are not automatically assigned during insertion anymore, all backend code explicitly assigns oids with GetNewOidWithIndex(). For the rare case that insertions into the catalog via SQL are called for the new pg_nextoid() function can be used (which only works on catalog tables). The fact that oid columns on system tables are now normal columns means that they will be included in the set of columns expanded by * (i.e. SELECT * FROM pg_class will now include the table's oid, previously it did not). It'd not technically be hard to hide oid column by default, but that'd mean confusing behavior would either have to be carried forward forever, or it'd cause breakage down the line. While it's not unlikely that further adjustments are needed, the scope/invasiveness of the patch makes it worthwhile to get merge this now. It's painful to maintain externally, too complicated to commit after the code code freeze, and a dependency of a number of other patches. Catversion bump, for obvious reasons. Author: Andres Freund, with contributions by John Naylor Discussion: https://postgr.es/m/20180930034810.ywp2c7awz7opzcfr@alap3.anarazel.de
2018-11-21 00:36:57 +01:00
tupdesc = CreateTemplateTupleDesc(3);
Rename the recently-added pg_timezonenames view to pg_timezone_abbrevs, and create a new view pg_timezone_names that provides information about the zones known in the 'zic' database. Magnus Hagander, with some additional work by Tom Lane.
2006-09-16 22:14:34 +02:00
TupleDescInitEntry(tupdesc, (AttrNumber) 1, "abbrev",
Remove hard-wired lists of timezone abbreviations in favor of providing configuration files that can be altered by a DBA. The australian_timezones GUC setting disappears, replaced by a timezone_abbreviations setting (set this to 'Australia' to get the effect of australian_timezones). The list of zone names defined by default has undergone a bit of cleanup, too. Documentation still needs some work --- in particular, should we fix Table B-4, or just get rid of it? Joachim Wieland, with some editorializing by moi.
2006-07-25 05:51:23 +02:00
TEXTOID, -1, 0);
TupleDescInitEntry(tupdesc, (AttrNumber) 2, "utc_offset",
INTERVALOID, -1, 0);
TupleDescInitEntry(tupdesc, (AttrNumber) 3, "is_dst",
BOOLOID, -1, 0);
funcctx->tuple_desc = BlessTupleDesc(tupdesc);
MemoryContextSwitchTo(oldcontext);
}
/* stuff done on every call of the function */
funcctx = SRF_PERCALL_SETUP();
pindex = (int *) funcctx->user_fctx;
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
if (zoneabbrevtbl == NULL ||
*pindex >= zoneabbrevtbl->numabbrevs)
Remove hard-wired lists of timezone abbreviations in favor of providing configuration files that can be altered by a DBA. The australian_timezones GUC setting disappears, replaced by a timezone_abbreviations setting (set this to 'Australia' to get the effect of australian_timezones). The list of zone names defined by default has undergone a bit of cleanup, too. Documentation still needs some work --- in particular, should we fix Table B-4, or just get rid of it? Joachim Wieland, with some editorializing by moi.
2006-07-25 05:51:23 +02:00
SRF_RETURN_DONE(funcctx);
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
tp = zoneabbrevtbl->abbrevs + *pindex;
switch (tp->type)
{
case TZ:
gmtoffset = tp->value;
is_dst = false;
break;
case DTZ:
gmtoffset = tp->value;
is_dst = true;
break;
case DYNTZ:
{
/* Determine the current meaning of the abbrev */
pg_tz *tzp;
TimestampTz now;
int isdst;
tzp = FetchDynamicTimeZone(zoneabbrevtbl, tp);
now = GetCurrentTransactionStartTimestamp();
gmtoffset = -DetermineTimeZoneAbbrevOffsetTS(now,
tp->token,
tzp,
&isdst);
is_dst = (bool) isdst;
break;
}
default:
elog(ERROR, "unrecognized timezone type %d", (int) tp->type);
gmtoffset = 0; /* keep compiler quiet */
is_dst = false;
break;
}
Remove hard-wired lists of timezone abbreviations in favor of providing configuration files that can be altered by a DBA. The australian_timezones GUC setting disappears, replaced by a timezone_abbreviations setting (set this to 'Australia' to get the effect of australian_timezones). The list of zone names defined by default has undergone a bit of cleanup, too. Documentation still needs some work --- in particular, should we fix Table B-4, or just get rid of it? Joachim Wieland, with some editorializing by moi.
