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/timestamp.c

5555 lines
135 KiB
C
Raw Normal View History

Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
/*-------------------------------------------------------------------------
*
* timestamp.c
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
* Functions for the built-in SQL types "timestamp" and "interval".
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +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
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
* Portions Copyright (c) 1994, Regents of the University of California
*
*
* IDENTIFICATION
Remove cvs keywords from all files.
2010-09-20 22:08:53 +02:00
* src/backend/utils/adt/timestamp.c
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
*
*-------------------------------------------------------------------------
*/
Second round of fmgr changes: triggers are now invoked in new style, CurrentTriggerData is history.
2000-05-29 03:59:17 +02:00
#include "postgres.h"
From: "D'Arcy J.M. Cain" <darcy@druid.net> Subject: [HACKERS] timestamp.c changes I sent in changes previously and they were rejected because they didn't follow ANSI spec. Here is the input part of the changes again. Even though it allows more flexibility for inputting different formats, it is also backwards compatible with the standard version. I have also not changed the output format so it will still output the ANSI forms. Is this acceptable to everyone?
1997-04-03 21:58:11 +02:00
#include <ctype.h>
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
#include <math.h>
#include <limits.h>
Fix build break on BSD, OSX, and other systems: add missing <sys/time.h> include.
2005-06-30 05:48:58 +02:00
#include <sys/time.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
Second round of fmgr changes: triggers are now invoked in new style, CurrentTriggerData is history.
2000-05-29 03:59:17 +02:00
#include "access/xact.h"
Modify array operations to include array's element type OID in the array header, and to compute sizing and alignment of array elements the same way normal tuple access operations do --- viz, using the tupmacs.h macros att_addlength and att_align. This makes the world safe for arrays of cstrings or intervals, and should make it much easier to write array-type-polymorphic functions; as examples see the cleanups of array_out and contrib/array_iterator. By Joe Conway and Tom Lane.
2002-08-26 19:54:02 +02:00
#include "catalog/pg_type.h"
Fix integer-overflow problems in interval comparison. When using integer timestamps, the interval-comparison functions tried to compute the overall magnitude of an interval as an int64 number of microseconds. As reported by Frazer McLean, this overflows for intervals exceeding about 296000 years, which is bad since we nominally allow intervals many times larger than that. That results in wrong comparison results, and possibly in corrupted btree indexes for columns containing such large interval values. To fix, compute the magnitude as int128 instead. Although some compilers have native support for int128 calculations, many don't, so create our own support functions that can do 128-bit addition and multiplication if the compiler support isn't there. These support functions are designed with an eye to allowing the int128 code paths in numeric.c to be rewritten for use on all platforms, although this patch doesn't do that, or even provide all the int128 primitives that will be needed for it. Back-patch as far as 9.4. Earlier releases did not guard against overflow of interval values at all (commit 146604ec4 fixed that), so it seems not very exciting to worry about overly-large intervals for them. Before 9.6, we did not assume that unreferenced "static inline" functions would not draw compiler warnings, so omit functions not directly referenced by timestamp.c, the only present consumer of int128.h. (We could have omitted these functions in HEAD too, but since they were written and debugged on the way to the present patch, and they look likely to be needed by numeric.c, let's keep them in HEAD.) I did not bother to try to prevent such warnings in a --disable-integer-datetimes build, though. Before 9.5, configure will never define HAVE_INT128, so the part of int128.h that exploits a native int128 implementation is dead code in the 9.4 branch. I didn't bother to remove it, thinking that keeping the file looking similar in different branches is more useful. In HEAD only, add a simple test harness for int128.h in src/tools/. In back branches, this does not change the float-timestamps code path. That's not subject to the same kind of overflow risk, since it computes the interval magnitude as float8. (No doubt, when this code was originally written, overflow was disregarded for exactly that reason.) There is a precision hazard instead :-(, but we'll avert our eyes from that question, since no complaints have been reported and that code's deprecated anyway. Kyotaro Horiguchi and Tom Lane Discussion: https://postgr.es/m/1490104629.422698.918452336.26FA96B7@webmail.messagingengine.com
2017-04-06 05:51:27 +02:00
#include "common/int128.h"
Add timestamp and timestamptz versions of generate_series(). Hitoshi Harada
2008-05-05 01:19:24 +02:00
#include "funcapi.h"
Add binary I/O routines for a bunch more datatypes. Still a few to go, but that was enough tedium for one day. Along the way, move the few support routines for types xid and cid into a more logical place.
2003-05-13 01:08:52 +02:00
#include "libpq/pqformat.h"
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
#include "miscadmin.h"
Add transform functions for AT TIME ZONE. This makes "ALTER TABLE tabname ALTER tscol TYPE ... USING tscol AT TIME ZONE 'UTC'" skip rewriting the table when altering from "timestamp" to "timestamptz" or vice versa. While it would be nicer still to optimize this in the absence of the USING clause given timezone==UTC, transform functions must consult IMMUTABLE facts only.
2015-03-01 19:22:34 +01:00
#include "nodes/makefuncs.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"
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
#include "nodes/supportnodes.h"
Solve the 'Turkish problem' with undesirable locale behavior for case conversion of basic ASCII letters. Remove all uses of strcasecmp and strncasecmp in favor of new functions pg_strcasecmp and pg_strncasecmp; remove most but not all direct uses of toupper and tolower in favor of pg_toupper and pg_tolower. These functions use the same notions of case folding already developed for identifier case conversion. I left the straight locale-based folding in place for situations where we are just manipulating user data and not trying to match it to built-in strings --- for example, the SQL upper() function is still locale dependent. Perhaps this will prove not to be what's wanted, but at the moment we can initdb and pass regression tests in Turkish locale.
2004-05-07 02:24:59 +02:00
#include "parser/scansup.h"
Revise aggregate functions per earlier discussions in pghackers. There's now only one transition value and transition function. NULL handling in aggregates is a lot cleaner. Also, use Numeric accumulators instead of integer accumulators for sum/avg on integer datatypes --- this avoids overflow at the cost of being a little slower. Implement VARIANCE() and STDDEV() aggregates in the standard backend. Also, enable new LIKE selectivity estimators by default. Unrelated change, but as long as I had to force initdb anyway...
2000-07-17 05:05:41 +02:00
#include "utils/array.h"
Missed another tar file... :(
1997-03-14 06:58:13 +01:00
#include "utils/builtins.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 "utils/datetime.h"
Provide separate header file for built-in float types Some data types under adt/ have separate header files, but most simple ones do not, and their public functions are defined in builtins.h. As the patches improving geometric types will require making additional functions public, this seems like a good opportunity to create a header for floats types. Commit 1acf757255 made _cmp functions public to solve NaN issues locally for GiST indexes. This patch reworks it in favour of a more widely applicable API. The API uses inline functions, as they are easier to use compared to macros, and avoid double-evaluation hazards. Author: Emre Hasegeli Reviewed-by: Kyotaro Horiguchi Discussion: https://www.postgresql.org/message-id/CAE2gYzxF7-5djV6-cEvqQu-fNsnt%3DEqbOURx7ZDg%2BVv6ZMTWbg%40mail.gmail.com
2018-07-29 03:30:48 +02:00
#include "utils/float.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
Add last-ditch defense against attempts to compile datetime code with -ffast-math.
2002-09-21 21:52:41 +02:00
/*
* gcc's -ffast-math switch breaks routines that expect exact results from
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
* expressions like timeval / SECS_PER_HOUR, where timeval is double.
Add last-ditch defense against attempts to compile datetime code with -ffast-math.
2002-09-21 21:52:41 +02:00
*/
#ifdef __FAST_MATH__
#error -ffast-math is known to break this code
#endif
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
#define SAMESIGN(a,b) (((a) < 0) == ((b) < 0))
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
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
/* Set at postmaster start */
TimestampTz PgStartTime;
8.4 pgindent run, with new combined Linux/FreeBSD/MinGW typedef list provided by Andrew.
2009-06-11 16:49:15 +02:00
Add pg_conf_load_time() function to report when the Postgres configuration files were last loaded. George Gensure
2008-05-04 23:13:36 +02:00
/* Set at configuration reload */
TimestampTz PgReloadTime;
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
Add timestamp and timestamptz versions of generate_series(). Hitoshi Harada
2008-05-05 01:19:24 +02:00
typedef struct
{
Timestamp current;
Timestamp finish;
Interval step;
int step_sign;
} generate_series_timestamp_fctx;
typedef struct
{
8.4 pgindent run, with new combined Linux/FreeBSD/MinGW typedef list provided by Andrew.
2009-06-11 16:49:15 +02:00
TimestampTz current;
TimestampTz finish;
Add timestamp and timestamptz versions of generate_series(). Hitoshi Harada
2008-05-05 01:19:24 +02:00
Interval step;
int step_sign;
} generate_series_timestamptz_fctx;
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
Get rid of a bunch of #ifdef HAVE_INT64_TIMESTAMP conditionals by inventing a new typedef TimeOffset to represent an intermediate time value. It's either int64 or double as appropriate, and in most usages will be measured in microseconds or seconds the same as Timestamp. We don't call it Timestamp, though, since the value doesn't necessarily represent an absolute time instant. Warren Turkal
2008-03-21 02:31:43 +01:00
static TimeOffset time2t(const int hour, const int min, const int sec, const fsec_t fsec);
Mark functions as static and ifdef NOT_USED as appropriate.
2000-06-09 00:38:00 +02:00
static Timestamp dt2local(Timestamp dt, int timezone);
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
static void AdjustIntervalForTypmod(Interval *interval, int32 typmod);
Add timestamp-versus-timestamptz cross-type comparison functions, flesh out the index operator classes to include these. In passing, fix erroneous volatility marking of ACL functions.
2004-03-22 02:38:18 +01:00
static TimestampTz timestamp2timestamptz(Timestamp timestamp);
Improve parsetree representation of special functions such as CURRENT_DATE. We implement a dozen or so parameterless functions that the SQL standard defines special syntax for. Up to now, that was done by converting them into more or less ad-hoc constructs such as "'now'::text::date". That's messy for multiple reasons: it exposes what should be implementation details to users, and performance is worse than it needs to be in several cases. To improve matters, invent a new expression node type SQLValueFunction that can represent any of these parameterless functions. Bump catversion because this changes stored parsetrees for rules. Discussion: <30058.1463091294@sss.pgh.pa.us>
2016-08-17 02:33:01 +02:00
static Timestamp timestamptz2timestamp(TimestampTz timestamp);
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
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
Support type modifiers for user-defined types, and pull most knowledge about typmod representation for standard types out into type-specific typmod I/O functions. Teodor Sigaev, with some editorialization by Tom Lane.
2006-12-30 22:21:56 +01:00
/* common code for timestamptypmodin and timestamptztypmodin */
static int32
anytimestamp_typmodin(bool istz, ArrayType *ta)
{
pgindent run for 8.3.
2007-11-15 22:14:46 +01:00
int32 *tl;
int n;
Support type modifiers for user-defined types, and pull most knowledge about typmod representation for standard types out into type-specific typmod I/O functions. Teodor Sigaev, with some editorialization by Tom Lane.
2006-12-30 22:21:56 +01:00
Tweak the API for per-datatype typmodin functions so that they are passed an array of strings rather than an array of integers, and allow any simple constant or identifier to be used in typmods; for example create table foo (f1 widget(42,'23skidoo',point)); Of course the typmodin function has still got to pack this info into a non-negative int32 for storage, but it's still a useful improvement in flexibility, especially considering that you can do nearly anything if you are willing to keep the info in a side table. We can get away with this change since we have not yet released a version providing user-definable typmods. Per discussion.
2007-06-15 22:56:52 +02:00
tl = ArrayGetIntegerTypmods(ta, &n);
Support type modifiers for user-defined types, and pull most knowledge about typmod representation for standard types out into type-specific typmod I/O functions. Teodor Sigaev, with some editorialization by Tom Lane.
2006-12-30 22:21:56 +01:00
/*
* we're not too tense about good error message here because grammar
* shouldn't allow wrong number of modifiers for TIMESTAMP
*/
if (n != 1)
ereport(ERROR,
(errcode(ERRCODE_INVALID_PARAMETER_VALUE),
errmsg("invalid type modifier")));
Improve parsetree representation of special functions such as CURRENT_DATE. We implement a dozen or so parameterless functions that the SQL standard defines special syntax for. Up to now, that was done by converting them into more or less ad-hoc constructs such as "'now'::text::date". That's messy for multiple reasons: it exposes what should be implementation details to users, and performance is worse than it needs to be in several cases. To improve matters, invent a new expression node type SQLValueFunction that can represent any of these parameterless functions. Bump catversion because this changes stored parsetrees for rules. Discussion: <30058.1463091294@sss.pgh.pa.us>
2016-08-17 02:33:01 +02:00
return anytimestamp_typmod_check(istz, tl[0]);
}
/* exported so parse_expr.c can use it */
int32
anytimestamp_typmod_check(bool istz, int32 typmod)
{
if (typmod < 0)
Support type modifiers for user-defined types, and pull most knowledge about typmod representation for standard types out into type-specific typmod I/O functions. Teodor Sigaev, with some editorialization by Tom Lane.
2006-12-30 22:21:56 +01:00
ereport(ERROR,
(errcode(ERRCODE_INVALID_PARAMETER_VALUE),
errmsg("TIMESTAMP(%d)%s precision must not be negative",
Improve parsetree representation of special functions such as CURRENT_DATE. We implement a dozen or so parameterless functions that the SQL standard defines special syntax for. Up to now, that was done by converting them into more or less ad-hoc constructs such as "'now'::text::date". That's messy for multiple reasons: it exposes what should be implementation details to users, and performance is worse than it needs to be in several cases. To improve matters, invent a new expression node type SQLValueFunction that can represent any of these parameterless functions. Bump catversion because this changes stored parsetrees for rules. Discussion: <30058.1463091294@sss.pgh.pa.us>
2016-08-17 02:33:01 +02:00
typmod, (istz ? " WITH TIME ZONE" : ""))));
if (typmod > MAX_TIMESTAMP_PRECISION)
Support type modifiers for user-defined types, and pull most knowledge about typmod representation for standard types out into type-specific typmod I/O functions. Teodor Sigaev, with some editorialization by Tom Lane.
2006-12-30 22:21:56 +01:00
{
ereport(WARNING,
(errcode(ERRCODE_INVALID_PARAMETER_VALUE),
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
errmsg("TIMESTAMP(%d)%s precision reduced to maximum allowed, %d",
typmod, (istz ? " WITH TIME ZONE" : ""),
MAX_TIMESTAMP_PRECISION)));
Support type modifiers for user-defined types, and pull most knowledge about typmod representation for standard types out into type-specific typmod I/O functions. Teodor Sigaev, with some editorialization by Tom Lane.
2006-12-30 22:21:56 +01:00
typmod = MAX_TIMESTAMP_PRECISION;
pgindent run for 8.3.
2007-11-15 22:14:46 +01:00
}
Support type modifiers for user-defined types, and pull most knowledge about typmod representation for standard types out into type-specific typmod I/O functions. Teodor Sigaev, with some editorialization by Tom Lane.
2006-12-30 22:21:56 +01:00
return typmod;
}
/* common code for timestamptypmodout and timestamptztypmodout */
static char *
anytimestamp_typmodout(bool istz, int32 typmod)
{
const char *tz = istz ? " with time zone" : " without time zone";
if (typmod >= 0)
Add more use of psprintf()
2014-01-07 03:30:26 +01:00
return psprintf("(%d)%s", (int) typmod, tz);
Support type modifiers for user-defined types, and pull most knowledge about typmod representation for standard types out into type-specific typmod I/O functions. Teodor Sigaev, with some editorialization by Tom Lane.
2006-12-30 22:21:56 +01:00
else
Add more use of psprintf()
2014-01-07 03:30:26 +01:00
return psprintf("%s", tz);
Support type modifiers for user-defined types, and pull most knowledge about typmod representation for standard types out into type-specific typmod I/O functions. Teodor Sigaev, with some editorialization by Tom Lane.
2006-12-30 22:21:56 +01:00
}
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
/*****************************************************************************
* USER I/O ROUTINES *
*****************************************************************************/
/* timestamp_in()
* Convert a string to internal form.
*/
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
Datum
timestamp_in(PG_FUNCTION_ARGS)
Missed another tar file... :(
1997-03-14 06:58:13 +01:00
{
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
char *str = PG_GETARG_CSTRING(0);
pgindent run on all C files. Java run to follow. initdb/regression tests pass.
2001-10-25 07:50:21 +02:00
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
#ifdef NOT_USED
Oid typelem = PG_GETARG_OID(1);
#endif
int32 typmod = PG_GETARG_INT32(2);
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
Timestamp result;
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;
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
struct pg_tm tt,
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
*tm = &tt;
int tz;
int dtype;
int nf;
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;
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
char *field[MAXDATEFIELDS];
int ftype[MAXDATEFIELDS];
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 workbuf[MAXDATELEN + MAXDATEFIELDS];
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
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
dterr = ParseDateTime(str, workbuf, sizeof(workbuf),
field, ftype, MAXDATEFIELDS, &nf);
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 == 0)
dterr = DecodeDateTime(field, ftype, nf, &dtype, tm, &fsec, &tz);
if (dterr != 0)
DateTimeParseError(dterr, str, "timestamp");
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
switch (dtype)
{
case DTK_DATE:
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 (tm2timestamp(tm, fsec, NULL, &result) != 0)
Error message editing in utils/adt. Again thanks to Joe Conway for doing the bulk of the heavy lifting ...
2003-07-27 06:53:12 +02:00
ereport(ERROR,
(errcode(ERRCODE_DATETIME_VALUE_OUT_OF_RANGE),
errmsg("timestamp out of range: \"%s\"", str)));
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
break;
case DTK_EPOCH:
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
result = SetEpochTimestamp();
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
break;
case DTK_LATE:
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
TIMESTAMP_NOEND(result);
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
break;
case DTK_EARLY:
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
TIMESTAMP_NOBEGIN(result);
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
break;
default:
Error message editing in utils/adt. Again thanks to Joe Conway for doing the bulk of the heavy lifting ...
2003-07-27 06:53:12 +02:00
elog(ERROR, "unexpected dtype %d while parsing timestamp \"%s\"",
dtype, 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
TIMESTAMP_NOEND(result);
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
}
Missed another tar file... :(
1997-03-14 06:58:13 +01:00
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
AdjustTimestampForTypmod(&result, typmod);
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
PG_RETURN_TIMESTAMP(result);
}
Missed another tar file... :(
1997-03-14 06:58:13 +01:00
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
/* timestamp_out()
* Convert a timestamp to external form.
*/
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
Datum
timestamp_out(PG_FUNCTION_ARGS)
Missed another tar file... :(
1997-03-14 06:58:13 +01:00
{
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
Timestamp timestamp = PG_GETARG_TIMESTAMP(0);
Another PGINDENT run that changes variable indenting and case label indenting. Also static variable indenting.
1997-09-08 04:41:22 +02:00
char *result;
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
struct pg_tm tt,
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 = &tt;
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;
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
char buf[MAXDATELEN + 1];
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
if (TIMESTAMP_NOT_FINITE(timestamp))
EncodeSpecialTimestamp(timestamp, buf);
This patch makes it possible to use the full set of timezones when doing "AT TIME ZONE", and not just the shorlist previously available. For example: SELECT CURRENT_TIMESTAMP AT TIME ZONE 'Europe/London'; works fine now. It will also obey whatever DST rules were in effect at just that date, which the previous implementation did not. It also supports the AT TIME ZONE on the timetz datatype. The whole handling of DST is a bit bogus there, so I chose to make it use whatever DST rules are in effect at the time of executig the query. not sure if anybody is actuallyi *using* timetz though, it seems pretty unpredictable just because of this... Magnus Hagander
2005-06-15 02:34:11 +02:00
else if (timestamp2tm(timestamp, NULL, tm, &fsec, NULL, NULL) == 0)
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
EncodeDateTime(tm, fsec, false, 0, NULL, DateStyle, buf);
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
Error message editing in utils/adt. Again thanks to Joe Conway for doing the bulk of the heavy lifting ...
2003-07-27 06:53:12 +02:00
ereport(ERROR,
(errcode(ERRCODE_DATETIME_VALUE_OUT_OF_RANGE),
errmsg("timestamp out of range")));
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
result = pstrdup(buf);
PG_RETURN_CSTRING(result);
}
Add binary I/O routines for a bunch more datatypes. Still a few to go, but that was enough tedium for one day. Along the way, move the few support routines for types xid and cid into a more logical place.
2003-05-13 01:08:52 +02:00
/*
* timestamp_recv - converts external binary format to timestamp
*/
Datum
timestamp_recv(PG_FUNCTION_ARGS)
{
StringInfo buf = (StringInfo) PG_GETARG_POINTER(0);
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
Change typreceive function API so that receive functions get the same optional arguments as text input functions, ie, typioparam OID and atttypmod. Make all the datatypes that use typmod enforce it the same way in typreceive as they do in typinput. This fixes a problem with failure to enforce length restrictions during COPY FROM BINARY.
2005-07-10 23:14:00 +02:00
#ifdef NOT_USED
Oid typelem = PG_GETARG_OID(1);
#endif
int32 typmod = PG_GETARG_INT32(2);
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
Timestamp timestamp;
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
struct pg_tm tt,
Add range-checking in timestamp_recv and timestamptz_recv, per Stephen Frost. Also tighten date range check in timestamp2tm.