2006-07-25 05:51:23 +02:00
MemSet(nulls, 0, sizeof(nulls));
/*
pgindent run for 8.2.
2006-10-04 02:30:14 +02:00
* Convert name to text, using upcasing conversion that is the inverse of
* what ParseDateTime() uses.
Remove hard-wired lists of timezone abbreviations in favor of providing configuration files that can be altered by a DBA. The australian_timezones GUC setting disappears, replaced by a timezone_abbreviations setting (set this to 'Australia' to get the effect of australian_timezones). The list of zone names defined by default has undergone a bit of cleanup, too. Documentation still needs some work --- in particular, should we fix Table B-4, or just get rid of it? Joachim Wieland, with some editorializing by moi.
2006-07-25 05:51:23 +02:00
*/
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
strlcpy(buffer, tp->token, sizeof(buffer));
Remove hard-wired lists of timezone abbreviations in favor of providing configuration files that can be altered by a DBA. The australian_timezones GUC setting disappears, replaced by a timezone_abbreviations setting (set this to 'Australia' to get the effect of australian_timezones). The list of zone names defined by default has undergone a bit of cleanup, too. Documentation still needs some work --- in particular, should we fix Table B-4, or just get rid of it? Joachim Wieland, with some editorializing by moi.
2006-07-25 05:51:23 +02:00
for (p = (unsigned char *) buffer; *p; p++)
*p = pg_toupper(*p);
Simplify and standardize conversions between TEXT datums and ordinary C strings. This patch introduces four support functions cstring_to_text, cstring_to_text_with_len, text_to_cstring, and text_to_cstring_buffer, and two macros CStringGetTextDatum and TextDatumGetCString. A number of existing macros that provided variants on these themes were removed. Most of the places that need to make such conversions now require just one function or macro call, in place of the multiple notational layers that used to be needed. There are no longer any direct calls of textout or textin, and we got most of the places that were using handmade conversions via memcpy (there may be a few still lurking, though). This commit doesn't make any serious effort to eliminate transient memory leaks caused by detoasting toasted text objects before they reach text_to_cstring. We changed PG_GETARG_TEXT_P to PG_GETARG_TEXT_PP in a few places where it was easy, but much more could be done. Brendan Jurd and Tom Lane
2008-03-25 23:42:46 +01:00
values[0] = CStringGetTextDatum(buffer);
Remove hard-wired lists of timezone abbreviations in favor of providing configuration files that can be altered by a DBA. The australian_timezones GUC setting disappears, replaced by a timezone_abbreviations setting (set this to 'Australia' to get the effect of australian_timezones). The list of zone names defined by default has undergone a bit of cleanup, too. Documentation still needs some work --- in particular, should we fix Table B-4, or just get rid of it? Joachim Wieland, with some editorializing by moi.
2006-07-25 05:51:23 +02:00
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
/* Convert offset (in seconds) to an interval */
Remove hard-wired lists of timezone abbreviations in favor of providing configuration files that can be altered by a DBA. The australian_timezones GUC setting disappears, replaced by a timezone_abbreviations setting (set this to 'Australia' to get the effect of australian_timezones). The list of zone names defined by default has undergone a bit of cleanup, too. Documentation still needs some work --- in particular, should we fix Table B-4, or just get rid of it? Joachim Wieland, with some editorializing by moi.
2006-07-25 05:51:23 +02:00
MemSet(&tm, 0, sizeof(struct pg_tm));
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
tm.tm_sec = gmtoffset;
Remove hard-wired lists of timezone abbreviations in favor of providing configuration files that can be altered by a DBA. The australian_timezones GUC setting disappears, replaced by a timezone_abbreviations setting (set this to 'Australia' to get the effect of australian_timezones). The list of zone names defined by default has undergone a bit of cleanup, too. Documentation still needs some work --- in particular, should we fix Table B-4, or just get rid of it? Joachim Wieland, with some editorializing by moi.
2006-07-25 05:51:23 +02:00
resInterval = (Interval *) palloc(sizeof(Interval));
tm2interval(&tm, 0, resInterval);
values[1] = IntervalPGetDatum(resInterval);
Support timezone abbreviations that sometimes change. Up to now, PG has assumed that any given timezone abbreviation (such as "EDT") represents a constant GMT offset in the usage of any particular region; we had a way to configure what that offset was, but not for it to be changeable over time. But, as with most things horological, this view of the world is too simplistic: there are numerous regions that have at one time or another switched to a different GMT offset but kept using the same timezone abbreviation. Almost the entire Russian Federation did that a few years ago, and later this month they're going to do it again. And there are similar examples all over the world. To cope with this, invent the notion of a "dynamic timezone abbreviation", which is one that is referenced to a particular underlying timezone (as defined in the IANA timezone database) and means whatever it currently means in that zone. For zones that use or have used daylight-savings time, the standard and DST abbreviations continue to have the property that you can specify standard or DST time and get that time offset whether or not DST was theoretically in effect at the time. However, the abbreviations mean what they meant at the time in question (or most recently before that time) rather than being absolutely fixed. The standard abbreviation-list files have been changed to use this behavior for abbreviations that have actually varied in meaning since 1970. The old simple-numeric definitions are kept for abbreviations that have not changed, since they are a bit faster to resolve. While this is clearly a new feature, it seems necessary to back-patch it into all active branches, because otherwise use of Russian zone abbreviations is going to become even more problematic than it already was. This change supersedes the changes in commit 513d06ded et al to modify the fixed meanings of the Russian abbreviations; since we've not shipped that yet, this will avoid an undesirably incompatible (not to mention incorrect) change in behavior for timestamps between 2011 and 2014. This patch makes some cosmetic changes in ecpglib to keep its usage of datetime lookup tables as similar as possible to the backend code, but doesn't do anything about the increasingly obsolete set of timezone abbreviation definitions that are hard-wired into ecpglib. Whatever we do about that will likely not be appropriate material for back-patching. Also, a potential free() of a garbage pointer after an out-of-memory failure in ecpglib has been fixed. This patch also fixes pre-existing bugs in DetermineTimeZoneOffset() that caused it to produce unexpected results near a timezone transition, if both the "before" and "after" states are marked as standard time. We'd only ever thought about or tested transitions between standard and DST time, but that's not what's happening when a zone simply redefines their base GMT offset. In passing, update the SGML documentation to refer to the Olson/zoneinfo/ zic timezone database as the "IANA" database, since it's now being maintained under the auspices of IANA.
2014-10-16 21:22:10 +02:00
values[2] = BoolGetDatum(is_dst);
Remove hard-wired lists of timezone abbreviations in favor of providing configuration files that can be altered by a DBA. The australian_timezones GUC setting disappears, replaced by a timezone_abbreviations setting (set this to 'Australia' to get the effect of australian_timezones). The list of zone names defined by default has undergone a bit of cleanup, too. Documentation still needs some work --- in particular, should we fix Table B-4, or just get rid of it? Joachim Wieland, with some editorializing by moi.
2006-07-25 05:51:23 +02:00
(*pindex)++;
tuple = heap_form_tuple(funcctx->tuple_desc, values, nulls);
result = HeapTupleGetDatum(tuple);
SRF_RETURN_NEXT(funcctx, result);
}
Rename the recently-added pg_timezonenames view to pg_timezone_abbrevs, and create a new view pg_timezone_names that provides information about the zones known in the 'zic' database. Magnus Hagander, with some additional work by Tom Lane.
2006-09-16 22:14:34 +02:00
/*
* This set-returning function reads all the available full time zones
* and returns a set of (name, abbrev, utc_offset, is_dst).