2004-06-03 19:57:09 +02:00
*tm = &tt;
fsec_t fsec;
Add binary I/O routines for a bunch more datatypes. Still a few to go, but that was enough tedium for one day. Along the way, move the few support routines for types xid and cid into a more logical place.
2003-05-13 01:08:52 +02:00
Add range-checking in timestamp_recv and timestamptz_recv, per Stephen Frost. Also tighten date range check in timestamp2tm.
2004-06-03 19:57:09 +02:00
timestamp = (Timestamp) pq_getmsgint64(buf);
Various minor corrections of and improvements to comments. Aleksander Alekseev
2016-03-18 14:38:59 +01:00
/* range check: see if timestamp_out would like it */
Add range-checking in timestamp_recv and timestamptz_recv, per Stephen Frost. Also tighten date range check in timestamp2tm.
2004-06-03 19:57:09 +02:00
if (TIMESTAMP_NOT_FINITE(timestamp))
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
/* ok */ ;
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
else if (timestamp2tm(timestamp, NULL, tm, &fsec, NULL, NULL) != 0 ||
!IS_VALID_TIMESTAMP(timestamp))
Add range-checking in timestamp_recv and timestamptz_recv, per Stephen Frost. Also tighten date range check in timestamp2tm.
2004-06-03 19:57:09 +02:00
ereport(ERROR,
(errcode(ERRCODE_DATETIME_VALUE_OUT_OF_RANGE),
errmsg("timestamp out of range")));
Change typreceive function API so that receive functions get the same optional arguments as text input functions, ie, typioparam OID and atttypmod. Make all the datatypes that use typmod enforce it the same way in typreceive as they do in typinput. This fixes a problem with failure to enforce length restrictions during COPY FROM BINARY.
2005-07-10 23:14:00 +02:00
AdjustTimestampForTypmod(&timestamp, typmod);
Add range-checking in timestamp_recv and timestamptz_recv, per Stephen Frost. Also tighten date range check in timestamp2tm.
2004-06-03 19:57:09 +02:00
PG_RETURN_TIMESTAMP(timestamp);
Add binary I/O routines for a bunch more datatypes. Still a few to go, but that was enough tedium for one day. Along the way, move the few support routines for types xid and cid into a more logical place.
2003-05-13 01:08:52 +02:00
}
/*
* timestamp_send - converts timestamp to binary format
*/
Datum
timestamp_send(PG_FUNCTION_ARGS)
{
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
Timestamp timestamp = PG_GETARG_TIMESTAMP(0);
Add binary I/O routines for a bunch more datatypes. Still a few to go, but that was enough tedium for one day. Along the way, move the few support routines for types xid and cid into a more logical place.
2003-05-13 01:08:52 +02:00
StringInfoData buf;
pq_begintypsend(&buf);
pq_sendint64(&buf, timestamp);
PG_RETURN_BYTEA_P(pq_endtypsend(&buf));
}
Support type modifiers for user-defined types, and pull most knowledge about typmod representation for standard types out into type-specific typmod I/O functions. Teodor Sigaev, with some editorialization by Tom Lane.
2006-12-30 22:21:56 +01:00
Datum
timestamptypmodin(PG_FUNCTION_ARGS)
{
pgindent run for 8.3.
2007-11-15 22:14:46 +01:00
ArrayType *ta = PG_GETARG_ARRAYTYPE_P(0);
Support type modifiers for user-defined types, and pull most knowledge about typmod representation for standard types out into type-specific typmod I/O functions. Teodor Sigaev, with some editorialization by Tom Lane.
2006-12-30 22:21:56 +01:00
PG_RETURN_INT32(anytimestamp_typmodin(false, ta));
}
Datum
timestamptypmodout(PG_FUNCTION_ARGS)
{
pgindent run for 8.3.
2007-11-15 22:14:46 +01:00
int32 typmod = PG_GETARG_INT32(0);
Support type modifiers for user-defined types, and pull most knowledge about typmod representation for standard types out into type-specific typmod I/O functions. Teodor Sigaev, with some editorialization by Tom Lane.
2006-12-30 22:21:56 +01:00
PG_RETURN_CSTRING(anytimestamp_typmodout(false, typmod));
}
Add binary I/O routines for a bunch more datatypes. Still a few to go, but that was enough tedium for one day. Along the way, move the few support routines for types xid and cid into a more logical place.
2003-05-13 01:08:52 +02:00
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
/*
* timestamp_support()
*
* Planner support function for the timestamp_scale() and timestamptz_scale()
* length coercion functions (we need not distinguish them here).
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
*/
Datum
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
timestamp_support(PG_FUNCTION_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
{
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
Node *rawreq = (Node *) PG_GETARG_POINTER(0);
Node *ret = NULL;
if (IsA(rawreq, SupportRequestSimplify))
{
SupportRequestSimplify *req = (SupportRequestSimplify *) rawreq;
ret = TemporalSimplify(MAX_TIMESTAMP_PRECISION, (Node *) req->fcall);
}
PG_RETURN_POINTER(ret);
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
}
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
/* timestamp_scale()
* Adjust time type for specified scale factor.
* Used by PostgreSQL type system to stuff columns.
*/
Datum
timestamp_scale(PG_FUNCTION_ARGS)
{
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
Timestamp timestamp = PG_GETARG_TIMESTAMP(0);
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
int32 typmod = PG_GETARG_INT32(1);
Timestamp result;
result = timestamp;
AdjustTimestampForTypmod does not work (at least not portably) on -infinity and +infinity. Put TIMESTAMP_NOT_FINITE guard into the routine, instead of forgetting it at some call sites. Fixes regression test failures here.
2001-10-03 17:50:48 +02:00
AdjustTimestampForTypmod(&result, typmod);
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
PG_RETURN_TIMESTAMP(result);
}
Improve parsetree representation of special functions such as CURRENT_DATE. We implement a dozen or so parameterless functions that the SQL standard defines special syntax for. Up to now, that was done by converting them into more or less ad-hoc constructs such as "'now'::text::date". That's messy for multiple reasons: it exposes what should be implementation details to users, and performance is worse than it needs to be in several cases. To improve matters, invent a new expression node type SQLValueFunction that can represent any of these parameterless functions. Bump catversion because this changes stored parsetrees for rules. Discussion: <30058.1463091294@sss.pgh.pa.us>
2016-08-17 02:33:01 +02:00
/*
Error suppression support for upcoming jsonpath .datetime() method Add support of error suppression in some date and time manipulation functions as it's required for jsonpath .datetime() method support. This commit doesn't use PG_TRY()/PG_CATCH() in order to implement that. Instead, it provides internal versions of date and time functions used, which support error suppression. Discussion: https://postgr.es/m/CAPpHfdsZgYEra_PeCLGNoXOWYx6iU-S3wF8aX0ObQUcZU%2B4XTw%40mail.gmail.com Author: Alexander Korotkov, Nikita Glukhov Reviewed-by: Anastasia Lubennikova, Peter Eisentraut
2019-09-25 20:51:47 +02:00
* AdjustTimestampForTypmodError --- round off a timestamp to suit given typmod
Improve parsetree representation of special functions such as CURRENT_DATE. We implement a dozen or so parameterless functions that the SQL standard defines special syntax for. Up to now, that was done by converting them into more or less ad-hoc constructs such as "'now'::text::date". That's messy for multiple reasons: it exposes what should be implementation details to users, and performance is worse than it needs to be in several cases. To improve matters, invent a new expression node type SQLValueFunction that can represent any of these parameterless functions. Bump catversion because this changes stored parsetrees for rules. Discussion: <30058.1463091294@sss.pgh.pa.us>
2016-08-17 02:33:01 +02:00
* Works for either timestamp or timestamptz.
*/
Error suppression support for upcoming jsonpath .datetime() method Add support of error suppression in some date and time manipulation functions as it's required for jsonpath .datetime() method support. This commit doesn't use PG_TRY()/PG_CATCH() in order to implement that. Instead, it provides internal versions of date and time functions used, which support error suppression. Discussion: https://postgr.es/m/CAPpHfdsZgYEra_PeCLGNoXOWYx6iU-S3wF8aX0ObQUcZU%2B4XTw%40mail.gmail.com Author: Alexander Korotkov, Nikita Glukhov Reviewed-by: Anastasia Lubennikova, Peter Eisentraut
2019-09-25 20:51:47 +02:00
bool
AdjustTimestampForTypmodError(Timestamp *time, int32 typmod, bool *error)
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
{
pgindent run.
2002-09-04 22:31:48 +02:00
static const int64 TimestampScales[MAX_TIMESTAMP_PRECISION + 1] = {
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
INT64CONST(1000000),
INT64CONST(100000),
INT64CONST(10000),
INT64CONST(1000),
INT64CONST(100),
INT64CONST(10),
INT64CONST(1)
};
pgindent run.
2002-09-04 22:31:48 +02:00
static const int64 TimestampOffsets[MAX_TIMESTAMP_PRECISION + 1] = {
Repair inconsistent rounding behavior for timestamp, time, interval, per gripe from Csaba Nagy. There is still potential for platform-specific behavior for values that are exactly halfway between integers, but at least we now get the expected answer for all other cases.
2003-01-09 02:06:57 +01:00
INT64CONST(500000),
INT64CONST(50000),
INT64CONST(5000),
INT64CONST(500),
INT64CONST(50),
INT64CONST(5),
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
INT64CONST(0)
};
if (!TIMESTAMP_NOT_FINITE(*time)
&& (typmod != -1) && (typmod != MAX_TIMESTAMP_PRECISION))
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
{
Remove unnecessary parentheses in datetime/timestamp code.
2005-05-23 19:13:14 +02:00
if (typmod < 0 || typmod > MAX_TIMESTAMP_PRECISION)
Error suppression support for upcoming jsonpath .datetime() method Add support of error suppression in some date and time manipulation functions as it's required for jsonpath .datetime() method support. This commit doesn't use PG_TRY()/PG_CATCH() in order to implement that. Instead, it provides internal versions of date and time functions used, which support error suppression. Discussion: https://postgr.es/m/CAPpHfdsZgYEra_PeCLGNoXOWYx6iU-S3wF8aX0ObQUcZU%2B4XTw%40mail.gmail.com Author: Alexander Korotkov, Nikita Glukhov Reviewed-by: Anastasia Lubennikova, Peter Eisentraut
2019-09-25 20:51:47 +02:00
{
if (error)
{
*error = true;
return false;
}
Error message editing in utils/adt. Again thanks to Joe Conway for doing the bulk of the heavy lifting ...
2003-07-27 06:53:12 +02:00
ereport(ERROR,
(errcode(ERRCODE_INVALID_PARAMETER_VALUE),
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
errmsg("timestamp(%d) precision must be between %d and %d",
typmod, 0, MAX_TIMESTAMP_PRECISION)));
Error suppression support for upcoming jsonpath .datetime() method Add support of error suppression in some date and time manipulation functions as it's required for jsonpath .datetime() method support. This commit doesn't use PG_TRY()/PG_CATCH() in order to implement that. Instead, it provides internal versions of date and time functions used, which support error suppression. Discussion: https://postgr.es/m/CAPpHfdsZgYEra_PeCLGNoXOWYx6iU-S3wF8aX0ObQUcZU%2B4XTw%40mail.gmail.com Author: Alexander Korotkov, Nikita Glukhov Reviewed-by: Anastasia Lubennikova, Peter Eisentraut
2019-09-25 20:51:47 +02:00
}
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
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
if (*time >= INT64CONST(0))
{
Remove more extraneous parentheses in date/time functions.
2005-05-24 04:09:45 +02:00
*time = ((*time + TimestampOffsets[typmod]) / TimestampScales[typmod]) *
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
TimestampScales[typmod];
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
{
pgindent run.
2003-08-04 02:43:34 +02:00
*time = -((((-*time) + TimestampOffsets[typmod]) / TimestampScales[typmod])
* TimestampScales[typmod]);
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
}
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
}
Error suppression support for upcoming jsonpath .datetime() method Add support of error suppression in some date and time manipulation functions as it's required for jsonpath .datetime() method support. This commit doesn't use PG_TRY()/PG_CATCH() in order to implement that. Instead, it provides internal versions of date and time functions used, which support error suppression. Discussion: https://postgr.es/m/CAPpHfdsZgYEra_PeCLGNoXOWYx6iU-S3wF8aX0ObQUcZU%2B4XTw%40mail.gmail.com Author: Alexander Korotkov, Nikita Glukhov Reviewed-by: Anastasia Lubennikova, Peter Eisentraut
2019-09-25 20:51:47 +02:00
return true;
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
}
Error suppression support for upcoming jsonpath .datetime() method Add support of error suppression in some date and time manipulation functions as it's required for jsonpath .datetime() method support. This commit doesn't use PG_TRY()/PG_CATCH() in order to implement that. Instead, it provides internal versions of date and time functions used, which support error suppression. Discussion: https://postgr.es/m/CAPpHfdsZgYEra_PeCLGNoXOWYx6iU-S3wF8aX0ObQUcZU%2B4XTw%40mail.gmail.com Author: Alexander Korotkov, Nikita Glukhov Reviewed-by: Anastasia Lubennikova, Peter Eisentraut
2019-09-25 20:51:47 +02:00
void
AdjustTimestampForTypmod(Timestamp *time, int32 typmod)
{
(void) AdjustTimestampForTypmodError(time, typmod, 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
/* timestamptz_in()
* Convert a string to internal form.
*/
Datum
timestamptz_in(PG_FUNCTION_ARGS)
{
char *str = PG_GETARG_CSTRING(0);
pgindent run on all C files. Java run to follow. initdb/regression tests pass.
2001-10-25 07:50:21 +02:00
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
#ifdef NOT_USED
Oid typelem = PG_GETARG_OID(1);
#endif
int32 typmod = PG_GETARG_INT32(2);
pgindent run on all C files. Java run to follow. initdb/regression tests pass.
2001-10-25 07:50:21 +02:00
TimestampTz result;
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;
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
struct pg_tm tt,
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 = &tt;
int tz;
int dtype;
int nf;
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;
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
char *field[MAXDATEFIELDS];
int ftype[MAXDATEFIELDS];
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 workbuf[MAXDATELEN + MAXDATEFIELDS];
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
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
dterr = ParseDateTime(str, workbuf, sizeof(workbuf),
field, ftype, MAXDATEFIELDS, &nf);
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 == 0)
dterr = DecodeDateTime(field, ftype, nf, &dtype, tm, &fsec, &tz);
if (dterr != 0)
DateTimeParseError(dterr, str, "timestamp with time zone");
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
switch (dtype)
{
case DTK_DATE:
if (tm2timestamp(tm, fsec, &tz, &result) != 0)
Error message editing in utils/adt. Again thanks to Joe Conway for doing the bulk of the heavy lifting ...
2003-07-27 06:53:12 +02:00
ereport(ERROR,
(errcode(ERRCODE_DATETIME_VALUE_OUT_OF_RANGE),
errmsg("timestamp out of range: \"%s\"", 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
break;
case DTK_EPOCH:
result = SetEpochTimestamp();
break;
case DTK_LATE:
TIMESTAMP_NOEND(result);
break;
case DTK_EARLY:
TIMESTAMP_NOBEGIN(result);
break;
default:
Error message editing in utils/adt. Again thanks to Joe Conway for doing the bulk of the heavy lifting ...
2003-07-27 06:53:12 +02:00
elog(ERROR, "unexpected dtype %d while parsing timestamptz \"%s\"",
dtype, 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
TIMESTAMP_NOEND(result);
}
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
AdjustTimestampForTypmod(&result, typmod);
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
PG_RETURN_TIMESTAMPTZ(result);
}
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
/*
* Try to parse a timezone specification, and return its timezone offset value
* if it's acceptable. Otherwise, an error is thrown.
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
*
* Note: some code paths update tm->tm_isdst, and some don't; current callers
* don't care, so we don't bother being consistent.
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
*/
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
parse_sane_timezone(struct pg_tm *tm, text *zone)
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
{
char tzname[TZ_STRLEN_MAX + 1];
int rt;
int tz;
text_to_cstring_buffer(zone, tzname, sizeof(tzname));
/*
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
* Look up the requested timezone. First we try to interpret it as a
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
* numeric timezone specification; if DecodeTimezone decides it doesn'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
* like the format, we look in the timezone abbreviation table (to handle
* cases like "EST"), and if that also fails, we look in the timezone
* database (to handle cases like "America/New_York"). (This matches the
* order in which timestamp input checks the cases; it's important because
* the timezone database unwisely uses a few zone names that are identical
* to offset abbreviations.)
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
*
* Note pg_tzset happily parses numeric input that DecodeTimezone would
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
* reject. To avoid having it accept input that would otherwise be seen
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
* as invalid, it's enough to disallow having a digit in the first
* position of our input string.
*/
isdigit() needs an unsigned char argument. Per the C standard, the routine should be passed an int, with a value that's representable as an unsigned char or EOF. Passing a signed char is wrong, because a negative value is not representable as an unsigned char. Unfortunately no compiler warns about that.
2014-03-06 20:19:00 +01:00
if (isdigit((unsigned char) *tzname))
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
ereport(ERROR,
(errcode(ERRCODE_INVALID_PARAMETER_VALUE),
Deduplicate "invalid input syntax" messages for various types. Previously a lot of the error messages referenced the type in the error message itself. That requires that the message is translated separately for each type. Note that currently a few smallint cases continue to reference the integer, rather than smallint, type. A later patch will create a separate routine for 16bit input. Author: Andres Freund Discussion: https://postgr.es/m/20180707200158.wpqkd7rjr4jxq5g7@alap3.anarazel.de
2018-07-22 23:58:01 +02:00
errmsg("invalid input syntax for type %s: \"%s\"",
"numeric time zone", tzname),
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
errhint("Numeric time zones must have \"-\" or \"+\" as first character.")));
rt = DecodeTimezone(tzname, &tz);
if (rt != 0)
{
char *lowzone;
int type,
val;
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 *tzp;
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
if (rt == DTERR_TZDISP_OVERFLOW)
ereport(ERROR,
(errcode(ERRCODE_INVALID_PARAMETER_VALUE),
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
errmsg("numeric time zone \"%s\" out of range", tzname)));
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
else if (rt != DTERR_BAD_FORMAT)
ereport(ERROR,
(errcode(ERRCODE_INVALID_PARAMETER_VALUE),
errmsg("time zone \"%s\" not recognized", tzname)));
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 requires lowercase input */
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
lowzone = downcase_truncate_identifier(tzname,
strlen(tzname),
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
type = DecodeTimezoneAbbrev(0, lowzone, &val, &tzp);
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
if (type == TZ || type == DTZ)
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
{
/* fixed-offset abbreviation */
tz = -val;
}
else if (type == DYNTZ)
{
/* dynamic-offset abbreviation, resolve using specified time */
tz = DetermineTimeZoneAbbrevOffset(tm, tzname, tzp);
}
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
else
{
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
/* try it as a full zone name */
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
tzp = pg_tzset(tzname);
if (tzp)
tz = DetermineTimeZoneOffset(tm, tzp);
else
ereport(ERROR,
(errcode(ERRCODE_INVALID_PARAMETER_VALUE),
errmsg("time zone \"%s\" not recognized", tzname)));
}
}
return tz;
}
/*
* make_timestamp_internal
* workhorse for make_timestamp and make_timestamptz
*/
static Timestamp
make_timestamp_internal(int year, int month, int day,
int hour, int min, double sec)
{
struct pg_tm tm;
TimeOffset date;
TimeOffset time;
int dterr;
Timestamp result;
tm.tm_year = year;
tm.tm_mon = month;
tm.tm_mday = day;
/*
* Note: we'll reject zero or negative year values. Perhaps negatives
* should be allowed to represent BC years?
*/
dterr = ValidateDate(DTK_DATE_M, false, false, false, &tm);
if (dterr != 0)
ereport(ERROR,
(errcode(ERRCODE_DATETIME_FIELD_OVERFLOW),
errmsg("date field value out of range: %d-%02d-%02d",
year, month, day)));
if (!IS_VALID_JULIAN(tm.tm_year, tm.tm_mon, tm.tm_mday))
ereport(ERROR,
(errcode(ERRCODE_DATETIME_VALUE_OUT_OF_RANGE),
errmsg("date out of range: %d-%02d-%02d",
year, month, day)));
date = date2j(tm.tm_year, tm.tm_mon, tm.tm_mday) - POSTGRES_EPOCH_JDATE;
Fix portability issues in recently added make_timestamp/make_interval code. Explicitly reject infinity/NaN inputs, rather than just assuming that something else will do it for us. Per buildfarm. While at it, make some over-parenthesized and under-legible code more readable.
2014-03-05 22:42:18 +01:00
/*
* This should match the checks in DecodeTimeOnly, except that since we're
* dealing with a float "sec" value, we also explicitly reject NaN. (An
* infinity input should get rejected by the range comparisons, but we
* can't be sure how those will treat a NaN.)
*/
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
if (hour < 0 || min < 0 || min > MINS_PER_HOUR - 1 ||
Fix portability issues in recently added make_timestamp/make_interval code. Explicitly reject infinity/NaN inputs, rather than just assuming that something else will do it for us. Per buildfarm. While at it, make some over-parenthesized and under-legible code more readable.