*/
Datum
pg_timezone_names(PG_FUNCTION_ARGS)
{
Avoid holding a directory FD open across assorted SRF calls. This extends the fixes made in commit 085b6b667 to other SRFs with the same bug, namely pg_logdir_ls(), pgrowlocks(), pg_timezone_names(), pg_ls_dir(), and pg_tablespace_databases(). Also adjust various comments and documentation to warn against expecting to clean up resources during a ValuePerCall SRF's final call. Back-patch to all supported branches, since these functions were all born broken. Justin Pryzby, with cosmetic tweaks by me Discussion: https://postgr.es/m/20200308173103.GC1357@telsasoft.com
2020-03-17 02:05:28 +01:00
ReturnSetInfo *rsinfo = (ReturnSetInfo *) fcinfo->resultinfo;
bool randomAccess;
TupleDesc tupdesc;
Tuplestorestate *tupstore;
pgindent run for 8.2.
2006-10-04 02:30:14 +02:00
pg_tzenum *tzenum;
pg_tz *tz;
Datum values[4];
bool nulls[4];
Rename the recently-added pg_timezonenames view to pg_timezone_abbrevs, and create a new view pg_timezone_names that provides information about the zones known in the 'zic' database. Magnus Hagander, with some additional work by Tom Lane.
2006-09-16 22:14:34 +02:00
int tzoff;
pgindent run for 8.2.
2006-10-04 02:30:14 +02:00
struct pg_tm tm;
Rename the recently-added pg_timezonenames view to pg_timezone_abbrevs, and create a new view pg_timezone_names that provides information about the zones known in the 'zic' database. Magnus Hagander, with some additional work by Tom Lane.
2006-09-16 22:14:34 +02:00
fsec_t fsec;
Add const qualifier to tzn returned by timestamp2tm() The tzn value might come from tm->tm_zone, which libc declares as const, so it's prudent that the upper layers know about this as well.
2012-03-15 20:17:19 +01:00
const char *tzn;
pgindent run for 8.2.
2006-10-04 02:30:14 +02:00
Interval *resInterval;
struct pg_tm itm;
Avoid holding a directory FD open across assorted SRF calls. This extends the fixes made in commit 085b6b667 to other SRFs with the same bug, namely pg_logdir_ls(), pgrowlocks(), pg_timezone_names(), pg_ls_dir(), and pg_tablespace_databases(). Also adjust various comments and documentation to warn against expecting to clean up resources during a ValuePerCall SRF's final call. Back-patch to all supported branches, since these functions were all born broken. Justin Pryzby, with cosmetic tweaks by me Discussion: https://postgr.es/m/20200308173103.GC1357@telsasoft.com
2020-03-17 02:05:28 +01:00
MemoryContext oldcontext;
Rename the recently-added pg_timezonenames view to pg_timezone_abbrevs, and create a new view pg_timezone_names that provides information about the zones known in the 'zic' database. Magnus Hagander, with some additional work by Tom Lane.
2006-09-16 22:14:34 +02:00
Avoid holding a directory FD open across assorted SRF calls. This extends the fixes made in commit 085b6b667 to other SRFs with the same bug, namely pg_logdir_ls(), pgrowlocks(), pg_timezone_names(), pg_ls_dir(), and pg_tablespace_databases(). Also adjust various comments and documentation to warn against expecting to clean up resources during a ValuePerCall SRF's final call. Back-patch to all supported branches, since these functions were all born broken. Justin Pryzby, with cosmetic tweaks by me Discussion: https://postgr.es/m/20200308173103.GC1357@telsasoft.com
2020-03-17 02:05:28 +01:00
/* check to see if caller supports us returning a tuplestore */
if (rsinfo == NULL || !IsA(rsinfo, ReturnSetInfo))
ereport(ERROR,
(errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
errmsg("set-valued function called in context that cannot accept a set")));
if (!(rsinfo->allowedModes & SFRM_Materialize))
ereport(ERROR,
(errcode(ERRCODE_SYNTAX_ERROR),
errmsg("materialize mode required, but it is not allowed in this context")));
Rename the recently-added pg_timezonenames view to pg_timezone_abbrevs, and create a new view pg_timezone_names that provides information about the zones known in the 'zic' database. Magnus Hagander, with some additional work by Tom Lane.
2006-09-16 22:14:34 +02:00
Avoid holding a directory FD open across assorted SRF calls. This extends the fixes made in commit 085b6b667 to other SRFs with the same bug, namely pg_logdir_ls(), pgrowlocks(), pg_timezone_names(), pg_ls_dir(), and pg_tablespace_databases(). Also adjust various comments and documentation to warn against expecting to clean up resources during a ValuePerCall SRF's final call. Back-patch to all supported branches, since these functions were all born broken. Justin Pryzby, with cosmetic tweaks by me Discussion: https://postgr.es/m/20200308173103.GC1357@telsasoft.com
2020-03-17 02:05:28 +01:00
/* The tupdesc and tuplestore must be created in ecxt_per_query_memory */
oldcontext = MemoryContextSwitchTo(rsinfo->econtext->ecxt_per_query_memory);
Rename the recently-added pg_timezonenames view to pg_timezone_abbrevs, and create a new view pg_timezone_names that provides information about the zones known in the 'zic' database. Magnus Hagander, with some additional work by Tom Lane.
2006-09-16 22:14:34 +02:00
Avoid holding a directory FD open across assorted SRF calls. This extends the fixes made in commit 085b6b667 to other SRFs with the same bug, namely pg_logdir_ls(), pgrowlocks(), pg_timezone_names(), pg_ls_dir(), and pg_tablespace_databases(). Also adjust various comments and documentation to warn against expecting to clean up resources during a ValuePerCall SRF's final call. Back-patch to all supported branches, since these functions were all born broken. Justin Pryzby, with cosmetic tweaks by me Discussion: https://postgr.es/m/20200308173103.GC1357@telsasoft.com
2020-03-17 02:05:28 +01:00
if (get_call_result_type(fcinfo, NULL, &tupdesc) != TYPEFUNC_COMPOSITE)
elog(ERROR, "return type must be a row type");
Rename the recently-added pg_timezonenames view to pg_timezone_abbrevs, and create a new view pg_timezone_names that provides information about the zones known in the 'zic' database. Magnus Hagander, with some additional work by Tom Lane.
2006-09-16 22:14:34 +02:00
Avoid holding a directory FD open across assorted SRF calls. This extends the fixes made in commit 085b6b667 to other SRFs with the same bug, namely pg_logdir_ls(), pgrowlocks(), pg_timezone_names(), pg_ls_dir(), and pg_tablespace_databases(). Also adjust various comments and documentation to warn against expecting to clean up resources during a ValuePerCall SRF's final call. Back-patch to all supported branches, since these functions were all born broken. Justin Pryzby, with cosmetic tweaks by me Discussion: https://postgr.es/m/20200308173103.GC1357@telsasoft.com
2020-03-17 02:05:28 +01:00
randomAccess = (rsinfo->allowedModes & SFRM_Materialize_Random) != 0;
tupstore = tuplestore_begin_heap(randomAccess, false, work_mem);
rsinfo->returnMode = SFRM_Materialize;
rsinfo->setResult = tupstore;
rsinfo->setDesc = tupdesc;
Rename the recently-added pg_timezonenames view to pg_timezone_abbrevs, and create a new view pg_timezone_names that provides information about the zones known in the 'zic' database. Magnus Hagander, with some additional work by Tom Lane.
2006-09-16 22:14:34 +02:00
Avoid holding a directory FD open across assorted SRF calls. This extends the fixes made in commit 085b6b667 to other SRFs with the same bug, namely pg_logdir_ls(), pgrowlocks(), pg_timezone_names(), pg_ls_dir(), and pg_tablespace_databases(). Also adjust various comments and documentation to warn against expecting to clean up resources during a ValuePerCall SRF's final call. Back-patch to all supported branches, since these functions were all born broken. Justin Pryzby, with cosmetic tweaks by me Discussion: https://postgr.es/m/20200308173103.GC1357@telsasoft.com
2020-03-17 02:05:28 +01:00
MemoryContextSwitchTo(oldcontext);
Rename the recently-added pg_timezonenames view to pg_timezone_abbrevs, and create a new view pg_timezone_names that provides information about the zones known in the 'zic' database. Magnus Hagander, with some additional work by Tom Lane.