2014-03-05 22:42:18 +01:00
isnan(sec) ||
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
sec < 0 || sec > SECS_PER_MINUTE ||
hour > HOURS_PER_DAY ||
/* test for > 24:00:00 */
(hour == HOURS_PER_DAY && (min > 0 || sec > 0)))
ereport(ERROR,
(errcode(ERRCODE_DATETIME_FIELD_OVERFLOW),
errmsg("time field value out of range: %d:%02d:%02g",
hour, min, sec)));
/* This should match tm2time */
time = (((hour * MINS_PER_HOUR + min) * SECS_PER_MINUTE)
* USECS_PER_SEC) + rint(sec * USECS_PER_SEC);
result = date * USECS_PER_DAY + time;
/* check for major overflow */
if ((result - time) / USECS_PER_DAY != date)
ereport(ERROR,
(errcode(ERRCODE_DATETIME_VALUE_OUT_OF_RANGE),
errmsg("timestamp out of range: %d-%02d-%02d %d:%02d:%02g",
year, month, day,
hour, min, sec)));
/* check for just-barely overflow (okay except time-of-day wraps) */
/* caution: we want to allow 1999-12-31 24:00:00 */
if ((result < 0 && date > 0) ||
(result > 0 && date < -1))
ereport(ERROR,
(errcode(ERRCODE_DATETIME_VALUE_OUT_OF_RANGE),
errmsg("timestamp out of range: %d-%02d-%02d %d:%02d:%02g",
year, month, day,
hour, min, sec)));
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
/* final range check catches just-out-of-range timestamps */
if (!IS_VALID_TIMESTAMP(result))
ereport(ERROR,
(errcode(ERRCODE_DATETIME_VALUE_OUT_OF_RANGE),
errmsg("timestamp out of range: %d-%02d-%02d %d:%02d:%02g",
year, month, day,
hour, min, sec)));
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
return result;
}
/*
* make_timestamp() - timestamp constructor
*/
Datum
make_timestamp(PG_FUNCTION_ARGS)
{
int32 year = PG_GETARG_INT32(0);
int32 month = PG_GETARG_INT32(1);
int32 mday = PG_GETARG_INT32(2);
int32 hour = PG_GETARG_INT32(3);
int32 min = PG_GETARG_INT32(4);
float8 sec = PG_GETARG_FLOAT8(5);
Timestamp result;
result = make_timestamp_internal(year, month, mday,
hour, min, sec);
PG_RETURN_TIMESTAMP(result);
}
/*
* make_timestamptz() - timestamp with time zone constructor
*/
Datum
make_timestamptz(PG_FUNCTION_ARGS)
{
int32 year = PG_GETARG_INT32(0);
int32 month = PG_GETARG_INT32(1);
int32 mday = PG_GETARG_INT32(2);
int32 hour = PG_GETARG_INT32(3);
int32 min = PG_GETARG_INT32(4);
float8 sec = PG_GETARG_FLOAT8(5);
Timestamp result;
result = make_timestamp_internal(year, month, mday,
hour, min, sec);
PG_RETURN_TIMESTAMPTZ(timestamp2timestamptz(result));
}
/*
* Construct a timestamp with time zone.
* As above, but the time zone is specified as seventh argument.
*/
Datum
make_timestamptz_at_timezone(PG_FUNCTION_ARGS)
{
int32 year = PG_GETARG_INT32(0);
int32 month = PG_GETARG_INT32(1);
int32 mday = PG_GETARG_INT32(2);
int32 hour = PG_GETARG_INT32(3);
int32 min = PG_GETARG_INT32(4);
float8 sec = PG_GETARG_FLOAT8(5);
text *zone = PG_GETARG_TEXT_PP(6);
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
TimestampTz result;
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
Timestamp timestamp;
struct pg_tm tt;
int tz;
fsec_t fsec;
timestamp = make_timestamp_internal(year, month, mday,
hour, min, sec);
if (timestamp2tm(timestamp, NULL, &tt, &fsec, NULL, NULL) != 0)
ereport(ERROR,
(errcode(ERRCODE_DATETIME_VALUE_OUT_OF_RANGE),
errmsg("timestamp out of range")));
tz = parse_sane_timezone(&tt, zone);
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
result = dt2local(timestamp, -tz);
if (!IS_VALID_TIMESTAMP(result))
ereport(ERROR,
(errcode(ERRCODE_DATETIME_VALUE_OUT_OF_RANGE),
errmsg("timestamp out of range")));
PG_RETURN_TIMESTAMPTZ(result);
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
}
Allow to_timestamp(float8) to convert float infinity to timestamp infinity. With the original SQL-function implementation, such cases failed because we don't support infinite intervals. Converting the function to C lets us bypass the interval representation, which should be a bit faster as well as more flexible. Vitaly Burovoy, reviewed by Anastasia Lubennikova
2016-03-29 23:09:21 +02:00
/*
* to_timestamp(double precision)
* Convert UNIX epoch to timestamptz.
*/
Datum
float8_timestamptz(PG_FUNCTION_ARGS)
{
float8 seconds = PG_GETARG_FLOAT8(0);
TimestampTz result;
/* Deal with NaN and infinite inputs ... */
if (isnan(seconds))
ereport(ERROR,
(errcode(ERRCODE_DATETIME_VALUE_OUT_OF_RANGE),
errmsg("timestamp cannot be NaN")));
if (isinf(seconds))
{
if (seconds < 0)
TIMESTAMP_NOBEGIN(result);
else
TIMESTAMP_NOEND(result);
}
else
{
/* Out of range? */
if (seconds <
Blind try to fix portability issue in commit 8f93bd851 et al. The S/390 members of the buildfarm are showing failures indicating that they're having trouble with the rint() calls I added yesterday. There's no good reason for that, and I wonder if it is a compiler bug similar to the one we worked around in d9476b838. Try to fix it using the same method as before, namely to store the result of rint() back into a "double" variable rather than immediately converting to int64. (This isn't entirely waving a dead chicken, since on machines with wider-than-double float registers, the extra store forces a width conversion. I don't know if S/390 is like that, but it seems worth trying.) In passing, merge duplicate ereport() calls in float8_timestamptz(). Per buildfarm.
2017-02-09 21:49:57 +01:00
(float8) SECS_PER_DAY * (DATETIME_MIN_JULIAN - UNIX_EPOCH_JDATE)
|| seconds >=
Allow to_timestamp(float8) to convert float infinity to timestamp infinity. With the original SQL-function implementation, such cases failed because we don't support infinite intervals. Converting the function to C lets us bypass the interval representation, which should be a bit faster as well as more flexible. Vitaly Burovoy, reviewed by Anastasia Lubennikova
2016-03-29 23:09:21 +02:00
(float8) SECS_PER_DAY * (TIMESTAMP_END_JULIAN - UNIX_EPOCH_JDATE))
ereport(ERROR,
(errcode(ERRCODE_DATETIME_VALUE_OUT_OF_RANGE),
errmsg("timestamp out of range: \"%g\"", seconds)));
/* Convert UNIX epoch to Postgres epoch */
seconds -= ((POSTGRES_EPOCH_JDATE - UNIX_EPOCH_JDATE) * SECS_PER_DAY);
Blind try to fix portability issue in commit 8f93bd851 et al. The S/390 members of the buildfarm are showing failures indicating that they're having trouble with the rint() calls I added yesterday. There's no good reason for that, and I wonder if it is a compiler bug similar to the one we worked around in d9476b838. Try to fix it using the same method as before, namely to store the result of rint() back into a "double" variable rather than immediately converting to int64. (This isn't entirely waving a dead chicken, since on machines with wider-than-double float registers, the extra store forces a width conversion. I don't know if S/390 is like that, but it seems worth trying.) In passing, merge duplicate ereport() calls in float8_timestamptz(). Per buildfarm.
2017-02-09 21:49:57 +01:00
seconds = rint(seconds * USECS_PER_SEC);
result = (int64) seconds;
Allow to_timestamp(float8) to convert float infinity to timestamp infinity. With the original SQL-function implementation, such cases failed because we don't support infinite intervals. Converting the function to C lets us bypass the interval representation, which should be a bit faster as well as more flexible. Vitaly Burovoy, reviewed by Anastasia Lubennikova
2016-03-29 23:09:21 +02:00
/* Recheck in case roundoff produces something just out of range */
if (!IS_VALID_TIMESTAMP(result))
ereport(ERROR,
(errcode(ERRCODE_DATETIME_VALUE_OUT_OF_RANGE),
errmsg("timestamp out of range: \"%g\"",
PG_GETARG_FLOAT8(0))));
}
PG_RETURN_TIMESTAMP(result);
}
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
/* timestamptz_out()
* Convert a timestamp to external form.
*/
Datum
timestamptz_out(PG_FUNCTION_ARGS)
{
Add binary I/O routines for a bunch more datatypes. Still a few to go, but that was enough tedium for one day. Along the way, move the few support routines for types xid and cid into a more logical place.
2003-05-13 01:08:52 +02:00
TimestampTz dt = PG_GETARG_TIMESTAMPTZ(0);
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
char *result;
Another PGINDENT run that changes variable indenting and case label indenting. Also static variable indenting.
1997-09-08 04:41:22 +02:00
int tz;
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
struct pg_tm tt,
Another PGINDENT run that changes variable indenting and case label indenting. Also static variable indenting.
1997-09-08 04:41:22 +02:00
*tm = &tt;
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;
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;
Another PGINDENT run that changes variable indenting and case label indenting. Also static variable indenting.
1997-09-08 04:41:22 +02:00
char buf[MAXDATELEN + 1];
Missed another tar file... :(
1997-03-14 06:58:13 +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
if (TIMESTAMP_NOT_FINITE(dt))
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
EncodeSpecialTimestamp(dt, buf);
This patch makes it possible to use the full set of timezones when doing "AT TIME ZONE", and not just the shorlist previously available. For example: SELECT CURRENT_TIMESTAMP AT TIME ZONE 'Europe/London'; works fine now. It will also obey whatever DST rules were in effect at just that date, which the previous implementation did not. It also supports the AT TIME ZONE on the timetz datatype. The whole handling of DST is a bit bogus there, so I chose to make it use whatever DST rules are in effect at the time of executig the query. not sure if anybody is actuallyi *using* timetz though, it seems pretty unpredictable just because of this... Magnus Hagander
2005-06-15 02:34:11 +02:00
else if (timestamp2tm(dt, &tz, tm, &fsec, &tzn, NULL) == 0)
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
EncodeDateTime(tm, fsec, true, tz, tzn, DateStyle, buf);
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
else
Error message editing in utils/adt. Again thanks to Joe Conway for doing the bulk of the heavy lifting ...
2003-07-27 06:53:12 +02:00
ereport(ERROR,
(errcode(ERRCODE_DATETIME_VALUE_OUT_OF_RANGE),
errmsg("timestamp out of range")));
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
result = pstrdup(buf);
PG_RETURN_CSTRING(result);
}
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
Add binary I/O routines for a bunch more datatypes. Still a few to go, but that was enough tedium for one day. Along the way, move the few support routines for types xid and cid into a more logical place.
2003-05-13 01:08:52 +02:00
/*
* timestamptz_recv - converts external binary format to timestamptz
*/
Datum
timestamptz_recv(PG_FUNCTION_ARGS)
{
StringInfo buf = (StringInfo) PG_GETARG_POINTER(0);
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
Change typreceive function API so that receive functions get the same optional arguments as text input functions, ie, typioparam OID and atttypmod. Make all the datatypes that use typmod enforce it the same way in typreceive as they do in typinput. This fixes a problem with failure to enforce length restrictions during COPY FROM BINARY.
2005-07-10 23:14:00 +02:00
#ifdef NOT_USED
Oid typelem = PG_GETARG_OID(1);
#endif
int32 typmod = PG_GETARG_INT32(2);
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
TimestampTz timestamp;
Add range-checking in timestamp_recv and timestamptz_recv, per Stephen Frost. Also tighten date range check in timestamp2tm.
2004-06-03 19:57:09 +02:00
int tz;
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
struct pg_tm tt,
Add range-checking in timestamp_recv and timestamptz_recv, per Stephen Frost. Also tighten date range check in timestamp2tm.
2004-06-03 19:57:09 +02:00
*tm = &tt;
fsec_t fsec;
Add binary I/O routines for a bunch more datatypes. Still a few to go, but that was enough tedium for one day. Along the way, move the few support routines for types xid and cid into a more logical place.
2003-05-13 01:08:52 +02:00
Add range-checking in timestamp_recv and timestamptz_recv, per Stephen Frost. Also tighten date range check in timestamp2tm.
2004-06-03 19:57:09 +02:00
timestamp = (TimestampTz) pq_getmsgint64(buf);
Various minor corrections of and improvements to comments. Aleksander Alekseev
2016-03-18 14:38:59 +01:00
/* range check: see if timestamptz_out would like it */
Add range-checking in timestamp_recv and timestamptz_recv, per Stephen Frost. Also tighten date range check in timestamp2tm.
2004-06-03 19:57:09 +02:00
if (TIMESTAMP_NOT_FINITE(timestamp))
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
/* ok */ ;
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
else if (timestamp2tm(timestamp, &tz, tm, &fsec, NULL, NULL) != 0 ||
!IS_VALID_TIMESTAMP(timestamp))
Add range-checking in timestamp_recv and timestamptz_recv, per Stephen Frost. Also tighten date range check in timestamp2tm.
2004-06-03 19:57:09 +02:00
ereport(ERROR,
(errcode(ERRCODE_DATETIME_VALUE_OUT_OF_RANGE),
errmsg("timestamp out of range")));
Change typreceive function API so that receive functions get the same optional arguments as text input functions, ie, typioparam OID and atttypmod. Make all the datatypes that use typmod enforce it the same way in typreceive as they do in typinput. This fixes a problem with failure to enforce length restrictions during COPY FROM BINARY.
2005-07-10 23:14:00 +02:00
AdjustTimestampForTypmod(&timestamp, typmod);
Add range-checking in timestamp_recv and timestamptz_recv, per Stephen Frost. Also tighten date range check in timestamp2tm.
2004-06-03 19:57:09 +02:00
PG_RETURN_TIMESTAMPTZ(timestamp);
Add binary I/O routines for a bunch more datatypes. Still a few to go, but that was enough tedium for one day. Along the way, move the few support routines for types xid and cid into a more logical place.
2003-05-13 01:08:52 +02:00
}
/*
* timestamptz_send - converts timestamptz to binary format
*/
Datum
timestamptz_send(PG_FUNCTION_ARGS)
{
pgindent run.
2003-08-04 02:43:34 +02:00
TimestampTz timestamp = PG_GETARG_TIMESTAMPTZ(0);
Add binary I/O routines for a bunch more datatypes. Still a few to go, but that was enough tedium for one day. Along the way, move the few support routines for types xid and cid into a more logical place.
2003-05-13 01:08:52 +02:00
StringInfoData buf;
pq_begintypsend(&buf);
pq_sendint64(&buf, timestamp);
PG_RETURN_BYTEA_P(pq_endtypsend(&buf));
}
Support type modifiers for user-defined types, and pull most knowledge about typmod representation for standard types out into type-specific typmod I/O functions. Teodor Sigaev, with some editorialization by Tom Lane.
2006-12-30 22:21:56 +01:00
Datum
timestamptztypmodin(PG_FUNCTION_ARGS)
{
pgindent run for 8.3.
2007-11-15 22:14:46 +01:00
ArrayType *ta = PG_GETARG_ARRAYTYPE_P(0);
Support type modifiers for user-defined types, and pull most knowledge about typmod representation for standard types out into type-specific typmod I/O functions. Teodor Sigaev, with some editorialization by Tom Lane.
2006-12-30 22:21:56 +01:00
PG_RETURN_INT32(anytimestamp_typmodin(true, ta));
}
Datum
timestamptztypmodout(PG_FUNCTION_ARGS)
{
pgindent run for 8.3.
2007-11-15 22:14:46 +01:00
int32 typmod = PG_GETARG_INT32(0);
Support type modifiers for user-defined types, and pull most knowledge about typmod representation for standard types out into type-specific typmod I/O functions. Teodor Sigaev, with some editorialization by Tom Lane.
2006-12-30 22:21:56 +01:00
PG_RETURN_CSTRING(anytimestamp_typmodout(true, typmod));
}
Add binary I/O routines for a bunch more datatypes. Still a few to go, but that was enough tedium for one day. Along the way, move the few support routines for types xid and cid into a more logical place.
2003-05-13 01:08:52 +02:00
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
/* timestamptz_scale()
* Adjust time type for specified scale factor.
* Used by PostgreSQL type system to stuff columns.
*/
Datum
timestamptz_scale(PG_FUNCTION_ARGS)
{
Add binary I/O routines for a bunch more datatypes. Still a few to go, but that was enough tedium for one day. Along the way, move the few support routines for types xid and cid into a more logical place.
2003-05-13 01:08:52 +02:00
TimestampTz timestamp = PG_GETARG_TIMESTAMPTZ(0);
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
int32 typmod = PG_GETARG_INT32(1);
pgindent run on all C files. Java run to follow. initdb/regression tests pass.
2001-10-25 07:50:21 +02:00
TimestampTz result;
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
result = timestamp;
AdjustTimestampForTypmod does not work (at least not portably) on -infinity and +infinity. Put TIMESTAMP_NOT_FINITE guard into the routine, instead of forgetting it at some call sites. Fixes regression test failures here.
2001-10-03 17:50:48 +02:00
AdjustTimestampForTypmod(&result, typmod);
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
PG_RETURN_TIMESTAMPTZ(result);
}
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
/* interval_in()
* Convert a string to internal form.
*
* External format(s):
* Uses the generic date/time parsing and decoding routines.
*/
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
Datum
interval_in(PG_FUNCTION_ARGS)
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
{
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
char *str = PG_GETARG_CSTRING(0);
pgindent run on all C files. Java run to follow. initdb/regression tests pass.
2001-10-25 07:50:21 +02:00
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
#ifdef NOT_USED
Oid typelem = PG_GETARG_OID(1);
#endif
int32 typmod = PG_GETARG_INT32(2);
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
Interval *result;
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;
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
struct pg_tm tt,
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
*tm = &tt;
int dtype;
int nf;
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
int 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
int dterr;
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
char *field[MAXDATEFIELDS];
int ftype[MAXDATEFIELDS];
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 workbuf[256];
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
tm->tm_year = 0;
tm->tm_mon = 0;
tm->tm_mday = 0;
tm->tm_hour = 0;
tm->tm_min = 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 (typmod >= 0)
range = INTERVAL_RANGE(typmod);
else
range = INTERVAL_FULL_RANGE;
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
dterr = ParseDateTime(str, workbuf, sizeof(workbuf), field,
ftype, MAXDATEFIELDS, &nf);
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 == 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
dterr = DecodeInterval(field, ftype, nf, range,
&dtype, tm, &fsec);
/* if those functions think it's a bad format, try ISO8601 style */
if (dterr == DTERR_BAD_FORMAT)
8.4 pgindent run, with new combined Linux/FreeBSD/MinGW typedef list provided by Andrew.
2009-06-11 16:49:15 +02:00
dterr = DecodeISO8601Interval(str,
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);
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 != 0)
{
if (dterr == DTERR_FIELD_OVERFLOW)
dterr = DTERR_INTERVAL_OVERFLOW;
DateTimeParseError(dterr, str, "interval");
}
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +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
result = (Interval *) palloc(sizeof(Interval));
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
switch (dtype)
{
case DTK_DELTA:
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 (tm2interval(tm, fsec, result) != 0)
Error message editing in utils/adt. Again thanks to Joe Conway for doing the bulk of the heavy lifting ...
2003-07-27 06:53:12 +02:00
ereport(ERROR,
(errcode(ERRCODE_DATETIME_VALUE_OUT_OF_RANGE),
errmsg("interval out of range")));
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;
Support special values 'now', 'current', etc on output.
1997-10-09 07:02:17 +02:00
default:
Error message editing in utils/adt. Again thanks to Joe Conway for doing the bulk of the heavy lifting ...
2003-07-27 06:53:12 +02:00
elog(ERROR, "unexpected dtype %d while parsing interval \"%s\"",
dtype, str);
Support special values 'now', 'current', etc on output.
1997-10-09 07:02:17 +02:00
}
pgindent run before 6.3 release, with Thomas' requested changes.
1998-02-26 05:46:47 +01:00
Change typreceive function API so that receive functions get the same optional arguments as text input functions, ie, typioparam OID and atttypmod. Make all the datatypes that use typmod enforce it the same way in typreceive as they do in typinput. This fixes a problem with failure to enforce length restrictions during COPY FROM BINARY.
2005-07-10 23:14:00 +02:00
AdjustIntervalForTypmod(result, typmod);
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
PG_RETURN_INTERVAL_P(result);
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
}
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
/* interval_out()
* Convert a time span to external form.
*/
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
Datum
interval_out(PG_FUNCTION_ARGS)
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
{
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
Interval *span = PG_GETARG_INTERVAL_P(0);
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
char *result;
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
struct pg_tm tt,
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
*tm = &tt;
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;
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
char buf[MAXDATELEN + 1];
if (interval2tm(*span, tm, &fsec) != 0)
Error message editing in utils/adt. Again thanks to Joe Conway for doing the bulk of the heavy lifting ...
2003-07-27 06:53:12 +02:00
elog(ERROR, "could not convert interval to tm");
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +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
EncodeInterval(tm, fsec, IntervalStyle, buf);
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
result = pstrdup(buf);
PG_RETURN_CSTRING(result);
}
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
Add binary I/O routines for a bunch more datatypes. Still a few to go, but that was enough tedium for one day. Along the way, move the few support routines for types xid and cid into a more logical place.
2003-05-13 01:08:52 +02:00
/*
* interval_recv - converts external binary format to interval
*/
Datum
interval_recv(PG_FUNCTION_ARGS)
{
StringInfo buf = (StringInfo) PG_GETARG_POINTER(0);
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
Change typreceive function API so that receive functions get the same optional arguments as text input functions, ie, typioparam OID and atttypmod. Make all the datatypes that use typmod enforce it the same way in typreceive as they do in typinput. This fixes a problem with failure to enforce length restrictions during COPY FROM BINARY.
2005-07-10 23:14:00 +02:00
#ifdef NOT_USED
Oid typelem = PG_GETARG_OID(1);
#endif
int32 typmod = PG_GETARG_INT32(2);
Add binary I/O routines for a bunch more datatypes. Still a few to go, but that was enough tedium for one day. Along the way, move the few support routines for types xid and cid into a more logical place.
2003-05-13 01:08:52 +02:00
Interval *interval;
interval = (Interval *) palloc(sizeof(Interval));
Remove unnecessary parentheses in datetime/timestamp code.
2005-05-23 19:13:14 +02:00
interval->time = pq_getmsgint64(buf);
Add 'day' field to INTERVAL so 1 day interval can be distinguished from 24 hours. This is very helpful for daylight savings time: select '2005-05-03 00:00:00 EST'::timestamp with time zone + '24 hours'; ?column? ---------------------- 2005-05-04 01:00:00-04 select '2005-05-03 00:00:00 EST'::timestamp with time zone + '1 day'; ?column? ---------------------- 2005-05-04 01:00:00-04 Michael Glaesemann
2005-07-20 18:42:32 +02:00
interval->day = pq_getmsgint(buf, sizeof(interval->day));
Remove unnecessary parentheses in datetime/timestamp code.
2005-05-23 19:13:14 +02:00
interval->month = pq_getmsgint(buf, sizeof(interval->month));
Add binary I/O routines for a bunch more datatypes. Still a few to go, but that was enough tedium for one day. Along the way, move the few support routines for types xid and cid into a more logical place.
2003-05-13 01:08:52 +02:00
Change typreceive function API so that receive functions get the same optional arguments as text input functions, ie, typioparam OID and atttypmod. Make all the datatypes that use typmod enforce it the same way in typreceive as they do in typinput. This fixes a problem with failure to enforce length restrictions during COPY FROM BINARY.
2005-07-10 23:14:00 +02:00
AdjustIntervalForTypmod(interval, typmod);
Add binary I/O routines for a bunch more datatypes. Still a few to go, but that was enough tedium for one day. Along the way, move the few support routines for types xid and cid into a more logical place.
2003-05-13 01:08:52 +02:00
PG_RETURN_INTERVAL_P(interval);
}
/*
* interval_send - converts interval to binary format
*/
Datum
interval_send(PG_FUNCTION_ARGS)
{
Interval *interval = PG_GETARG_INTERVAL_P(0);
StringInfoData buf;
pq_begintypsend(&buf);
pq_sendint64(&buf, interval->time);
Replace remaining uses of pq_sendint with pq_sendint{8,16,32}. pq_sendint() remains, so extension code doesn't unnecessarily break. Author: Andres Freund Discussion: https://postgr.es/m/20170914063418.sckdzgjfrsbekae4@alap3.anarazel.de
2017-10-12 06:00:46 +02:00
pq_sendint32(&buf, interval->day);
pq_sendint32(&buf, interval->month);
Add binary I/O routines for a bunch more datatypes. Still a few to go, but that was enough tedium for one day. Along the way, move the few support routines for types xid and cid into a more logical place.
2003-05-13 01:08:52 +02:00
PG_RETURN_BYTEA_P(pq_endtypsend(&buf));
}
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
/*
* The interval typmod stores a "range" in its high 16 bits and a "precision"
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
* in its low 16 bits. Both contribute to defining the resolution of the
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
* type. Range addresses resolution granules larger than one second, and
* precision specifies resolution below one second. This representation can
* express all SQL standard resolutions, but we implement them all in terms of
* truncating rightward from some position. Range is a bitmap of permitted
* fields, but only the temporally-smallest such field is significant to our
* calculations. Precision is a count of sub-second decimal places to retain.
* Setting all bits (INTERVAL_FULL_PRECISION) gives the same truncation
* semantics as choosing MAX_INTERVAL_PRECISION.
*/
Support type modifiers for user-defined types, and pull most knowledge about typmod representation for standard types out into type-specific typmod I/O functions. Teodor Sigaev, with some editorialization by Tom Lane.
2006-12-30 22:21:56 +01:00
Datum
intervaltypmodin(PG_FUNCTION_ARGS)
{
pgindent run for 8.3.
2007-11-15 22:14:46 +01:00
ArrayType *ta = PG_GETARG_ARRAYTYPE_P(0);
int32 *tl;
int n;
Support type modifiers for user-defined types, and pull most knowledge about typmod representation for standard types out into type-specific typmod I/O functions. Teodor Sigaev, with some editorialization by Tom Lane.
2006-12-30 22:21:56 +01:00
int32 typmod;
Tweak the API for per-datatype typmodin functions so that they are passed an array of strings rather than an array of integers, and allow any simple constant or identifier to be used in typmods; for example create table foo (f1 widget(42,'23skidoo',point)); Of course the typmodin function has still got to pack this info into a non-negative int32 for storage, but it's still a useful improvement in flexibility, especially considering that you can do nearly anything if you are willing to keep the info in a side table. We can get away with this change since we have not yet released a version providing user-definable typmods. Per discussion.
2007-06-15 22:56:52 +02:00
tl = ArrayGetIntegerTypmods(ta, &n);
Support type modifiers for user-defined types, and pull most knowledge about typmod representation for standard types out into type-specific typmod I/O functions. Teodor Sigaev, with some editorialization by Tom Lane.
2006-12-30 22:21:56 +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
* tl[0] - interval range (fields bitmask) tl[1] - precision (optional)
Support type modifiers for user-defined types, and pull most knowledge about typmod representation for standard types out into type-specific typmod I/O functions. Teodor Sigaev, with some editorialization by Tom Lane.
2006-12-30 22:21:56 +01:00
*
* Note we must validate tl[0] even though it's normally guaranteed
* correct by the grammar --- consider SELECT 'foo'::"interval"(1000).
*/
if (n > 0)
{
switch (tl[0])
{
case INTERVAL_MASK(YEAR):
case INTERVAL_MASK(MONTH):
case INTERVAL_MASK(DAY):
case INTERVAL_MASK(HOUR):
case INTERVAL_MASK(MINUTE):
case INTERVAL_MASK(SECOND):
case INTERVAL_MASK(YEAR) | INTERVAL_MASK(MONTH):
case INTERVAL_MASK(DAY) | INTERVAL_MASK(HOUR):
case INTERVAL_MASK(DAY) | INTERVAL_MASK(HOUR) | INTERVAL_MASK(MINUTE):
case INTERVAL_MASK(DAY) | INTERVAL_MASK(HOUR) | INTERVAL_MASK(MINUTE) | INTERVAL_MASK(SECOND):
case INTERVAL_MASK(HOUR) | INTERVAL_MASK(MINUTE):
case INTERVAL_MASK(HOUR) | INTERVAL_MASK(MINUTE) | INTERVAL_MASK(SECOND):
case INTERVAL_MASK(MINUTE) | INTERVAL_MASK(SECOND):
case INTERVAL_FULL_RANGE:
/* all OK */
break;
default:
ereport(ERROR,
(errcode(ERRCODE_INVALID_PARAMETER_VALUE),
errmsg("invalid INTERVAL type modifier")));
}
}
if (n == 1)
{
if (tl[0] != INTERVAL_FULL_RANGE)
typmod = INTERVAL_TYPMOD(INTERVAL_FULL_PRECISION, tl[0]);
else
typmod = -1;
}
else if (n == 2)
{
if (tl[1] < 0)
ereport(ERROR,
(errcode(ERRCODE_INVALID_PARAMETER_VALUE),
errmsg("INTERVAL(%d) precision must not be negative",
pgindent run for 8.3.
2007-11-15 22:14:46 +01:00
tl[1])));
Support type modifiers for user-defined types, and pull most knowledge about typmod representation for standard types out into type-specific typmod I/O functions. Teodor Sigaev, with some editorialization by Tom Lane.
2006-12-30 22:21:56 +01:00
if (tl[1] > MAX_INTERVAL_PRECISION)
{
ereport(WARNING,
(errcode(ERRCODE_INVALID_PARAMETER_VALUE),
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
errmsg("INTERVAL(%d) precision reduced to maximum allowed, %d",
tl[1], MAX_INTERVAL_PRECISION)));
Support type modifiers for user-defined types, and pull most knowledge about typmod representation for standard types out into type-specific typmod I/O functions. Teodor Sigaev, with some editorialization by Tom Lane.
2006-12-30 22:21:56 +01:00
typmod = INTERVAL_TYPMOD(MAX_INTERVAL_PRECISION, tl[0]);
}
else
typmod = INTERVAL_TYPMOD(tl[1], tl[0]);
}
else
{
ereport(ERROR,
(errcode(ERRCODE_INVALID_PARAMETER_VALUE),
pgindent run for 8.3.
2007-11-15 22:14:46 +01:00
errmsg("invalid INTERVAL type modifier")));
Support type modifiers for user-defined types, and pull most knowledge about typmod representation for standard types out into type-specific typmod I/O functions. Teodor Sigaev, with some editorialization by Tom Lane.
2006-12-30 22:21:56 +01:00
typmod = 0; /* keep compiler quiet */
}
PG_RETURN_INT32(typmod);
}
Datum
intervaltypmodout(PG_FUNCTION_ARGS)
{
pgindent run for 8.3.
2007-11-15 22:14:46 +01:00
int32 typmod = PG_GETARG_INT32(0);
Support type modifiers for user-defined types, and pull most knowledge about typmod representation for standard types out into type-specific typmod I/O functions. Teodor Sigaev, with some editorialization by Tom Lane.
2006-12-30 22:21:56 +01:00
char *res = (char *) palloc(64);
pgindent run for 8.3.
2007-11-15 22:14:46 +01:00
int fields;
int precision;
Support type modifiers for user-defined types, and pull most knowledge about typmod representation for standard types out into type-specific typmod I/O functions. Teodor Sigaev, with some editorialization by Tom Lane.
2006-12-30 22:21:56 +01:00
const char *fieldstr;
if (typmod < 0)
{
*res = '\0';
PG_RETURN_CSTRING(res);
}
fields = INTERVAL_RANGE(typmod);
precision = INTERVAL_PRECISION(typmod);
switch (fields)
{
case INTERVAL_MASK(YEAR):
fieldstr = " year";
break;
case INTERVAL_MASK(MONTH):
fieldstr = " month";
break;
case INTERVAL_MASK(DAY):
fieldstr = " day";
break;
case INTERVAL_MASK(HOUR):
fieldstr = " hour";
break;
case INTERVAL_MASK(MINUTE):
fieldstr = " minute";
break;
case INTERVAL_MASK(SECOND):
fieldstr = " second";
break;
case INTERVAL_MASK(YEAR) | INTERVAL_MASK(MONTH):
fieldstr = " year to month";
break;
case INTERVAL_MASK(DAY) | INTERVAL_MASK(HOUR):
fieldstr = " day to hour";
break;
case INTERVAL_MASK(DAY) | INTERVAL_MASK(HOUR) | INTERVAL_MASK(MINUTE):
fieldstr = " day to minute";
break;
case INTERVAL_MASK(DAY) | INTERVAL_MASK(HOUR) | INTERVAL_MASK(MINUTE) | INTERVAL_MASK(SECOND):
fieldstr = " day to second";
break;
case INTERVAL_MASK(HOUR) | INTERVAL_MASK(MINUTE):
fieldstr = " hour to minute";
break;
case INTERVAL_MASK(HOUR) | INTERVAL_MASK(MINUTE) | INTERVAL_MASK(SECOND):
fieldstr = " hour to second";
break;
case INTERVAL_MASK(MINUTE) | INTERVAL_MASK(SECOND):
fieldstr = " minute to second";
break;
case INTERVAL_FULL_RANGE:
fieldstr = "";
break;
default:
elog(ERROR, "invalid INTERVAL typmod: 0x%x", typmod);
fieldstr = "";
break;
}
if (precision != INTERVAL_FULL_PRECISION)
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
snprintf(res, 64, "%s(%d)", fieldstr, precision);
Support type modifiers for user-defined types, and pull most knowledge about typmod representation for standard types out into type-specific typmod I/O functions. Teodor Sigaev, with some editorialization by Tom Lane.
2006-12-30 22:21:56 +01:00
else
snprintf(res, 64, "%s", fieldstr);
PG_RETURN_CSTRING(res);
}
Fix interval_transform so it doesn't throw away non-no-op casts. interval_transform() contained two separate bugs that caused it to sometimes mistakenly decide that a cast from interval to restricted interval is a no-op and throw it away. First, it was wrong to rely on dt.h's field type macros to have an ordering consistent with the field's significance; in one case they do not. This led to mistakenly treating YEAR as less significant than MONTH, so that a cast from INTERVAL MONTH to INTERVAL YEAR was incorrectly discarded. Second, fls(1<<k) produces k+1 not k, so comparing its output directly to SECOND was wrong. This led to supposing that a cast to INTERVAL MINUTE was really a cast to INTERVAL SECOND and so could be discarded. To fix, get rid of the use of fls(), and make a function based on intervaltypmodout to produce a field ID code adapted to the need here. Per bug #14479 from Piotr Stefaniak. Back-patch to 9.2 where transform functions were introduced, because this code was born broken. Discussion: https://postgr.es/m/20161227172307.10135.7747@wrigleys.postgresql.org
2016-12-27 21:43:54 +01:00
/*
* Given an interval typmod value, return a code for the least-significant
* field that the typmod allows to be nonzero, for instance given
* INTERVAL DAY TO HOUR we want to identify "hour".
*
* The results should be ordered by field significance, which means
* we can't use the dt.h macros YEAR etc, because for some odd reason
* they aren't ordered that way. Instead, arbitrarily represent
* SECOND = 0, MINUTE = 1, HOUR = 2, DAY = 3, MONTH = 4, YEAR = 5.
*/
static int
intervaltypmodleastfield(int32 typmod)
{
if (typmod < 0)
return 0; /* SECOND */
switch (INTERVAL_RANGE(typmod))
{
case INTERVAL_MASK(YEAR):
return 5; /* YEAR */
case INTERVAL_MASK(MONTH):
return 4; /* MONTH */
case INTERVAL_MASK(DAY):
return 3; /* DAY */
case INTERVAL_MASK(HOUR):
return 2; /* HOUR */
case INTERVAL_MASK(MINUTE):
return 1; /* MINUTE */
case INTERVAL_MASK(SECOND):
return 0; /* SECOND */
case INTERVAL_MASK(YEAR) | INTERVAL_MASK(MONTH):
return 4; /* MONTH */
case INTERVAL_MASK(DAY) | INTERVAL_MASK(HOUR):
return 2; /* HOUR */
case INTERVAL_MASK(DAY) | INTERVAL_MASK(HOUR) | INTERVAL_MASK(MINUTE):
return 1; /* MINUTE */
case INTERVAL_MASK(DAY) | INTERVAL_MASK(HOUR) | INTERVAL_MASK(MINUTE) | INTERVAL_MASK(SECOND):
return 0; /* SECOND */
case INTERVAL_MASK(HOUR) | INTERVAL_MASK(MINUTE):
return 1; /* MINUTE */
case INTERVAL_MASK(HOUR) | INTERVAL_MASK(MINUTE) | INTERVAL_MASK(SECOND):
return 0; /* SECOND */
case INTERVAL_MASK(MINUTE) | INTERVAL_MASK(SECOND):
return 0; /* SECOND */
case INTERVAL_FULL_RANGE:
return 0; /* SECOND */
default:
elog(ERROR, "invalid INTERVAL typmod: 0x%x", typmod);
break;
}
return 0; /* can't get here, but keep compiler quiet */
}
Add binary I/O routines for a bunch more datatypes. Still a few to go, but that was enough tedium for one day. Along the way, move the few support routines for types xid and cid into a more logical place.
2003-05-13 01:08:52 +02:00
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
/*
* interval_support()
*
* Planner support function for interval_scale().
*
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
* Flatten superfluous calls to interval_scale(). The interval typmod is
* complex to permit accepting and regurgitating all SQL standard variations.
* For truncation purposes, it boils down to a single, simple granularity.
*/
Datum
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
interval_support(PG_FUNCTION_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
{
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
Node *rawreq = (Node *) PG_GETARG_POINTER(0);
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;
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
if (IsA(rawreq, SupportRequestSimplify))
{
SupportRequestSimplify *req = (SupportRequestSimplify *) rawreq;
FuncExpr *expr = req->fcall;
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
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
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
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
typmod = (Node *) lsecond(expr->args);
Fix interval_transform so it doesn't throw away non-no-op casts. interval_transform() contained two separate bugs that caused it to sometimes mistakenly decide that a cast from interval to restricted interval is a no-op and throw it away. First, it was wrong to rely on dt.h's field type macros to have an ordering consistent with the field's significance; in one case they do not. This led to mistakenly treating YEAR as less significant than MONTH, so that a cast from INTERVAL MONTH to INTERVAL YEAR was incorrectly discarded. Second, fls(1<<k) produces k+1 not k, so comparing its output directly to SECOND was wrong. This led to supposing that a cast to INTERVAL MINUTE was really a cast to INTERVAL SECOND and so could be discarded. To fix, get rid of the use of fls(), and make a function based on intervaltypmodout to produce a field ID code adapted to the need here. Per bug #14479 from Piotr Stefaniak. Back-patch to 9.2 where transform functions were introduced, because this code was born broken. Discussion: https://postgr.es/m/20161227172307.10135.7747@wrigleys.postgresql.org
2016-12-27 21:43:54 +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
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
{
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
Node *source = (Node *) linitial(expr->args);
int32 new_typmod = DatumGetInt32(((Const *) typmod)->constvalue);
bool noop;
if (new_typmod < 0)
noop = true;
Fix interval_transform so it doesn't throw away non-no-op casts. interval_transform() contained two separate bugs that caused it to sometimes mistakenly decide that a cast from interval to restricted interval is a no-op and throw it away. First, it was wrong to rely on dt.h's field type macros to have an ordering consistent with the field's significance; in one case they do not. This led to mistakenly treating YEAR as less significant than MONTH, so that a cast from INTERVAL MONTH to INTERVAL YEAR was incorrectly discarded. Second, fls(1<<k) produces k+1 not k, so comparing its output directly to SECOND was wrong. This led to supposing that a cast to INTERVAL MINUTE was really a cast to INTERVAL SECOND and so could be discarded. To fix, get rid of the use of fls(), and make a function based on intervaltypmodout to produce a field ID code adapted to the need here. Per bug #14479 from Piotr Stefaniak. Back-patch to 9.2 where transform functions were introduced, because this code was born broken. Discussion: https://postgr.es/m/20161227172307.10135.7747@wrigleys.postgresql.org
2016-12-27 21:43:54 +01:00
else
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
{
int32 old_typmod = exprTypmod(source);
int old_least_field;
int new_least_field;
int old_precis;
int new_precis;
old_least_field = intervaltypmodleastfield(old_typmod);
new_least_field = intervaltypmodleastfield(new_typmod);
if (old_typmod < 0)
old_precis = INTERVAL_FULL_PRECISION;
else
old_precis = INTERVAL_PRECISION(old_typmod);
new_precis = INTERVAL_PRECISION(new_typmod);
/*
* Cast is a no-op if least field stays the same or decreases
* while precision stays the same or increases. But
* precision, which is to say, sub-second precision, only
* affects ranges that include SECOND.
*/
noop = (new_least_field <= old_least_field) &&
(old_least_field > 0 /* SECOND */ ||
new_precis >= MAX_INTERVAL_PRECISION ||
new_precis >= old_precis);
}
if (noop)
ret = relabel_to_typmod(source, new_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
}
}
PG_RETURN_POINTER(ret);
}
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
/* interval_scale()
* Adjust interval type for specified fields.
* Used by PostgreSQL type system to stuff columns.
*/
Datum
interval_scale(PG_FUNCTION_ARGS)
{
Interval *interval = PG_GETARG_INTERVAL_P(0);
int32 typmod = PG_GETARG_INT32(1);
Interval *result;
result = palloc(sizeof(Interval));
*result = *interval;
AdjustIntervalForTypmod(result, typmod);
PG_RETURN_INTERVAL_P(result);
}
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
Improve comments for AdjustIntervalForTypmod. Blank line adjustments.
2005-07-12 18:05:12 +02:00
/*
* Adjust interval for specified precision, in both YEAR to SECOND
* range and sub-second precision.
*/
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
static void
AdjustIntervalForTypmod(Interval *interval, int32 typmod)
{
pgindent run.
2002-09-04 22:31:48 +02:00
static const int64 IntervalScales[MAX_INTERVAL_PRECISION + 1] = {
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
INT64CONST(1000000),
INT64CONST(100000),
INT64CONST(10000),
INT64CONST(1000),
INT64CONST(100),
INT64CONST(10),
INT64CONST(1)
};
pgindent run.
2002-09-04 22:31:48 +02:00
static const int64 IntervalOffsets[MAX_INTERVAL_PRECISION + 1] = {
Repair inconsistent rounding behavior for timestamp, time, interval, per gripe from Csaba Nagy. There is still potential for platform-specific behavior for values that are exactly halfway between integers, but at least we now get the expected answer for all other cases.