2006-09-16 22:14:34 +02:00
Avoid holding a directory FD open across assorted SRF calls. This extends the fixes made in commit 085b6b667 to other SRFs with the same bug, namely pg_logdir_ls(), pgrowlocks(), pg_timezone_names(), pg_ls_dir(), and pg_tablespace_databases(). Also adjust various comments and documentation to warn against expecting to clean up resources during a ValuePerCall SRF's final call. Back-patch to all supported branches, since these functions were all born broken. Justin Pryzby, with cosmetic tweaks by me Discussion: https://postgr.es/m/20200308173103.GC1357@telsasoft.com
2020-03-17 02:05:28 +01:00
/* initialize timezone scanning code */
tzenum = pg_tzenumerate_start();
Rename the recently-added pg_timezonenames view to pg_timezone_abbrevs, and create a new view pg_timezone_names that provides information about the zones known in the 'zic' database. Magnus Hagander, with some additional work by Tom Lane.
2006-09-16 22:14:34 +02:00
/* search for another zone to display */
for (;;)
{
tz = pg_tzenumerate_next(tzenum);
if (!tz)
Avoid holding a directory FD open across assorted SRF calls. This extends the fixes made in commit 085b6b667 to other SRFs with the same bug, namely pg_logdir_ls(), pgrowlocks(), pg_timezone_names(), pg_ls_dir(), and pg_tablespace_databases(). Also adjust various comments and documentation to warn against expecting to clean up resources during a ValuePerCall SRF's final call. Back-patch to all supported branches, since these functions were all born broken. Justin Pryzby, with cosmetic tweaks by me Discussion: https://postgr.es/m/20200308173103.GC1357@telsasoft.com
2020-03-17 02:05:28 +01:00
break;
Rename the recently-added pg_timezonenames view to pg_timezone_abbrevs, and create a new view pg_timezone_names that provides information about the zones known in the 'zic' database. Magnus Hagander, with some additional work by Tom Lane.
2006-09-16 22:14:34 +02:00
/* Convert now() to local time in this zone */
if (timestamp2tm(GetCurrentTransactionStartTimestamp(),
&tzoff, &tm, &fsec, &tzn, tz) != 0)
continue; /* ignore if conversion fails */
Suppress "Factory" zone in pg_timezone_names view for tzdata >= 2016g. IANA got rid of the really silly "abbreviation" and replaced it with one that's only moderately silly. But it's still pointless, so keep on not showing it.
2016-10-20 00:11:49 +02:00
/*
Tweak our special-case logic for the IANA "Factory" timezone. pg_timezone_names() tries to avoid showing the "Factory" zone in the view, mainly because that has traditionally had a very long "abbreviation" such as "Local time zone must be set--see zic manual page", so that showing it messes up psql's formatting of the whole view. Since tzdb version 2016g, IANA instead uses the abbreviation "-00", which is sane enough that there's no reason to discriminate against it. On the other hand, it emerges that FreeBSD and possibly other packagers are so wedded to backwards compatibility that they hack the IANA data to keep the old spelling --- and not just that old spelling, but even older spellings that IANA used back in the stone age. This caused the filter logic to fail to suppress "Factory" at all on such platforms, though the formatting problem is definitely real in that case. To solve both problems, get rid of the hard-wired assumption about exactly what Factory's abbreviation is, and instead reject abbreviations exceeding 31 characters. This will allow Factory to appear in the view if and only if it's using the modern abbreviation. In passing, simplify the code we add to zic.c to support "zic -P" to remove its now-obsolete hacks to not print the Factory zone's abbreviation. Unlike pg_timezone_names(), there's no reason for that code to support old/nonstandard timezone data. Since we generally prefer to keep timezone-related behavior the same in all branches, and since this is arguably a bug fix, back-patch to all supported branches. Discussion: https://postgr.es/m/3961.1564086915@sss.pgh.pa.us
2019-07-26 19:07:08 +02:00
* IANA's rather silly "Factory" time zone used to emit ridiculously
* long "abbreviations" such as "Local time zone must be set--see zic
* manual page" or "Local time zone must be set--use tzsetup". While
* modern versions of tzdb emit the much saner "-00", it seems some
* benighted packagers are hacking the IANA data so that it continues
* to produce these strings. To prevent producing a weirdly wide
* abbrev column, reject ridiculously long abbreviations.