2003-01-09 02:06:57 +01:00
INT64CONST(500000),
INT64CONST(50000),
INT64CONST(5000),
INT64CONST(500),
INT64CONST(50),
INT64CONST(5),
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
INT64CONST(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
* Unspecified range and precision? Then not necessary to adjust. Setting
Change AdjustIntervalForTypmod to not discard higher-order field values on the grounds that they don't fit into the specified interval qualifier (typmod). This behavior, while of long standing, is clearly wrong per spec --- for example the value INTERVAL '999' SECOND means 999 seconds and should not be reduced to less than 60 seconds. In some cases there could be grounds to raise an error if higher-order field values are not given as zero; for example '1 year 1 month'::INTERVAL MONTH should arguably be taken as an error rather than equivalent to 13 months. However our internal representation doesn't allow us to do that in a fashion that would consistently reject all and only the cases that a strict reading of the spec would suggest. Also, seeing that for example INTERVAL '13' MONTH will print out as '1 year 1 mon', we have to be careful not to create a situation where valid data will fail to dump and reload. The present patch therefore takes the attitude of not throwing an error in any such case. We might want to revisit that in future but it would take more redesign than seems prudent in late beta. Per a complaint from Sebastien Flaesch and subsequent discussion. While at other times we might have just postponed such an issue to the next development cycle, 8.4 already has changed the parsing of interval literals quite a bit in an effort to accept all spec-compliant cases correctly. This seems like a change that should be part of that rather than coming along later.
2009-06-02 01:55:15 +02:00
* typmod to -1 is the convention for all data types.
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
*/
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 (typmod >= 0)
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
{
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
int range = INTERVAL_RANGE(typmod);
int precision = INTERVAL_PRECISION(typmod);
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
Change AdjustIntervalForTypmod to not discard higher-order field values on the grounds that they don't fit into the specified interval qualifier (typmod). This behavior, while of long standing, is clearly wrong per spec --- for example the value INTERVAL '999' SECOND means 999 seconds and should not be reduced to less than 60 seconds. In some cases there could be grounds to raise an error if higher-order field values are not given as zero; for example '1 year 1 month'::INTERVAL MONTH should arguably be taken as an error rather than equivalent to 13 months. However our internal representation doesn't allow us to do that in a fashion that would consistently reject all and only the cases that a strict reading of the spec would suggest. Also, seeing that for example INTERVAL '13' MONTH will print out as '1 year 1 mon', we have to be careful not to create a situation where valid data will fail to dump and reload. The present patch therefore takes the attitude of not throwing an error in any such case. We might want to revisit that in future but it would take more redesign than seems prudent in late beta. Per a complaint from Sebastien Flaesch and subsequent discussion. While at other times we might have just postponed such an issue to the next development cycle, 8.4 already has changed the parsing of interval literals quite a bit in an effort to accept all spec-compliant cases correctly. This seems like a change that should be part of that rather than coming along later.
2009-06-02 01:55:15 +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
* Our interpretation of intervals with a limited set of fields is
* that fields to the right of the last one specified are zeroed out,
* but those to the left of it remain valid. Thus for example there
* is no operational difference between INTERVAL YEAR TO MONTH and
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
* INTERVAL MONTH. In some cases we could meaningfully enforce that
8.4 pgindent run, with new combined Linux/FreeBSD/MinGW typedef list provided by Andrew.
2009-06-11 16:49:15 +02:00
* higher-order fields are zero; for example INTERVAL DAY could reject
* nonzero "month" field. However that seems a bit pointless when we
* can't do it consistently. (We cannot enforce a range limit on the
* highest expected field, since we do not have any equivalent of
Add a comment to AdjustIntervalForTypmod to reduce chance of future bugs. It's not entirely evident how the logic here relates to the interval_transform function, so let's clue people in that they need to check that if the rules change.
2012-02-09 18:22:21 +01:00
* SQL's <interval leading field precision>.) If we ever decide to
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
* revisit this, interval_support will likely require adjusting.
Change AdjustIntervalForTypmod to not discard higher-order field values on the grounds that they don't fit into the specified interval qualifier (typmod). This behavior, while of long standing, is clearly wrong per spec --- for example the value INTERVAL '999' SECOND means 999 seconds and should not be reduced to less than 60 seconds. In some cases there could be grounds to raise an error if higher-order field values are not given as zero; for example '1 year 1 month'::INTERVAL MONTH should arguably be taken as an error rather than equivalent to 13 months. However our internal representation doesn't allow us to do that in a fashion that would consistently reject all and only the cases that a strict reading of the spec would suggest. Also, seeing that for example INTERVAL '13' MONTH will print out as '1 year 1 mon', we have to be careful not to create a situation where valid data will fail to dump and reload. The present patch therefore takes the attitude of not throwing an error in any such case. We might want to revisit that in future but it would take more redesign than seems prudent in late beta. Per a complaint from Sebastien Flaesch and subsequent discussion. While at other times we might have just postponed such an issue to the next development cycle, 8.4 already has changed the parsing of interval literals quite a bit in an effort to accept all spec-compliant cases correctly. This seems like a change that should be part of that rather than coming along later.
2009-06-02 01:55:15 +02:00
*
* Note: before PG 8.4 we interpreted a limited set of fields as
* actually causing a "modulo" operation on a given value, potentially
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
* losing high-order as well as low-order information. But there is
Change AdjustIntervalForTypmod to not discard higher-order field values on the grounds that they don't fit into the specified interval qualifier (typmod). This behavior, while of long standing, is clearly wrong per spec --- for example the value INTERVAL '999' SECOND means 999 seconds and should not be reduced to less than 60 seconds. In some cases there could be grounds to raise an error if higher-order field values are not given as zero; for example '1 year 1 month'::INTERVAL MONTH should arguably be taken as an error rather than equivalent to 13 months. However our internal representation doesn't allow us to do that in a fashion that would consistently reject all and only the cases that a strict reading of the spec would suggest. Also, seeing that for example INTERVAL '13' MONTH will print out as '1 year 1 mon', we have to be careful not to create a situation where valid data will fail to dump and reload. The present patch therefore takes the attitude of not throwing an error in any such case. We might want to revisit that in future but it would take more redesign than seems prudent in late beta. Per a complaint from Sebastien Flaesch and subsequent discussion. While at other times we might have just postponed such an issue to the next development cycle, 8.4 already has changed the parsing of interval literals quite a bit in an effort to accept all spec-compliant cases correctly. This seems like a change that should be part of that rather than coming along later.
2009-06-02 01:55:15 +02:00
* no support for such behavior in the standard, and it seems fairly
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
* undesirable on data consistency grounds anyway. Now we only
Change AdjustIntervalForTypmod to not discard higher-order field values on the grounds that they don't fit into the specified interval qualifier (typmod). This behavior, while of long standing, is clearly wrong per spec --- for example the value INTERVAL '999' SECOND means 999 seconds and should not be reduced to less than 60 seconds. In some cases there could be grounds to raise an error if higher-order field values are not given as zero; for example '1 year 1 month'::INTERVAL MONTH should arguably be taken as an error rather than equivalent to 13 months. However our internal representation doesn't allow us to do that in a fashion that would consistently reject all and only the cases that a strict reading of the spec would suggest. Also, seeing that for example INTERVAL '13' MONTH will print out as '1 year 1 mon', we have to be careful not to create a situation where valid data will fail to dump and reload. The present patch therefore takes the attitude of not throwing an error in any such case. We might want to revisit that in future but it would take more redesign than seems prudent in late beta. Per a complaint from Sebastien Flaesch and subsequent discussion. While at other times we might have just postponed such an issue to the next development cycle, 8.4 already has changed the parsing of interval literals quite a bit in an effort to accept all spec-compliant cases correctly. This seems like a change that should be part of that rather than coming along later.
2009-06-02 01:55:15 +02:00
* perform truncation or rounding of low-order fields.
*/
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
if (range == INTERVAL_FULL_RANGE)
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
{
/* Do nothing... */
}
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
else if (range == INTERVAL_MASK(YEAR))
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
{
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
interval->month = (interval->month / MONTHS_PER_YEAR) * MONTHS_PER_YEAR;
Add 'day' field to INTERVAL so 1 day interval can be distinguished from 24 hours. This is very helpful for daylight savings time: select '2005-05-03 00:00:00 EST'::timestamp with time zone + '24 hours'; ?column? ---------------------- 2005-05-04 01:00:00-04 select '2005-05-03 00:00:00 EST'::timestamp with time zone + '1 day'; ?column? ---------------------- 2005-05-04 01:00:00-04 Michael Glaesemann
2005-07-20 18:42:32 +02:00
interval->day = 0;
Remove unnecessary parentheses in datetime/timestamp code.
2005-05-23 19:13:14 +02:00
interval->time = 0;
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
}
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
else if (range == INTERVAL_MASK(MONTH))
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
{
Add 'day' field to INTERVAL so 1 day interval can be distinguished from 24 hours. This is very helpful for daylight savings time: select '2005-05-03 00:00:00 EST'::timestamp with time zone + '24 hours'; ?column? ---------------------- 2005-05-04 01:00:00-04 select '2005-05-03 00:00:00 EST'::timestamp with time zone + '1 day'; ?column? ---------------------- 2005-05-04 01:00:00-04 Michael Glaesemann
2005-07-20 18:42:32 +02:00
interval->day = 0;
Remove unnecessary parentheses in datetime/timestamp code.
2005-05-23 19:13:14 +02:00
interval->time = 0;
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
}
/* YEAR TO MONTH */
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
else if (range == (INTERVAL_MASK(YEAR) | INTERVAL_MASK(MONTH)))
Add 'day' field to INTERVAL so 1 day interval can be distinguished from 24 hours. This is very helpful for daylight savings time: select '2005-05-03 00:00:00 EST'::timestamp with time zone + '24 hours'; ?column? ---------------------- 2005-05-04 01:00:00-04 select '2005-05-03 00:00:00 EST'::timestamp with time zone + '1 day'; ?column? ---------------------- 2005-05-04 01:00:00-04 Michael Glaesemann
2005-07-20 18:42:32 +02:00
{
interval->day = 0;
Remove unnecessary parentheses in datetime/timestamp code.
2005-05-23 19:13:14 +02:00
interval->time = 0;
Add 'day' field to INTERVAL so 1 day interval can be distinguished from 24 hours. This is very helpful for daylight savings time: select '2005-05-03 00:00:00 EST'::timestamp with time zone + '24 hours'; ?column? ---------------------- 2005-05-04 01:00:00-04 select '2005-05-03 00:00:00 EST'::timestamp with time zone + '1 day'; ?column? ---------------------- 2005-05-04 01:00:00-04 Michael Glaesemann
2005-07-20 18:42:32 +02:00
}
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
else if (range == INTERVAL_MASK(DAY))
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
{
Add 'day' field to INTERVAL so 1 day interval can be distinguished from 24 hours. This is very helpful for daylight savings time: select '2005-05-03 00:00:00 EST'::timestamp with time zone + '24 hours'; ?column? ---------------------- 2005-05-04 01:00:00-04 select '2005-05-03 00:00:00 EST'::timestamp with time zone + '1 day'; ?column? ---------------------- 2005-05-04 01:00:00-04 Michael Glaesemann
2005-07-20 18:42:32 +02:00
interval->time = 0;
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
}
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
else if (range == INTERVAL_MASK(HOUR))
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
{
Add datetime macros for constants, for clarity: #define SECS_PER_DAY 86400 #define USECS_PER_DAY INT64CONST(86400000000) #define USECS_PER_HOUR INT64CONST(3600000000) #define USECS_PER_MINUTE INT64CONST(60000000) #define USECS_PER_SEC INT64CONST(1000000)
2005-05-23 20:56:55 +02:00
interval->time = (interval->time / USECS_PER_HOUR) *
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
USECS_PER_HOUR;
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
}
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
else if (range == INTERVAL_MASK(MINUTE))
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
{
Add datetime macros for constants, for clarity: #define SECS_PER_DAY 86400 #define USECS_PER_DAY INT64CONST(86400000000) #define USECS_PER_HOUR INT64CONST(3600000000) #define USECS_PER_MINUTE INT64CONST(60000000) #define USECS_PER_SEC INT64CONST(1000000)
2005-05-23 20:56:55 +02:00
interval->time = (interval->time / USECS_PER_MINUTE) *
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
USECS_PER_MINUTE;
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
}
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
else if (range == INTERVAL_MASK(SECOND))
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
{
Change AdjustIntervalForTypmod to not discard higher-order field values on the grounds that they don't fit into the specified interval qualifier (typmod). This behavior, while of long standing, is clearly wrong per spec --- for example the value INTERVAL '999' SECOND means 999 seconds and should not be reduced to less than 60 seconds. In some cases there could be grounds to raise an error if higher-order field values are not given as zero; for example '1 year 1 month'::INTERVAL MONTH should arguably be taken as an error rather than equivalent to 13 months. However our internal representation doesn't allow us to do that in a fashion that would consistently reject all and only the cases that a strict reading of the spec would suggest. Also, seeing that for example INTERVAL '13' MONTH will print out as '1 year 1 mon', we have to be careful not to create a situation where valid data will fail to dump and reload. The present patch therefore takes the attitude of not throwing an error in any such case. We might want to revisit that in future but it would take more redesign than seems prudent in late beta. Per a complaint from Sebastien Flaesch and subsequent discussion. While at other times we might have just postponed such an issue to the next development cycle, 8.4 already has changed the parsing of interval literals quite a bit in an effort to accept all spec-compliant cases correctly. This seems like a change that should be part of that rather than coming along later.
2009-06-02 01:55:15 +02:00
/* fractional-second rounding will be dealt with below */
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
}
/* DAY TO HOUR */
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
else if (range == (INTERVAL_MASK(DAY) |
INTERVAL_MASK(HOUR)))
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
{
Add datetime macros for constants, for clarity: #define SECS_PER_DAY 86400 #define USECS_PER_DAY INT64CONST(86400000000) #define USECS_PER_HOUR INT64CONST(3600000000) #define USECS_PER_MINUTE INT64CONST(60000000) #define USECS_PER_SEC INT64CONST(1000000)
2005-05-23 20:56:55 +02:00
interval->time = (interval->time / USECS_PER_HOUR) *
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
USECS_PER_HOUR;
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
}
/* DAY TO MINUTE */
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
else if (range == (INTERVAL_MASK(DAY) |
INTERVAL_MASK(HOUR) |
INTERVAL_MASK(MINUTE)))
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
{
Add datetime macros for constants, for clarity: #define SECS_PER_DAY 86400 #define USECS_PER_DAY INT64CONST(86400000000) #define USECS_PER_HOUR INT64CONST(3600000000) #define USECS_PER_MINUTE INT64CONST(60000000) #define USECS_PER_SEC INT64CONST(1000000)
2005-05-23 20:56:55 +02:00
interval->time = (interval->time / USECS_PER_MINUTE) *
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
USECS_PER_MINUTE;
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
}
/* DAY TO SECOND */
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
else if (range == (INTERVAL_MASK(DAY) |
INTERVAL_MASK(HOUR) |
INTERVAL_MASK(MINUTE) |
INTERVAL_MASK(SECOND)))
Change AdjustIntervalForTypmod to not discard higher-order field values on the grounds that they don't fit into the specified interval qualifier (typmod). This behavior, while of long standing, is clearly wrong per spec --- for example the value INTERVAL '999' SECOND means 999 seconds and should not be reduced to less than 60 seconds. In some cases there could be grounds to raise an error if higher-order field values are not given as zero; for example '1 year 1 month'::INTERVAL MONTH should arguably be taken as an error rather than equivalent to 13 months. However our internal representation doesn't allow us to do that in a fashion that would consistently reject all and only the cases that a strict reading of the spec would suggest. Also, seeing that for example INTERVAL '13' MONTH will print out as '1 year 1 mon', we have to be careful not to create a situation where valid data will fail to dump and reload. The present patch therefore takes the attitude of not throwing an error in any such case. We might want to revisit that in future but it would take more redesign than seems prudent in late beta. Per a complaint from Sebastien Flaesch and subsequent discussion. While at other times we might have just postponed such an issue to the next development cycle, 8.4 already has changed the parsing of interval literals quite a bit in an effort to accept all spec-compliant cases correctly. This seems like a change that should be part of that rather than coming along later.
2009-06-02 01:55:15 +02:00
{
/* fractional-second rounding will be dealt with below */
}
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
/* HOUR TO MINUTE */
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
else if (range == (INTERVAL_MASK(HOUR) |
INTERVAL_MASK(MINUTE)))
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
{
Add datetime macros for constants, for clarity: #define SECS_PER_DAY 86400 #define USECS_PER_DAY INT64CONST(86400000000) #define USECS_PER_HOUR INT64CONST(3600000000) #define USECS_PER_MINUTE INT64CONST(60000000) #define USECS_PER_SEC INT64CONST(1000000)
2005-05-23 20:56:55 +02:00
interval->time = (interval->time / USECS_PER_MINUTE) *
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
USECS_PER_MINUTE;
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
}
/* HOUR TO SECOND */
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
else if (range == (INTERVAL_MASK(HOUR) |
INTERVAL_MASK(MINUTE) |
INTERVAL_MASK(SECOND)))
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
{
Change AdjustIntervalForTypmod to not discard higher-order field values on the grounds that they don't fit into the specified interval qualifier (typmod). This behavior, while of long standing, is clearly wrong per spec --- for example the value INTERVAL '999' SECOND means 999 seconds and should not be reduced to less than 60 seconds. In some cases there could be grounds to raise an error if higher-order field values are not given as zero; for example '1 year 1 month'::INTERVAL MONTH should arguably be taken as an error rather than equivalent to 13 months. However our internal representation doesn't allow us to do that in a fashion that would consistently reject all and only the cases that a strict reading of the spec would suggest. Also, seeing that for example INTERVAL '13' MONTH will print out as '1 year 1 mon', we have to be careful not to create a situation where valid data will fail to dump and reload. The present patch therefore takes the attitude of not throwing an error in any such case. We might want to revisit that in future but it would take more redesign than seems prudent in late beta. Per a complaint from Sebastien Flaesch and subsequent discussion. While at other times we might have just postponed such an issue to the next development cycle, 8.4 already has changed the parsing of interval literals quite a bit in an effort to accept all spec-compliant cases correctly. This seems like a change that should be part of that rather than coming along later.
2009-06-02 01:55:15 +02:00
/* fractional-second rounding will be dealt with below */
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
}
/* MINUTE TO SECOND */
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
else if (range == (INTERVAL_MASK(MINUTE) |
INTERVAL_MASK(SECOND)))
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
{
Change AdjustIntervalForTypmod to not discard higher-order field values on the grounds that they don't fit into the specified interval qualifier (typmod). This behavior, while of long standing, is clearly wrong per spec --- for example the value INTERVAL '999' SECOND means 999 seconds and should not be reduced to less than 60 seconds. In some cases there could be grounds to raise an error if higher-order field values are not given as zero; for example '1 year 1 month'::INTERVAL MONTH should arguably be taken as an error rather than equivalent to 13 months. However our internal representation doesn't allow us to do that in a fashion that would consistently reject all and only the cases that a strict reading of the spec would suggest. Also, seeing that for example INTERVAL '13' MONTH will print out as '1 year 1 mon', we have to be careful not to create a situation where valid data will fail to dump and reload. The present patch therefore takes the attitude of not throwing an error in any such case. We might want to revisit that in future but it would take more redesign than seems prudent in late beta. Per a complaint from Sebastien Flaesch and subsequent discussion. While at other times we might have just postponed such an issue to the next development cycle, 8.4 already has changed the parsing of interval literals quite a bit in an effort to accept all spec-compliant cases correctly. This seems like a change that should be part of that rather than coming along later.
2009-06-02 01:55:15 +02:00
/* fractional-second rounding will be dealt with below */
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
}
else
Error message editing in utils/adt. Again thanks to Joe Conway for doing the bulk of the heavy lifting ...
2003-07-27 06:53:12 +02:00
elog(ERROR, "unrecognized interval typmod: %d", typmod);
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
Various minor corrections of and improvements to comments. Aleksander Alekseev
2016-03-18 14:38:59 +01:00
/* Need to adjust sub-second precision? */
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
if (precision != INTERVAL_FULL_PRECISION)
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
{
Remove unnecessary parentheses in datetime/timestamp code.
2005-05-23 19:13:14 +02:00
if (precision < 0 || precision > MAX_INTERVAL_PRECISION)
Error message editing in utils/adt. Again thanks to Joe Conway for doing the bulk of the heavy lifting ...
2003-07-27 06:53:12 +02:00
ereport(ERROR,
(errcode(ERRCODE_INVALID_PARAMETER_VALUE),
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
errmsg("interval(%d) precision must be between %d and %d",
precision, 0, MAX_INTERVAL_PRECISION)));
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
if (interval->time >= INT64CONST(0))
{
Remove more extraneous parentheses in date/time functions.
2005-05-24 04:09:45 +02:00
interval->time = ((interval->time +
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
IntervalOffsets[precision]) /
IntervalScales[precision]) *
IntervalScales[precision];
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
{
Remove unnecessary parentheses in datetime/timestamp code.
2005-05-23 19:13:14 +02:00
interval->time = -(((-interval->time +
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
IntervalOffsets[precision]) /
Remove unnecessary parentheses in datetime/timestamp code.
2005-05-23 19:13:14 +02:00
IntervalScales[precision]) *
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
IntervalScales[precision]);
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
}
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
}
}
}
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
/*
* make_interval - numeric Interval constructor
*/
Datum
make_interval(PG_FUNCTION_ARGS)
{
int32 years = PG_GETARG_INT32(0);
int32 months = PG_GETARG_INT32(1);
int32 weeks = PG_GETARG_INT32(2);
int32 days = PG_GETARG_INT32(3);
int32 hours = PG_GETARG_INT32(4);
int32 mins = PG_GETARG_INT32(5);
double secs = PG_GETARG_FLOAT8(6);
Interval *result;
Fix portability issues in recently added make_timestamp/make_interval code. Explicitly reject infinity/NaN inputs, rather than just assuming that something else will do it for us. Per buildfarm. While at it, make some over-parenthesized and under-legible code more readable.
2014-03-05 22:42:18 +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
* Reject out-of-range inputs. We really ought to check the integer
Fix portability issues in recently added make_timestamp/make_interval code. Explicitly reject infinity/NaN inputs, rather than just assuming that something else will do it for us. Per buildfarm. While at it, make some over-parenthesized and under-legible code more readable.
2014-03-05 22:42:18 +01:00
* inputs as well, but it's not entirely clear what limits to apply.
*/
if (isinf(secs) || isnan(secs))
ereport(ERROR,
(errcode(ERRCODE_DATETIME_VALUE_OUT_OF_RANGE),
errmsg("interval out of range")));
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
result = (Interval *) palloc(sizeof(Interval));
result->month = years * MONTHS_PER_YEAR + months;
result->day = weeks * 7 + days;
Blind try to fix portability issue in commit 8f93bd851 et al. The S/390 members of the buildfarm are showing failures indicating that they're having trouble with the rint() calls I added yesterday. There's no good reason for that, and I wonder if it is a compiler bug similar to the one we worked around in d9476b838. Try to fix it using the same method as before, namely to store the result of rint() back into a "double" variable rather than immediately converting to int64. (This isn't entirely waving a dead chicken, since on machines with wider-than-double float registers, the extra store forces a width conversion. I don't know if S/390 is like that, but it seems worth trying.) In passing, merge duplicate ereport() calls in float8_timestamptz(). Per buildfarm.
2017-02-09 21:49:57 +01:00
secs = rint(secs * USECS_PER_SEC);
Fix roundoff problems in float8_timestamptz() and make_interval(). When converting a float value to integer microseconds, we should be careful to round the value to the nearest integer, typically with rint(); simply assigning to an int64 variable will truncate, causing apparently off-by-one values in cases that should work. Most places in the datetime code got this right, but not these two. float8_timestamptz() is new as of commit e511d878f (9.6). Previous versions effectively depended on interval_mul() to do roundoff correctly, which it does, so this fixes an accuracy regression in 9.6. The problem in make_interval() dates to its introduction in 9.4. Aside from being careful to round not truncate, let's incorporate the hours and minutes inputs into the result with exact integer arithmetic, rather than risk introducing roundoff error where there need not have been any. float8_timestamptz() problem reported by Erik Nordström, though this is not his proposed patch. make_interval() problem found by me. Discussion: https://postgr.es/m/CAHuQZDS76jTYk3LydPbKpNfw9KbACmD=49dC4BrzHcfPv6yA1A@mail.gmail.com
2017-02-09 00:04:59 +01:00
result->time = hours * ((int64) SECS_PER_HOUR * USECS_PER_SEC) +
mins * ((int64) SECS_PER_MINUTE * USECS_PER_SEC) +
Blind try to fix portability issue in commit 8f93bd851 et al. The S/390 members of the buildfarm are showing failures indicating that they're having trouble with the rint() calls I added yesterday. There's no good reason for that, and I wonder if it is a compiler bug similar to the one we worked around in d9476b838. Try to fix it using the same method as before, namely to store the result of rint() back into a "double" variable rather than immediately converting to int64. (This isn't entirely waving a dead chicken, since on machines with wider-than-double float registers, the extra store forces a width conversion. I don't know if S/390 is like that, but it seems worth trying.) In passing, merge duplicate ereport() calls in float8_timestamptz(). Per buildfarm.
2017-02-09 21:49:57 +01:00
(int64) secs;
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
PG_RETURN_INTERVAL_P(result);
}
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
/* EncodeSpecialTimestamp()
* Convert reserved timestamp data type to string.
*/
Fix incorrect translation of minus-infinity datetimes for json/jsonb. Commit bda76c1c8cfb1d11751ba6be88f0242850481733 caused both plus and minus infinity to be rendered as "infinity", which is not only wrong but inconsistent with the pre-9.4 behavior of to_json(). Fix that by duplicating the coding in date_out/timestamp_out/timestamptz_out more closely. Per bug #13687 from Stepan Perlov. Back-patch to 9.4, like the previous commit. In passing, also re-pgindent json.c, since it had gotten a bit messed up by recent patches (and I was already annoyed by indentation-related problems in back-patching this fix ...)
2015-10-20 20:06:24 +02:00
void
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
EncodeSpecialTimestamp(Timestamp dt, char *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
if (TIMESTAMP_IS_NOBEGIN(dt))
strcpy(str, EARLY);
else if (TIMESTAMP_IS_NOEND(dt))
strcpy(str, LATE);
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
else /* shouldn't happen */
Fix EncodeSpecialTimestamp to throw error on unrecognized input, rather than returning a failure code that none of its callers bothered to check for.
2008-10-14 17:44:29 +02:00
elog(ERROR, "invalid argument for EncodeSpecialTimestamp");
}
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
Datum
now(PG_FUNCTION_ARGS)
Missed another tar file... :(
1997-03-14 06:58:13 +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
PG_RETURN_TIMESTAMPTZ(GetCurrentTransactionStartTimestamp());
Missed another tar file... :(
1997-03-14 06:58:13 +01:00
}
Add statement_timestamp(), clock_timestamp(), and transaction_timestamp() (just like now()). Also update statement_timeout() to mention it is statement arrival time that is measured. Catalog version updated.
2006-04-25 02:25:22 +02:00
Datum
statement_timestamp(PG_FUNCTION_ARGS)
{
PG_RETURN_TIMESTAMPTZ(GetCurrentStatementStartTimestamp());
}
Datum
clock_timestamp(PG_FUNCTION_ARGS)
{
PG_RETURN_TIMESTAMPTZ(GetCurrentTimestamp());
}
Add pg_postmaster_start_time() function. Euler Taveira de Oliveira Matthias Schmidt
2005-06-14 23:04:42 +02:00
Datum
Add pg_conf_load_time() function to report when the Postgres configuration files were last loaded. George Gensure
2008-05-04 23:13:36 +02:00
pg_postmaster_start_time(PG_FUNCTION_ARGS)
Add pg_postmaster_start_time() function. Euler Taveira de Oliveira Matthias Schmidt
2005-06-14 23:04:42 +02:00
{
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
PG_RETURN_TIMESTAMPTZ(PgStartTime);
Add pg_conf_load_time() function to report when the Postgres configuration files were last loaded. George Gensure
2008-05-04 23:13:36 +02:00
}
Datum
pg_conf_load_time(PG_FUNCTION_ARGS)
{
PG_RETURN_TIMESTAMPTZ(PgReloadTime);
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
}
/*
* GetCurrentTimestamp -- get the current operating system time
*
* Result is in the form of a TimestampTz value, and is expressed to the
* full precision of the gettimeofday() syscall
*/
TimestampTz
GetCurrentTimestamp(void)
{
TimestampTz result;
struct timeval tp;
gettimeofday(&tp, NULL);
Code & docs review for server instrumentation patch. File timestamps should surely be timestamptz not timestamp; fix some but not all of the holes in check_and_make_absolute(); other minor cleanup. Also put in the missed catversion bump.
2005-08-12 20:23:56 +02:00
result = (TimestampTz) tp.tv_sec -
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
((POSTGRES_EPOCH_JDATE - UNIX_EPOCH_JDATE) * SECS_PER_DAY);
Make the streaming replication protocol messages architecture-independent. We used to send structs wrapped in CopyData messages, which works as long as the client and server agree on things like endianess, timestamp format and alignment. That's good enough for running a standby server, which has to run on the same platform anyway, but it's useful for tools like pg_receivexlog to work across platforms. This breaks protocol compatibility of streaming replication, but we never promised that to be compatible across versions, anyway.
2012-11-07 17:59:12 +01:00
result = (result * USECS_PER_SEC) + tp.tv_usec;
return result;
}
Improve parsetree representation of special functions such as CURRENT_DATE. We implement a dozen or so parameterless functions that the SQL standard defines special syntax for. Up to now, that was done by converting them into more or less ad-hoc constructs such as "'now'::text::date". That's messy for multiple reasons: it exposes what should be implementation details to users, and performance is worse than it needs to be in several cases. To improve matters, invent a new expression node type SQLValueFunction that can represent any of these parameterless functions. Bump catversion because this changes stored parsetrees for rules. Discussion: <30058.1463091294@sss.pgh.pa.us>
2016-08-17 02:33:01 +02:00
/*
* GetSQLCurrentTimestamp -- implements CURRENT_TIMESTAMP, CURRENT_TIMESTAMP(n)
*/
TimestampTz
GetSQLCurrentTimestamp(int32 typmod)
{
TimestampTz ts;
ts = GetCurrentTransactionStartTimestamp();
if (typmod >= 0)
AdjustTimestampForTypmod(&ts, typmod);
return ts;
}
/*
* GetSQLLocalTimestamp -- implements LOCALTIMESTAMP, LOCALTIMESTAMP(n)
*/
Timestamp
GetSQLLocalTimestamp(int32 typmod)
{
Timestamp ts;
ts = timestamptz2timestamp(GetCurrentTransactionStartTimestamp());
if (typmod >= 0)
AdjustTimestampForTypmod(&ts, typmod);
return ts;
}
Move timeofday() implementation out of nabstime.c. nabstime.c is about to be removed, but timeofday() isn't related to the rest of the functionality therein, and some find it useful. Move to timestamp.c. Discussion: https://postgr.es/m/20181009192237.34wjp3nmw7oynmmr@alap3.anarazel.de https://postgr.es/m/20180928223240.kgwc4czzzekrpsid%40alap3.anarazel.de
2018-10-11 20:30:59 +02:00
/*
* timeofday(*) -- returns the current time as a text.
*/
Datum
timeofday(PG_FUNCTION_ARGS)
{
struct timeval tp;
char templ[128];
char buf[128];
pg_time_t tt;
gettimeofday(&tp, NULL);
tt = (pg_time_t) tp.tv_sec;
pg_strftime(templ, sizeof(templ), "%a %b %d %H:%M:%S.%%06d %Y %Z",
pg_localtime(&tt, session_timezone));
snprintf(buf, sizeof(buf), templ, tp.tv_usec);
PG_RETURN_TEXT_P(cstring_to_text(buf));
}
Remove redundant gettimeofday() calls to the extent practical without changing semantics too much. statement_timestamp is now set immediately upon receipt of a client command message, and the various places that used to do their own gettimeofday() calls to mark command startup are referenced to that instead. I have also made stats_command_string use that same value for pg_stat_activity.query_start for both the command itself and its eventual replacement by <IDLE> or <idle in transaction>. There was some debate about that, but no argument that seemed convincing enough to justify an extra gettimeofday() call.
2006-06-21 00:52:00 +02:00
/*
* TimestampDifference -- convert the difference between two timestamps
* into integer seconds and microseconds
*
* Both inputs must be ordinary finite timestamps (in current usage,
* they'll be results from GetCurrentTimestamp()).
*
Various minor corrections of and improvements to comments. Aleksander Alekseev
2016-03-18 14:38:59 +01:00
* We expect start_time <= stop_time. If not, we return zeros; for current
Remove redundant gettimeofday() calls to the extent practical without changing semantics too much. statement_timestamp is now set immediately upon receipt of a client command message, and the various places that used to do their own gettimeofday() calls to mark command startup are referenced to that instead. I have also made stats_command_string use that same value for pg_stat_activity.query_start for both the command itself and its eventual replacement by <IDLE> or <idle in transaction>. There was some debate about that, but no argument that seemed convincing enough to justify an extra gettimeofday() call.
2006-06-21 00:52:00 +02:00
* callers there is no need to be tense about which way division rounds on
* negative inputs.
*/
void
TimestampDifference(TimestampTz start_time, TimestampTz stop_time,
long *secs, int *microsecs)
{
TimestampTz diff = stop_time - start_time;
if (diff <= 0)
{
*secs = 0;
*microsecs = 0;
}
else
{
*secs = (long) (diff / USECS_PER_SEC);
*microsecs = (int) (diff % USECS_PER_SEC);
}
}
Code & docs review for server instrumentation patch. File timestamps should surely be timestamptz not timestamp; fix some but not all of the holes in check_and_make_absolute(); other minor cleanup. Also put in the missed catversion bump.
2005-08-12 20:23:56 +02:00
Implement rate-limiting logic on how often backends will attempt to send messages to the stats collector. This avoids the problem that enabling stats_row_level for autovacuum has a significant overhead for short read-only transactions, as noted by Arjen van der Meijden. We can avoid an extra gettimeofday call by piggybacking on the one done for WAL-logging xact commit or abort (although that doesn't help read-only transactions, since they don't WAL-log anything). In my proposal for this, I noted that we could change the WAL log entries for commit/abort to record full TimestampTz precision, instead of only time_t as at present. That's not done in this patch, but will be committed separately.
2007-04-30 05:23:49 +02:00
/*
* TimestampDifferenceExceeds -- report whether the difference between two
* timestamps is >= a threshold (expressed in milliseconds)
*
* Both inputs must be ordinary finite timestamps (in current usage,
* they'll be results from GetCurrentTimestamp()).
*/
bool
TimestampDifferenceExceeds(TimestampTz start_time,
TimestampTz stop_time,
int msec)
{
TimestampTz diff = stop_time - start_time;
return (diff >= msec * INT64CONST(1000));
}
Code & docs review for server instrumentation patch. File timestamps should surely be timestamptz not timestamp; fix some but not all of the holes in check_and_make_absolute(); other minor cleanup. Also put in the missed catversion bump.
2005-08-12 20:23:56 +02:00
/*
* Convert a time_t to TimestampTz.
*
* We do not use time_t internally in Postgres, but this is provided for use
* by functions that need to interpret, say, a stat(2) result.
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
*
* To avoid having the function's ABI vary depending on the width of time_t,
* we declare the argument as pg_time_t, which is cast-compatible with
* time_t but always 64 bits wide (unless the platform has no 64-bit type).
* This detail should be invisible to callers, at least at source code level.
Code & docs review for server instrumentation patch. File timestamps should surely be timestamptz not timestamp; fix some but not all of the holes in check_and_make_absolute(); other minor cleanup. Also put in the missed catversion bump.
2005-08-12 20:23:56 +02:00
*/
TimestampTz
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
time_t_to_timestamptz(pg_time_t tm)
Code & docs review for server instrumentation patch. File timestamps should surely be timestamptz not timestamp; fix some but not all of the holes in check_and_make_absolute(); other minor cleanup. Also put in the missed catversion bump.
2005-08-12 20:23:56 +02:00
{
TimestampTz result;
result = (TimestampTz) tm -
((POSTGRES_EPOCH_JDATE - UNIX_EPOCH_JDATE) * SECS_PER_DAY);
result *= USECS_PER_SEC;
return result;
}
Remove redundant gettimeofday() calls to the extent practical without changing semantics too much. statement_timestamp is now set immediately upon receipt of a client command message, and the various places that used to do their own gettimeofday() calls to mark command startup are referenced to that instead. I have also made stats_command_string use that same value for pg_stat_activity.query_start for both the command itself and its eventual replacement by <IDLE> or <idle in transaction>. There was some debate about that, but no argument that seemed convincing enough to justify an extra gettimeofday() call.
2006-06-21 00:52:00 +02:00
/*
* Convert a TimestampTz to time_t.
*
* This too is just marginally useful, but some places need it.
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
*
* To avoid having the function's ABI vary depending on the width of time_t,
* we declare the result as pg_time_t, which is cast-compatible with
* time_t but always 64 bits wide (unless the platform has no 64-bit type).
* This detail should be invisible to callers, at least at source code level.
Remove redundant gettimeofday() calls to the extent practical without changing semantics too much. statement_timestamp is now set immediately upon receipt of a client command message, and the various places that used to do their own gettimeofday() calls to mark command startup are referenced to that instead. I have also made stats_command_string use that same value for pg_stat_activity.query_start for both the command itself and its eventual replacement by <IDLE> or <idle in transaction>. There was some debate about that, but no argument that seemed convincing enough to justify an extra gettimeofday() call.
2006-06-21 00:52:00 +02:00
*/
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
pg_time_t
Remove redundant gettimeofday() calls to the extent practical without changing semantics too much. statement_timestamp is now set immediately upon receipt of a client command message, and the various places that used to do their own gettimeofday() calls to mark command startup are referenced to that instead. I have also made stats_command_string use that same value for pg_stat_activity.query_start for both the command itself and its eventual replacement by <IDLE> or <idle in transaction>. There was some debate about that, but no argument that seemed convincing enough to justify an extra gettimeofday() call.
2006-06-21 00:52:00 +02:00
timestamptz_to_time_t(TimestampTz t)
{
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
pg_time_t result;
Remove redundant gettimeofday() calls to the extent practical without changing semantics too much. statement_timestamp is now set immediately upon receipt of a client command message, and the various places that used to do their own gettimeofday() calls to mark command startup are referenced to that instead. I have also made stats_command_string use that same value for pg_stat_activity.query_start for both the command itself and its eventual replacement by <IDLE> or <idle in transaction>. There was some debate about that, but no argument that seemed convincing enough to justify an extra gettimeofday() call.
2006-06-21 00:52:00 +02:00
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
result = (pg_time_t) (t / USECS_PER_SEC +
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
((POSTGRES_EPOCH_JDATE - UNIX_EPOCH_JDATE) * SECS_PER_DAY));
Remove redundant gettimeofday() calls to the extent practical without changing semantics too much. statement_timestamp is now set immediately upon receipt of a client command message, and the various places that used to do their own gettimeofday() calls to mark command startup are referenced to that instead. I have also made stats_command_string use that same value for pg_stat_activity.query_start for both the command itself and its eventual replacement by <IDLE> or <idle in transaction>. There was some debate about that, but no argument that seemed convincing enough to justify an extra gettimeofday() call.
2006-06-21 00:52:00 +02:00
return result;
}
Change the timestamps recorded in transaction commit/abort xlog records from time_t to TimestampTz representation. This provides full gettimeofday() resolution of the timestamps, which might be useful when attempting to do point-in-time recovery --- previously it was not possible to specify the stop point with sub-second resolution. But mostly this is to get rid of TimestampTz-to-time_t conversion overhead during commit. Per my proposal of a day or two back.
2007-04-30 23:01:53 +02:00
/*
* Produce a C-string representation of a TimestampTz.
*
* This is mostly for use in emitting messages. The primary difference
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
* from timestamptz_out is that we force the output format to ISO. Note
Change the timestamps recorded in transaction commit/abort xlog records from time_t to TimestampTz representation. This provides full gettimeofday() resolution of the timestamps, which might be useful when attempting to do point-in-time recovery --- previously it was not possible to specify the stop point with sub-second resolution. But mostly this is to get rid of TimestampTz-to-time_t conversion overhead during commit. Per my proposal of a day or two back.
2007-04-30 23:01:53 +02:00
* also that the result is in a static buffer, not pstrdup'd.
*/
const char *
timestamptz_to_str(TimestampTz t)
{
pgindent run for 8.3.
2007-11-15 22:14:46 +01:00
static char buf[MAXDATELEN + 1];
Change the timestamps recorded in transaction commit/abort xlog records from time_t to TimestampTz representation. This provides full gettimeofday() resolution of the timestamps, which might be useful when attempting to do point-in-time recovery --- previously it was not possible to specify the stop point with sub-second resolution. But mostly this is to get rid of TimestampTz-to-time_t conversion overhead during commit. Per my proposal of a day or two back.
2007-04-30 23:01:53 +02:00
int tz;
struct pg_tm tt,
*tm = &tt;
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;
Change the timestamps recorded in transaction commit/abort xlog records from time_t to TimestampTz representation. This provides full gettimeofday() resolution of the timestamps, which might be useful when attempting to do point-in-time recovery --- previously it was not possible to specify the stop point with sub-second resolution. But mostly this is to get rid of TimestampTz-to-time_t conversion overhead during commit. Per my proposal of a day or two back.
2007-04-30 23:01:53 +02:00
if (TIMESTAMP_NOT_FINITE(t))
EncodeSpecialTimestamp(t, buf);
else if (timestamp2tm(t, &tz, tm, &fsec, &tzn, NULL) == 0)
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
EncodeDateTime(tm, fsec, true, tz, tzn, USE_ISO_DATES, buf);
Change the timestamps recorded in transaction commit/abort xlog records from time_t to TimestampTz representation. This provides full gettimeofday() resolution of the timestamps, which might be useful when attempting to do point-in-time recovery --- previously it was not possible to specify the stop point with sub-second resolution. But mostly this is to get rid of TimestampTz-to-time_t conversion overhead during commit. Per my proposal of a day or two back.
2007-04-30 23:01:53 +02:00
else
strlcpy(buf, "(timestamp out of range)", sizeof(buf));
return buf;
}
Code & docs review for server instrumentation patch. File timestamps should surely be timestamptz not timestamp; fix some but not all of the holes in check_and_make_absolute(); other minor cleanup. Also put in the missed catversion bump.