Suppress "Factory" zone in pg_timezone_names view for tzdata >= 2016g. IANA got rid of the really silly "abbreviation" and replaced it with one that's only moderately silly. But it's still pointless, so keep on not showing it.
2016-10-20 00:11:49 +02:00
*/
Tweak our special-case logic for the IANA "Factory" timezone. pg_timezone_names() tries to avoid showing the "Factory" zone in the view, mainly because that has traditionally had a very long "abbreviation" such as "Local time zone must be set--see zic manual page", so that showing it messes up psql's formatting of the whole view. Since tzdb version 2016g, IANA instead uses the abbreviation "-00", which is sane enough that there's no reason to discriminate against it. On the other hand, it emerges that FreeBSD and possibly other packagers are so wedded to backwards compatibility that they hack the IANA data to keep the old spelling --- and not just that old spelling, but even older spellings that IANA used back in the stone age. This caused the filter logic to fail to suppress "Factory" at all on such platforms, though the formatting problem is definitely real in that case. To solve both problems, get rid of the hard-wired assumption about exactly what Factory's abbreviation is, and instead reject abbreviations exceeding 31 characters. This will allow Factory to appear in the view if and only if it's using the modern abbreviation. In passing, simplify the code we add to zic.c to support "zic -P" to remove its now-obsolete hacks to not print the Factory zone's abbreviation. Unlike pg_timezone_names(), there's no reason for that code to support old/nonstandard timezone data. Since we generally prefer to keep timezone-related behavior the same in all branches, and since this is arguably a bug fix, back-patch to all supported branches. Discussion: https://postgr.es/m/3961.1564086915@sss.pgh.pa.us
2019-07-26 19:07:08 +02:00
if (tzn && strlen(tzn) > 31)
Rename the recently-added pg_timezonenames view to pg_timezone_abbrevs, and create a new view pg_timezone_names that provides information about the zones known in the 'zic' database. Magnus Hagander, with some additional work by Tom Lane.
2006-09-16 22:14:34 +02:00
continue;
Avoid holding a directory FD open across assorted SRF calls. This extends the fixes made in commit 085b6b667 to other SRFs with the same bug, namely pg_logdir_ls(), pgrowlocks(), pg_timezone_names(), pg_ls_dir(), and pg_tablespace_databases(). Also adjust various comments and documentation to warn against expecting to clean up resources during a ValuePerCall SRF's final call. Back-patch to all supported branches, since these functions were all born broken. Justin Pryzby, with cosmetic tweaks by me Discussion: https://postgr.es/m/20200308173103.GC1357@telsasoft.com
2020-03-17 02:05:28 +01:00
MemSet(nulls, 0, sizeof(nulls));
Rename the recently-added pg_timezonenames view to pg_timezone_abbrevs, and create a new view pg_timezone_names that provides information about the zones known in the 'zic' database. Magnus Hagander, with some additional work by Tom Lane.
2006-09-16 22:14:34 +02:00
Avoid holding a directory FD open across assorted SRF calls. This extends the fixes made in commit 085b6b667 to other SRFs with the same bug, namely pg_logdir_ls(), pgrowlocks(), pg_timezone_names(), pg_ls_dir(), and pg_tablespace_databases(). Also adjust various comments and documentation to warn against expecting to clean up resources during a ValuePerCall SRF's final call. Back-patch to all supported branches, since these functions were all born broken. Justin Pryzby, with cosmetic tweaks by me Discussion: https://postgr.es/m/20200308173103.GC1357@telsasoft.com
2020-03-17 02:05:28 +01:00
values[0] = CStringGetTextDatum(pg_get_timezone_name(tz));
values[1] = CStringGetTextDatum(tzn ? tzn : "");
Rename the recently-added pg_timezonenames view to pg_timezone_abbrevs, and create a new view pg_timezone_names that provides information about the zones known in the 'zic' database. Magnus Hagander, with some additional work by Tom Lane.