2005-08-12 20:23:56 +02:00
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
void
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
dt2time(Timestamp jd, int *hour, int *min, int *sec, fsec_t *fsec)
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
{
Get rid of a bunch of #ifdef HAVE_INT64_TIMESTAMP conditionals by inventing a new typedef TimeOffset to represent an intermediate time value. It's either int64 or double as appropriate, and in most usages will be measured in microseconds or seconds the same as Timestamp. We don't call it Timestamp, though, since the value doesn't necessarily represent an absolute time instant. Warren Turkal
2008-03-21 02:31:43 +01:00
TimeOffset time;
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
time = jd;
Add datetime macros for constants, for clarity: #define SECS_PER_DAY 86400 #define USECS_PER_DAY INT64CONST(86400000000) #define USECS_PER_HOUR INT64CONST(3600000000) #define USECS_PER_MINUTE INT64CONST(60000000) #define USECS_PER_SEC INT64CONST(1000000)
2005-05-23 20:56:55 +02:00
*hour = time / USECS_PER_HOUR;
time -= (*hour) * USECS_PER_HOUR;
*min = time / USECS_PER_MINUTE;
time -= (*min) * USECS_PER_MINUTE;
*sec = time / USECS_PER_SEC;
*fsec = time - (*sec * USECS_PER_SEC);
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
} /* dt2time() */
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +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
/*
* timestamp2tm() - Convert timestamp data type to POSIX time structure.
*
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
* Note that year is _not_ 1900-based, but is an explicit full value.
* Also, month is one-based, _not_ zero-based.
* Returns:
* 0 on success
* -1 on out of range
This patch makes it possible to use the full set of timezones when doing "AT TIME ZONE", and not just the shorlist previously available. For example: SELECT CURRENT_TIMESTAMP AT TIME ZONE 'Europe/London'; works fine now. It will also obey whatever DST rules were in effect at just that date, which the previous implementation did not. It also supports the AT TIME ZONE on the timetz datatype. The whole handling of DST is a bit bogus there, so I chose to make it use whatever DST rules are in effect at the time of executig the query. not sure if anybody is actuallyi *using* timetz though, it seems pretty unpredictable just because of this... Magnus Hagander
2005-06-15 02:34:11 +02:00
*
Remove internal uses of CTimeZone/HasCTZSet. The only remaining places where we actually look at CTimeZone/HasCTZSet are abstime2tm() and timestamp2tm(). Now that session_timezone is always valid, we can remove these special cases. The caller-visible impact of this is that these functions now always return a valid zone abbreviation if requested, whereas before they'd return a NULL pointer if a brute-force timezone was in use. In the existing code, the only place I can find that changes behavior is to_char(), whose TZ format code will now print something useful rather than nothing for such zones. (In the places where the returned zone abbreviation is passed to EncodeDateTime, the lack of visible change is because we've chosen the abbreviation used for these zones to match what EncodeTimezone would have printed.) It's likely that there is now a fair amount of removable dead code around the call sites, namely anything that's meant to cope with getting a NULL timezone abbreviation, but I've not made an effort to root that out. This could be back-patched if we decide we'd like to fix to_char()'s behavior in the back branches, but there doesn't seem to be much enthusiasm for that at present.
2013-11-01 17:51:27 +01:00
* If attimezone is NULL, the global timezone setting will be used.
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
*/
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
timestamp2tm(Timestamp dt, int *tzp, struct pg_tm *tm, fsec_t *fsec, const char **tzn, pg_tz *attimezone)
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
{
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
Timestamp date;
Add range-checking in timestamp_recv and timestamptz_recv, per Stephen Frost. Also tighten date range check in timestamp2tm.
2004-06-03 19:57:09 +02:00
Timestamp time;
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
pg_time_t utime;
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
Remove internal uses of CTimeZone/HasCTZSet. The only remaining places where we actually look at CTimeZone/HasCTZSet are abstime2tm() and timestamp2tm(). Now that session_timezone is always valid, we can remove these special cases. The caller-visible impact of this is that these functions now always return a valid zone abbreviation if requested, whereas before they'd return a NULL pointer if a brute-force timezone was in use. In the existing code, the only place I can find that changes behavior is to_char(), whose TZ format code will now print something useful rather than nothing for such zones. (In the places where the returned zone abbreviation is passed to EncodeDateTime, the lack of visible change is because we've chosen the abbreviation used for these zones to match what EncodeTimezone would have printed.) It's likely that there is now a fair amount of removable dead code around the call sites, namely anything that's meant to cope with getting a NULL timezone abbreviation, but I've not made an effort to root that out. This could be back-patched if we decide we'd like to fix to_char()'s behavior in the back branches, but there doesn't seem to be much enthusiasm for that at present.
2013-11-01 17:51:27 +01:00
/* Use session timezone if caller asks for default */
if (attimezone == NULL)
attimezone = session_timezone;
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
Fix (hopefully for the last time) problems with datetime values displaying like '23:59:60' because of fractional-second roundoff problems. Trying to control this upstream of the actual display code was hopeless; the right way is to explicitly round fractional seconds in the display code and then refigure the results if the fraction rounds up to 1. Per bug #1927.
2005-10-09 19:21:47 +02:00
time = dt;
Add datetime macros for constants, for clarity: #define SECS_PER_DAY 86400 #define USECS_PER_DAY INT64CONST(86400000000) #define USECS_PER_HOUR INT64CONST(3600000000) #define USECS_PER_MINUTE INT64CONST(60000000) #define USECS_PER_SEC INT64CONST(1000000)
2005-05-23 20:56:55 +02:00
TMODULO(time, date, USECS_PER_DAY);
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
if (time < INT64CONST(0))
{
Add datetime macros for constants, for clarity: #define SECS_PER_DAY 86400 #define USECS_PER_DAY INT64CONST(86400000000) #define USECS_PER_HOUR INT64CONST(3600000000) #define USECS_PER_MINUTE INT64CONST(60000000) #define USECS_PER_SEC INT64CONST(1000000)
2005-05-23 20:56:55 +02:00
time += USECS_PER_DAY;
Remove unnecessary parentheses in datetime/timestamp code.
2005-05-23 19:13:14 +02:00
date -= 1;
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
}
Fix (hopefully for the last time) problems with datetime values displaying like '23:59:60' because of fractional-second roundoff problems. Trying to control this upstream of the actual display code was hopeless; the right way is to explicitly round fractional seconds in the display code and then refigure the results if the fraction rounds up to 1. Per bug #1927.
2005-10-09 19:21:47 +02:00
/* add offset to go from J2000 back to standard Julian date */
date += POSTGRES_EPOCH_JDATE;
/* Julian day routine does not work for negative Julian days */
if (date < 0 || date > (Timestamp) INT_MAX)
return -1;
j2date((int) date, &tm->tm_year, &tm->tm_mon, &tm->tm_mday);
dt2time(time, &tm->tm_hour, &tm->tm_min, &tm->tm_sec, fsec);
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
/* Done if no TZ conversion wanted */
if (tzp == NULL)
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +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
tm->tm_isdst = -1;
tm->tm_gmtoff = 0;
tm->tm_zone = NULL;
if (tzn != NULL)
*tzn = NULL;
return 0;
}
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
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
* If the time falls within the range of pg_time_t, use pg_localtime() to
* rotate to the local time zone.
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, convert to an integral timestamp, avoiding possibly
* platform-specific roundoff-in-wrong-direction errors, and adjust to
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
* Unix epoch. Then see if we can convert to pg_time_t without loss. This
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* coding avoids hardwiring any assumptions about the width of pg_time_t,
* so it should behave sanely on machines without int64.
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
*/
Add datetime macros for constants, for clarity: #define SECS_PER_DAY 86400 #define USECS_PER_DAY INT64CONST(86400000000) #define USECS_PER_HOUR INT64CONST(3600000000) #define USECS_PER_MINUTE INT64CONST(60000000) #define USECS_PER_SEC INT64CONST(1000000)
2005-05-23 20:56:55 +02:00
dt = (dt - *fsec) / USECS_PER_SEC +
More macro cleanups for date/time.
2005-05-23 23:54:02 +02:00
(POSTGRES_EPOCH_JDATE - UNIX_EPOCH_JDATE) * SECS_PER_DAY;
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
utime = (pg_time_t) dt;
if ((Timestamp) utime == dt)
{
Remove internal uses of CTimeZone/HasCTZSet. The only remaining places where we actually look at CTimeZone/HasCTZSet are abstime2tm() and timestamp2tm(). Now that session_timezone is always valid, we can remove these special cases. The caller-visible impact of this is that these functions now always return a valid zone abbreviation if requested, whereas before they'd return a NULL pointer if a brute-force timezone was in use. In the existing code, the only place I can find that changes behavior is to_char(), whose TZ format code will now print something useful rather than nothing for such zones. (In the places where the returned zone abbreviation is passed to EncodeDateTime, the lack of visible change is because we've chosen the abbreviation used for these zones to match what EncodeTimezone would have printed.) It's likely that there is now a fair amount of removable dead code around the call sites, namely anything that's meant to cope with getting a NULL timezone abbreviation, but I've not made an effort to root that out. This could be back-patched if we decide we'd like to fix to_char()'s behavior in the back branches, but there doesn't seem to be much enthusiasm for that at present.
2013-11-01 17:51:27 +01:00
struct pg_tm *tx = pg_localtime(&utime, attimezone);
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
tm->tm_year = tx->tm_year + 1900;
tm->tm_mon = tx->tm_mon + 1;
tm->tm_mday = tx->tm_mday;
tm->tm_hour = tx->tm_hour;
tm->tm_min = tx->tm_min;
tm->tm_sec = tx->tm_sec;
tm->tm_isdst = tx->tm_isdst;
tm->tm_gmtoff = tx->tm_gmtoff;
tm->tm_zone = tx->tm_zone;
More removal of unneeded parentheses.
2005-07-22 21:00:55 +02:00
*tzp = -tm->tm_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
if (tzn != NULL)
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
*tzn = tm->tm_zone;
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
}
else
{
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
/*
* When out of range of pg_time_t, treat as GMT
*/
*tzp = 0;
/* Mark this as *no* time zone available */
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
tm->tm_isdst = -1;
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
tm->tm_gmtoff = 0;
tm->tm_zone = NULL;
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
if (tzn != NULL)
*tzn = NULL;
}
return 0;
Error message editing in utils/adt. Again thanks to Joe Conway for doing the bulk of the heavy lifting ...
2003-07-27 06:53:12 +02:00
}
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
/* tm2timestamp()
* Convert a tm structure to a timestamp data type.
* Note that year is _not_ 1900-based, but is an explicit full value.
* Also, month is one-based, _not_ zero-based.
tm2timestamp should return -1, not elog, on overflow. (In the backend this is merely an API inconsistency, but in ecpg it's fatal.) Also, fix misconceived overflow test in HAVE_INT64_TIMESTAMP case.
2003-07-04 20:21:14 +02:00
*
Error message editing in utils/adt. Again thanks to Joe Conway for doing the bulk of the heavy lifting ...
2003-07-27 06:53:12 +02:00
* Returns -1 on failure (value out of range).
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
*/
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
tm2timestamp(struct pg_tm *tm, fsec_t fsec, int *tzp, Timestamp *result)
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
{
Get rid of a bunch of #ifdef HAVE_INT64_TIMESTAMP conditionals by inventing a new typedef TimeOffset to represent an intermediate time value. It's either int64 or double as appropriate, and in most usages will be measured in microseconds or seconds the same as Timestamp. We don't call it Timestamp, though, since the value doesn't necessarily represent an absolute time instant. Warren Turkal
2008-03-21 02:31:43 +01:00
TimeOffset date;
TimeOffset time;
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
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
/* Prevent overflow in Julian-day routines */
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
if (!IS_VALID_JULIAN(tm->tm_year, tm->tm_mon, tm->tm_mday))
Suppress a few 'uninitialized variable' warnings that gcc emits only at -O3 or higher (presumably because it inlines more things). Per gripe from Mark Mielke.
2006-11-11 02:14:19 +01:00
{
*result = 0; /* keep compiler quiet */
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
return -1;
Suppress a few 'uninitialized variable' warnings that gcc emits only at -O3 or higher (presumably because it inlines more things). Per gripe from Mark Mielke.
2006-11-11 02:14:19 +01:00
}
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +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
date = date2j(tm->tm_year, tm->tm_mon, tm->tm_mday) - POSTGRES_EPOCH_JDATE;
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
time = time2t(tm->tm_hour, tm->tm_min, tm->tm_sec, fsec);
Fix AT TIME ZONE for timestamps without time zones: test=> select ('2005-07-20 00:00:00'::timestamp without time zone) at time zone 'Europe/Paris'; timezone ------------------------ 2005-07-19 22:00:00-04 Udpate documentation.
2005-07-22 23:16:15 +02:00
Add datetime macros for constants, for clarity: #define SECS_PER_DAY 86400 #define USECS_PER_DAY INT64CONST(86400000000) #define USECS_PER_HOUR INT64CONST(3600000000) #define USECS_PER_MINUTE INT64CONST(60000000) #define USECS_PER_SEC INT64CONST(1000000)
2005-05-23 20:56:55 +02:00
*result = date * USECS_PER_DAY + time;
tm2timestamp should return -1, not elog, on overflow. (In the backend this is merely an API inconsistency, but in ecpg it's fatal.) Also, fix misconceived overflow test in HAVE_INT64_TIMESTAMP case.
2003-07-04 20:21:14 +02:00
/* check for major overflow */
Add datetime macros for constants, for clarity: #define SECS_PER_DAY 86400 #define USECS_PER_DAY INT64CONST(86400000000) #define USECS_PER_HOUR INT64CONST(3600000000) #define USECS_PER_MINUTE INT64CONST(60000000) #define USECS_PER_SEC INT64CONST(1000000)
2005-05-23 20:56:55 +02:00
if ((*result - time) / USECS_PER_DAY != date)
Suppress a few 'uninitialized variable' warnings that gcc emits only at -O3 or higher (presumably because it inlines more things). Per gripe from Mark Mielke.
2006-11-11 02:14:19 +01:00
{
*result = 0; /* keep compiler quiet */
tm2timestamp should return -1, not elog, on overflow. (In the backend this is merely an API inconsistency, but in ecpg it's fatal.) Also, fix misconceived overflow test in HAVE_INT64_TIMESTAMP case.
2003-07-04 20:21:14 +02:00
return -1;
Suppress a few 'uninitialized variable' warnings that gcc emits only at -O3 or higher (presumably because it inlines more things). Per gripe from Mark Mielke.
2006-11-11 02:14:19 +01:00
}
tm2timestamp should return -1, not elog, on overflow. (In the backend this is merely an API inconsistency, but in ecpg it's fatal.) Also, fix misconceived overflow test in HAVE_INT64_TIMESTAMP case.
2003-07-04 20:21:14 +02:00
/* check for just-barely overflow (okay except time-of-day wraps) */
Fix overflow check in tm2timestamp (this time for sure). I fixed this code back in commit 841b4a2d5, but didn't think carefully enough about the behavior near zero, which meant it improperly rejected 1999-12-31 24:00:00. Per report from Magnus Hagander.
2013-03-04 21:13:31 +01:00
/* caution: we want to allow 1999-12-31 24:00:00 */
if ((*result < 0 && date > 0) ||
(*result > 0 && date < -1))
Suppress a few 'uninitialized variable' warnings that gcc emits only at -O3 or higher (presumably because it inlines more things). Per gripe from Mark Mielke.
2006-11-11 02:14:19 +01:00
{
*result = 0; /* keep compiler quiet */
tm2timestamp should return -1, not elog, on overflow. (In the backend this is merely an API inconsistency, but in ecpg it's fatal.) Also, fix misconceived overflow test in HAVE_INT64_TIMESTAMP case.
2003-07-04 20:21:14 +02:00
return -1;
Suppress a few 'uninitialized variable' warnings that gcc emits only at -O3 or higher (presumably because it inlines more things). Per gripe from Mark Mielke.
2006-11-11 02:14:19 +01:00
}
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
if (tzp != NULL)
*result = dt2local(*result, -(*tzp));
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
/* final range check catches just-out-of-range timestamps */
if (!IS_VALID_TIMESTAMP(*result))
{
*result = 0; /* keep compiler quiet */
return -1;
}
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
return 0;
Error message editing in utils/adt. Again thanks to Joe Conway for doing the bulk of the heavy lifting ...
2003-07-27 06:53:12 +02:00
}
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
/* interval2tm()
Collection of typo fixes. Use "a" and "an" correctly, mostly in comments. Two error messages were also fixed (they were just elogs, so no translation work required). Two function comments in pg_proc.h were also fixed. Etsuro Fujita reported one of these, but I found a lot more with grep. Also fix a few other typos spotted while grepping for the a/an typos. For example, "consists out of ..." -> "consists of ...". Plus a "though"/ "through" mixup reported by Euler Taveira. Many of these typos were in old code, which would be nice to backpatch to make future backpatching easier. But much of the code was new, and I didn't feel like crafting separate patches for each branch. So no backpatching.
2015-05-20 15:18:11 +02:00
* Convert an interval data type to a tm structure.
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
*/
- new to_char(interval, text) - new millisecond (ms) and microsecond (us) support - more robus parsing from string - used is separator checking for non-exact formats like to_date('2001-9-1', 'YYYY-MM-DD') - SGML docs are included Karel Zak
2001-09-06 05:22:42 +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
interval2tm(Interval span, struct pg_tm *tm, fsec_t *fsec)
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
{
Get rid of a bunch of #ifdef HAVE_INT64_TIMESTAMP conditionals by inventing a new typedef TimeOffset to represent an intermediate time value. It's either int64 or double as appropriate, and in most usages will be measured in microseconds or seconds the same as Timestamp. We don't call it Timestamp, though, since the value doesn't necessarily represent an absolute time instant. Warren Turkal
2008-03-21 02:31:43 +01:00
TimeOffset time;
TimeOffset tfrac;
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
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
tm->tm_year = span.month / MONTHS_PER_YEAR;
tm->tm_mon = span.month % MONTHS_PER_YEAR;
Add 'day' field to INTERVAL so 1 day interval can be distinguished from 24 hours. This is very helpful for daylight savings time: select '2005-05-03 00:00:00 EST'::timestamp with time zone + '24 hours'; ?column? ---------------------- 2005-05-04 01:00:00-04 select '2005-05-03 00:00:00 EST'::timestamp with time zone + '1 day'; ?column? ---------------------- 2005-05-04 01:00:00-04 Michael Glaesemann
2005-07-20 18:42:32 +02:00
tm->tm_mday = span.day;
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
time = span.time;
Remove justify_hours call from interval_mul and interval_div, and make some small stylistic improvements in these functions. Also fix several places where TMODULO() was being used with wrong-sized quotient argument, creating a risk of overflow --- interval2tm was actually capable of going into an infinite loop because of this.
2005-10-25 19:13:07 +02:00
tfrac = time / USECS_PER_HOUR;
time -= tfrac * USECS_PER_HOUR;
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
tm->tm_hour = tfrac;
if (!SAMESIGN(tm->tm_hour, tfrac))
ereport(ERROR,
(errcode(ERRCODE_DATETIME_VALUE_OUT_OF_RANGE),
errmsg("interval out of range")));
Remove justify_hours call from interval_mul and interval_div, and make some small stylistic improvements in these functions. Also fix several places where TMODULO() was being used with wrong-sized quotient argument, creating a risk of overflow --- interval2tm was actually capable of going into an infinite loop because of this.
2005-10-25 19:13:07 +02:00
tfrac = time / USECS_PER_MINUTE;
time -= tfrac * USECS_PER_MINUTE;
tm->tm_min = tfrac;
tfrac = time / USECS_PER_SEC;
*fsec = time - (tfrac * USECS_PER_SEC);
tm->tm_sec = tfrac;
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
return 0;
Error message editing in utils/adt. Again thanks to Joe Conway for doing the bulk of the heavy lifting ...
2003-07-27 06:53:12 +02:00
}
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
- new to_char(interval, text) - new millisecond (ms) and microsecond (us) support - more robus parsing from string - used is separator checking for non-exact formats like to_date('2001-9-1', 'YYYY-MM-DD') - SGML docs are included Karel Zak
2001-09-06 05:22:42 +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
tm2interval(struct pg_tm *tm, fsec_t fsec, Interval *span)
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +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
double total_months = (double) tm->tm_year * MONTHS_PER_YEAR + tm->tm_mon;
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
if (total_months > INT_MAX || total_months < INT_MIN)
return -1;
span->month = total_months;
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
span->day = tm->tm_mday;
Fix integer timestamp build for macro changes.
2005-07-21 07:18:26 +02:00
span->time = (((((tm->tm_hour * INT64CONST(60)) +
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
tm->tm_min) * INT64CONST(60)) +
tm->tm_sec) * USECS_PER_SEC) + fsec;
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
return 0;
Error message editing in utils/adt. Again thanks to Joe Conway for doing the bulk of the heavy lifting ...
2003-07-27 06:53:12 +02:00
}
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
Get rid of a bunch of #ifdef HAVE_INT64_TIMESTAMP conditionals by inventing a new typedef TimeOffset to represent an intermediate time value. It's either int64 or double as appropriate, and in most usages will be measured in microseconds or seconds the same as Timestamp. We don't call it Timestamp, though, since the value doesn't necessarily represent an absolute time instant. Warren Turkal
2008-03-21 02:31:43 +01:00
static TimeOffset
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
time2t(const int hour, const int min, const int sec, const fsec_t fsec)
{
Update DAYS_PER_MONTH comment. Add SECS_PER_YEAR and MINS_PER_HOUR macros.