2006-09-16 22:14:34 +02:00
Avoid holding a directory FD open across assorted SRF calls. This extends the fixes made in commit 085b6b667 to other SRFs with the same bug, namely pg_logdir_ls(), pgrowlocks(), pg_timezone_names(), pg_ls_dir(), and pg_tablespace_databases(). Also adjust various comments and documentation to warn against expecting to clean up resources during a ValuePerCall SRF's final call. Back-patch to all supported branches, since these functions were all born broken. Justin Pryzby, with cosmetic tweaks by me Discussion: https://postgr.es/m/20200308173103.GC1357@telsasoft.com
2020-03-17 02:05:28 +01:00
MemSet(&itm, 0, sizeof(struct pg_tm));
itm.tm_sec = -tzoff;
resInterval = (Interval *) palloc(sizeof(Interval));
tm2interval(&itm, 0, resInterval);
values[2] = IntervalPGetDatum(resInterval);
Rename the recently-added pg_timezonenames view to pg_timezone_abbrevs, and create a new view pg_timezone_names that provides information about the zones known in the 'zic' database. Magnus Hagander, with some additional work by Tom Lane.
2006-09-16 22:14:34 +02:00
Avoid holding a directory FD open across assorted SRF calls. This extends the fixes made in commit 085b6b667 to other SRFs with the same bug, namely pg_logdir_ls(), pgrowlocks(), pg_timezone_names(), pg_ls_dir(), and pg_tablespace_databases(). Also adjust various comments and documentation to warn against expecting to clean up resources during a ValuePerCall SRF's final call. Back-patch to all supported branches, since these functions were all born broken. Justin Pryzby, with cosmetic tweaks by me Discussion: https://postgr.es/m/20200308173103.GC1357@telsasoft.com
2020-03-17 02:05:28 +01:00
values[3] = BoolGetDatum(tm.tm_isdst > 0);
Rename the recently-added pg_timezonenames view to pg_timezone_abbrevs, and create a new view pg_timezone_names that provides information about the zones known in the 'zic' database. Magnus Hagander, with some additional work by Tom Lane.
2006-09-16 22:14:34 +02:00
Avoid holding a directory FD open across assorted SRF calls. This extends the fixes made in commit 085b6b667 to other SRFs with the same bug, namely pg_logdir_ls(), pgrowlocks(), pg_timezone_names(), pg_ls_dir(), and pg_tablespace_databases(). Also adjust various comments and documentation to warn against expecting to clean up resources during a ValuePerCall SRF's final call. Back-patch to all supported branches, since these functions were all born broken. Justin Pryzby, with cosmetic tweaks by me Discussion: https://postgr.es/m/20200308173103.GC1357@telsasoft.com
2020-03-17 02:05:28 +01:00
tuplestore_putvalues(tupstore, tupdesc, values, nulls);
}
Rename the recently-added pg_timezonenames view to pg_timezone_abbrevs, and create a new view pg_timezone_names that provides information about the zones known in the 'zic' database. Magnus Hagander, with some additional work by Tom Lane.
2006-09-16 22:14:34 +02:00
Avoid holding a directory FD open across assorted SRF calls. This extends the fixes made in commit 085b6b667 to other SRFs with the same bug, namely pg_logdir_ls(), pgrowlocks(), pg_timezone_names(), pg_ls_dir(), and pg_tablespace_databases(). Also adjust various comments and documentation to warn against expecting to clean up resources during a ValuePerCall SRF's final call. Back-patch to all supported branches, since these functions were all born broken. Justin Pryzby, with cosmetic tweaks by me Discussion: https://postgr.es/m/20200308173103.GC1357@telsasoft.com
2020-03-17 02:05:28 +01:00
pg_tzenumerate_end(tzenum);
return (Datum) 0;
Rename the recently-added pg_timezonenames view to pg_timezone_abbrevs, and create a new view pg_timezone_names that provides information about the zones known in the 'zic' database. Magnus Hagander, with some additional work by Tom Lane.
2006-09-16 22:14:34 +02:00
}