2005-07-21 20:06:13 +02:00
return (((((hour * MINS_PER_HOUR) + min) * SECS_PER_MINUTE) + sec) * USECS_PER_SEC) + fsec;
Get rid of a bunch of #ifdef HAVE_INT64_TIMESTAMP conditionals by inventing a new typedef TimeOffset to represent an intermediate time value. It's either int64 or double as appropriate, and in most usages will be measured in microseconds or seconds the same as Timestamp. We don't call it Timestamp, though, since the value doesn't necessarily represent an absolute time instant. Warren Turkal
2008-03-21 02:31:43 +01:00
}
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
Mark functions as static and ifdef NOT_USED as appropriate.
2000-06-09 00:38:00 +02:00
static Timestamp
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
dt2local(Timestamp dt, int tz)
{
Add datetime macros for constants, for clarity: #define SECS_PER_DAY 86400 #define USECS_PER_DAY INT64CONST(86400000000) #define USECS_PER_HOUR INT64CONST(3600000000) #define USECS_PER_MINUTE INT64CONST(60000000) #define USECS_PER_SEC INT64CONST(1000000)
2005-05-23 20:56:55 +02:00
dt -= (tz * USECS_PER_SEC);
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
return dt;
Get rid of a bunch of #ifdef HAVE_INT64_TIMESTAMP conditionals by inventing a new typedef TimeOffset to represent an intermediate time value. It's either int64 or double as appropriate, and in most usages will be measured in microseconds or seconds the same as Timestamp. We don't call it Timestamp, though, since the value doesn't necessarily represent an absolute time instant. Warren Turkal
2008-03-21 02:31:43 +01:00
}
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
/*****************************************************************************
* PUBLIC ROUTINES *
*****************************************************************************/
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
Datum
timestamp_finite(PG_FUNCTION_ARGS)
Missed another tar file... :(
1997-03-14 06:58:13 +01:00
{
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
Timestamp timestamp = PG_GETARG_TIMESTAMP(0);
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
pgindent run. Make it all clean.
2001-03-22 05:01:46 +01:00
PG_RETURN_BOOL(!TIMESTAMP_NOT_FINITE(timestamp));
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
}
Missed another tar file... :(
1997-03-14 06:58:13 +01:00
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
Datum
interval_finite(PG_FUNCTION_ARGS)
Missed another tar file... :(
1997-03-14 06:58:13 +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
PG_RETURN_BOOL(true);
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
}
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
/*----------------------------------------------------------
* Relational operators for timestamp.
*---------------------------------------------------------*/
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
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
GetEpochTime(struct pg_tm *tm)
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
{
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
struct pg_tm *t0;
pg_time_t epoch = 0;
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
Integrate src/timezone library for all platforms. There is more we can and should do now that we control our own destiny for timezone handling, but this commit gets the bulk of the picayune diffs in place. Magnus Hagander and Tom Lane.
2004-05-21 07:08:06 +02:00
t0 = pg_gmtime(&epoch);
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
Avoid statically allocating gmtsub()'s timezone workspace. localtime.c's "struct state" is a rather large object, ~23KB. We were statically allocating one for gmtsub() to use to represent the GMT timezone, even though that function is not at all heavily used and is never reached in most backends. Let's malloc it on-demand, instead. This does pose the question of how to handle a malloc failure, but there's already a well-defined error report convention here, ie set errno and return NULL. We have but one caller of pg_gmtime in HEAD, and two in back branches, neither of which were troubling to check for error. Make them do so. The possible errors are sufficiently unlikely (out-of-range timestamp, and now malloc failure) that I think elog() is adequate. Back-patch to all supported branches to keep our copies of the IANA timezone code in sync. This particular change is in a stanza that already differs from upstream, so it's a wash for maintenance purposes --- but only as long as we keep the branches the same. Discussion: https://postgr.es/m/20181015200754.7y7zfuzsoux2c4ya@alap3.anarazel.de
2018-10-16 17:50:18 +02:00
if (t0 == NULL)
elog(ERROR, "could not convert epoch to timestamp: %m");
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
tm->tm_year = t0->tm_year;
tm->tm_mon = t0->tm_mon;
tm->tm_mday = t0->tm_mday;
tm->tm_hour = t0->tm_hour;
tm->tm_min = t0->tm_min;
tm->tm_sec = t0->tm_sec;
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
tm->tm_year += 1900;
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
tm->tm_mon++;
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
}
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
Timestamp
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
SetEpochTimestamp(void)
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +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
Timestamp dt;
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
struct pg_tm tt,
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 = &tt;
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +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
GetEpochTime(tm);
Error message editing in utils/adt. Again thanks to Joe Conway for doing the bulk of the heavy lifting ...
2003-07-27 06:53:12 +02:00
/* we don't bother to test for failure ... */
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
tm2timestamp(tm, 0, NULL, &dt);
Missed another tar file... :(
1997-03-14 06:58:13 +01:00
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
return dt;
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
} /* SetEpochTimestamp() */
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
/*
Fix floating-point timestamp comparisons to not go nuts if NaN is encountered; per bug report from Christian van der Leeden 8/7/03. Also, adjust larger/smaller routines (MAX/MIN) to share code with comparisons for timestamp, interval, timetz.
2003-08-08 02:10:31 +02:00
* We are currently sharing some code between timestamp and timestamptz.
* The comparison functions are among them. - thomas 2001-09-25
*
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
* timestamp_relop - is timestamp1 relop timestamp2
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
*/
Create crosstype comparison operators for date vs. timestamp and date vs. timestamptz. This allows use of indexes for expressions like datecol >= date 'today' - interval '1 month' which were formerly not indexable without casting the righthand side down from timestamp to date.
2004-02-14 21:16:18 +01:00
int
Ensure that btree sort ordering functions and boolean comparison operators give consistent results for all datatypes. Types float4, float8, and numeric were broken for NaN values; abstime, timestamp, and interval were broken for INVALID values; timetz was just plain broken (some possible pairs of values were neither < nor = nor >). Also clean up text, bpchar, varchar, and bit/varbit to eliminate duplicate code and thereby reduce the probability of similar inconsistencies arising in the future.
2001-05-03 21:00:37 +02:00
timestamp_cmp_internal(Timestamp dt1, Timestamp dt2)
{
Remove more extraneous parentheses in date/time functions.
2005-05-24 04:09:45 +02:00
return (dt1 < dt2) ? -1 : ((dt1 > dt2) ? 1 : 0);
Ensure that btree sort ordering functions and boolean comparison operators give consistent results for all datatypes. Types float4, float8, and numeric were broken for NaN values; abstime, timestamp, and interval were broken for INVALID values; timetz was just plain broken (some possible pairs of values were neither < nor = nor >). Also clean up text, bpchar, varchar, and bit/varbit to eliminate duplicate code and thereby reduce the probability of similar inconsistencies arising in the future.
2001-05-03 21:00:37 +02:00
}
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
Datum
timestamp_eq(PG_FUNCTION_ARGS)
Missed another tar file... :(
1997-03-14 06:58:13 +01:00
{
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
Timestamp dt1 = PG_GETARG_TIMESTAMP(0);
Timestamp dt2 = PG_GETARG_TIMESTAMP(1);
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
Ensure that btree sort ordering functions and boolean comparison operators give consistent results for all datatypes. Types float4, float8, and numeric were broken for NaN values; abstime, timestamp, and interval were broken for INVALID values; timetz was just plain broken (some possible pairs of values were neither < nor = nor >). Also clean up text, bpchar, varchar, and bit/varbit to eliminate duplicate code and thereby reduce the probability of similar inconsistencies arising in the future.
2001-05-03 21:00:37 +02:00
PG_RETURN_BOOL(timestamp_cmp_internal(dt1, dt2) == 0);
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
}
Missed another tar file... :(
1997-03-14 06:58:13 +01:00
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
Datum
timestamp_ne(PG_FUNCTION_ARGS)
Missed another tar file... :(
1997-03-14 06:58:13 +01:00
{
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
Timestamp dt1 = PG_GETARG_TIMESTAMP(0);
Timestamp dt2 = PG_GETARG_TIMESTAMP(1);
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
Ensure that btree sort ordering functions and boolean comparison operators give consistent results for all datatypes. Types float4, float8, and numeric were broken for NaN values; abstime, timestamp, and interval were broken for INVALID values; timetz was just plain broken (some possible pairs of values were neither < nor = nor >). Also clean up text, bpchar, varchar, and bit/varbit to eliminate duplicate code and thereby reduce the probability of similar inconsistencies arising in the future.
2001-05-03 21:00:37 +02:00
PG_RETURN_BOOL(timestamp_cmp_internal(dt1, dt2) != 0);
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
}
Missed another tar file... :(
1997-03-14 06:58:13 +01:00
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
Datum
timestamp_lt(PG_FUNCTION_ARGS)
Missed another tar file... :(
1997-03-14 06:58:13 +01:00
{
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
Timestamp dt1 = PG_GETARG_TIMESTAMP(0);
Timestamp dt2 = PG_GETARG_TIMESTAMP(1);
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
Ensure that btree sort ordering functions and boolean comparison operators give consistent results for all datatypes. Types float4, float8, and numeric were broken for NaN values; abstime, timestamp, and interval were broken for INVALID values; timetz was just plain broken (some possible pairs of values were neither < nor = nor >). Also clean up text, bpchar, varchar, and bit/varbit to eliminate duplicate code and thereby reduce the probability of similar inconsistencies arising in the future.
2001-05-03 21:00:37 +02:00
PG_RETURN_BOOL(timestamp_cmp_internal(dt1, dt2) < 0);
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
}
Missed another tar file... :(
1997-03-14 06:58:13 +01:00
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
Datum
timestamp_gt(PG_FUNCTION_ARGS)
Missed another tar file... :(
1997-03-14 06:58:13 +01:00
{
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
Timestamp dt1 = PG_GETARG_TIMESTAMP(0);
Timestamp dt2 = PG_GETARG_TIMESTAMP(1);
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
Ensure that btree sort ordering functions and boolean comparison operators give consistent results for all datatypes. Types float4, float8, and numeric were broken for NaN values; abstime, timestamp, and interval were broken for INVALID values; timetz was just plain broken (some possible pairs of values were neither < nor = nor >). Also clean up text, bpchar, varchar, and bit/varbit to eliminate duplicate code and thereby reduce the probability of similar inconsistencies arising in the future.
2001-05-03 21:00:37 +02:00
PG_RETURN_BOOL(timestamp_cmp_internal(dt1, dt2) > 0);
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
}
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
Datum
timestamp_le(PG_FUNCTION_ARGS)
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
{
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
Timestamp dt1 = PG_GETARG_TIMESTAMP(0);
Timestamp dt2 = PG_GETARG_TIMESTAMP(1);
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
Ensure that btree sort ordering functions and boolean comparison operators give consistent results for all datatypes. Types float4, float8, and numeric were broken for NaN values; abstime, timestamp, and interval were broken for INVALID values; timetz was just plain broken (some possible pairs of values were neither < nor = nor >). Also clean up text, bpchar, varchar, and bit/varbit to eliminate duplicate code and thereby reduce the probability of similar inconsistencies arising in the future.
2001-05-03 21:00:37 +02:00
PG_RETURN_BOOL(timestamp_cmp_internal(dt1, dt2) <= 0);
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
}
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
Datum
timestamp_ge(PG_FUNCTION_ARGS)
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
{
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
Timestamp dt1 = PG_GETARG_TIMESTAMP(0);
Timestamp dt2 = PG_GETARG_TIMESTAMP(1);
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
Ensure that btree sort ordering functions and boolean comparison operators give consistent results for all datatypes. Types float4, float8, and numeric were broken for NaN values; abstime, timestamp, and interval were broken for INVALID values; timetz was just plain broken (some possible pairs of values were neither < nor = nor >). Also clean up text, bpchar, varchar, and bit/varbit to eliminate duplicate code and thereby reduce the probability of similar inconsistencies arising in the future.
2001-05-03 21:00:37 +02:00
PG_RETURN_BOOL(timestamp_cmp_internal(dt1, dt2) >= 0);
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
}
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
Datum
timestamp_cmp(PG_FUNCTION_ARGS)
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
{
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
Timestamp dt1 = PG_GETARG_TIMESTAMP(0);
Timestamp dt2 = PG_GETARG_TIMESTAMP(1);
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
Ensure that btree sort ordering functions and boolean comparison operators give consistent results for all datatypes. Types float4, float8, and numeric were broken for NaN values; abstime, timestamp, and interval were broken for INVALID values; timetz was just plain broken (some possible pairs of values were neither < nor = nor >). Also clean up text, bpchar, varchar, and bit/varbit to eliminate duplicate code and thereby reduce the probability of similar inconsistencies arising in the future.
2001-05-03 21:00:37 +02:00
PG_RETURN_INT32(timestamp_cmp_internal(dt1, dt2));
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
}
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
Create a "sort support" interface API for faster sorting. This patch creates an API whereby a btree index opclass can optionally provide non-SQL-callable support functions for sorting. In the initial patch, we only use this to provide a directly-callable comparator function, which can be invoked with a bit less overhead than the traditional SQL-callable comparator. While that should be of value in itself, the real reason for doing this is to provide a datatype-extensible framework for more aggressive optimizations, as in Peter Geoghegan's recent work. Robert Haas and Tom Lane
2011-12-07 06:18:38 +01:00
/* note: this is used for timestamptz also */
static int
timestamp_fastcmp(Datum x, Datum y, SortSupport ssup)
{
Timestamp a = DatumGetTimestamp(x);
Timestamp b = DatumGetTimestamp(y);
return timestamp_cmp_internal(a, b);
}
Datum
timestamp_sortsupport(PG_FUNCTION_ARGS)
{
Run pgindent on 9.2 source tree in preparation for first 9.3 commit-fest.
2012-06-10 21:20:04 +02:00
SortSupport ssup = (SortSupport) PG_GETARG_POINTER(0);
Create a "sort support" interface API for faster sorting. This patch creates an API whereby a btree index opclass can optionally provide non-SQL-callable support functions for sorting. In the initial patch, we only use this to provide a directly-callable comparator function, which can be invoked with a bit less overhead than the traditional SQL-callable comparator. While that should be of value in itself, the real reason for doing this is to provide a datatype-extensible framework for more aggressive optimizations, as in Peter Geoghegan's recent work. Robert Haas and Tom Lane
2011-12-07 06:18:38 +01:00
ssup->comparator = timestamp_fastcmp;
PG_RETURN_VOID();
}
Fix up hash functions for datetime datatypes so that they don't take unwarranted liberties with int8 vs float8 values for these types. Specifically, be sure to apply either hashint8 or hashfloat8 depending on HAVE_INT64_TIMESTAMP. Per my gripe of even date.
2007-07-06 06:16:00 +02:00
Datum
timestamp_hash(PG_FUNCTION_ARGS)
{
return hashint8(fcinfo);
}
Introduce 64-bit hash functions with a 64-bit seed. This will be useful for hash partitioning, which needs a way to seed the hash functions to avoid problems such as a hash index on a hash partitioned table clumping all values into a small portion of the bucket space; it's also useful for anything that wants a 64-bit hash value rather than a 32-bit hash value. Just in case somebody wants a 64-bit hash value that is compatible with the existing 32-bit hash values, make the low 32-bits of the 64-bit hash value match the 32-bit hash value when the seed is 0. Robert Haas and Amul Sul Discussion: http://postgr.es/m/CA+Tgmoafx2yoJuhCQQOL5CocEi-w_uG4S2xT0EtgiJnPGcHW3g@mail.gmail.com
2017-09-01 04:21:21 +02:00
Datum
timestamp_hash_extended(PG_FUNCTION_ARGS)
{
return hashint8extended(fcinfo);
}
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
Add timestamp-versus-timestamptz cross-type comparison functions, flesh out the index operator classes to include these. In passing, fix erroneous volatility marking of ACL functions.
2004-03-22 02:38:18 +01:00
/*
Various minor corrections of and improvements to comments. Aleksander Alekseev
2016-03-18 14:38:59 +01:00
* Cross-type comparison functions for timestamp vs timestamptz
Add timestamp-versus-timestamptz cross-type comparison functions, flesh out the index operator classes to include these. In passing, fix erroneous volatility marking of ACL functions.
2004-03-22 02:38:18 +01:00
*/
Datum
timestamp_eq_timestamptz(PG_FUNCTION_ARGS)
{
Timestamp timestampVal = PG_GETARG_TIMESTAMP(0);
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
TimestampTz dt2 = PG_GETARG_TIMESTAMPTZ(1);
TimestampTz dt1;
Add timestamp-versus-timestamptz cross-type comparison functions, flesh out the index operator classes to include these. In passing, fix erroneous volatility marking of ACL functions.
2004-03-22 02:38:18 +01:00
dt1 = timestamp2timestamptz(timestampVal);
PG_RETURN_BOOL(timestamp_cmp_internal(dt1, dt2) == 0);
}
Datum
timestamp_ne_timestamptz(PG_FUNCTION_ARGS)
{
Timestamp timestampVal = PG_GETARG_TIMESTAMP(0);
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
TimestampTz dt2 = PG_GETARG_TIMESTAMPTZ(1);
TimestampTz dt1;
Add timestamp-versus-timestamptz cross-type comparison functions, flesh out the index operator classes to include these. In passing, fix erroneous volatility marking of ACL functions.
2004-03-22 02:38:18 +01:00
dt1 = timestamp2timestamptz(timestampVal);
PG_RETURN_BOOL(timestamp_cmp_internal(dt1, dt2) != 0);
}
Datum
timestamp_lt_timestamptz(PG_FUNCTION_ARGS)
{
Timestamp timestampVal = PG_GETARG_TIMESTAMP(0);
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
TimestampTz dt2 = PG_GETARG_TIMESTAMPTZ(1);
TimestampTz dt1;
Add timestamp-versus-timestamptz cross-type comparison functions, flesh out the index operator classes to include these. In passing, fix erroneous volatility marking of ACL functions.
2004-03-22 02:38:18 +01:00
dt1 = timestamp2timestamptz(timestampVal);
PG_RETURN_BOOL(timestamp_cmp_internal(dt1, dt2) < 0);
}
Datum
timestamp_gt_timestamptz(PG_FUNCTION_ARGS)
{
Timestamp timestampVal = PG_GETARG_TIMESTAMP(0);
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
TimestampTz dt2 = PG_GETARG_TIMESTAMPTZ(1);
TimestampTz dt1;
Add timestamp-versus-timestamptz cross-type comparison functions, flesh out the index operator classes to include these. In passing, fix erroneous volatility marking of ACL functions.
2004-03-22 02:38:18 +01:00
dt1 = timestamp2timestamptz(timestampVal);
PG_RETURN_BOOL(timestamp_cmp_internal(dt1, dt2) > 0);
}
Datum
timestamp_le_timestamptz(PG_FUNCTION_ARGS)
{
Timestamp timestampVal = PG_GETARG_TIMESTAMP(0);
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
TimestampTz dt2 = PG_GETARG_TIMESTAMPTZ(1);
TimestampTz dt1;
Add timestamp-versus-timestamptz cross-type comparison functions, flesh out the index operator classes to include these. In passing, fix erroneous volatility marking of ACL functions.
2004-03-22 02:38:18 +01:00
dt1 = timestamp2timestamptz(timestampVal);
PG_RETURN_BOOL(timestamp_cmp_internal(dt1, dt2) <= 0);
}
Datum
timestamp_ge_timestamptz(PG_FUNCTION_ARGS)
{
Timestamp timestampVal = PG_GETARG_TIMESTAMP(0);
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
TimestampTz dt2 = PG_GETARG_TIMESTAMPTZ(1);
TimestampTz dt1;
Add timestamp-versus-timestamptz cross-type comparison functions, flesh out the index operator classes to include these. In passing, fix erroneous volatility marking of ACL functions.
2004-03-22 02:38:18 +01:00
dt1 = timestamp2timestamptz(timestampVal);
PG_RETURN_BOOL(timestamp_cmp_internal(dt1, dt2) >= 0);
}
Datum
timestamp_cmp_timestamptz(PG_FUNCTION_ARGS)
{
Timestamp timestampVal = PG_GETARG_TIMESTAMP(0);
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
TimestampTz dt2 = PG_GETARG_TIMESTAMPTZ(1);
TimestampTz dt1;
Add timestamp-versus-timestamptz cross-type comparison functions, flesh out the index operator classes to include these. In passing, fix erroneous volatility marking of ACL functions.
2004-03-22 02:38:18 +01:00
dt1 = timestamp2timestamptz(timestampVal);
PG_RETURN_INT32(timestamp_cmp_internal(dt1, dt2));
}
Datum
timestamptz_eq_timestamp(PG_FUNCTION_ARGS)
{
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
TimestampTz dt1 = PG_GETARG_TIMESTAMPTZ(0);
Add timestamp-versus-timestamptz cross-type comparison functions, flesh out the index operator classes to include these. In passing, fix erroneous volatility marking of ACL functions.
2004-03-22 02:38:18 +01:00
Timestamp timestampVal = PG_GETARG_TIMESTAMP(1);
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
TimestampTz dt2;
Add timestamp-versus-timestamptz cross-type comparison functions, flesh out the index operator classes to include these. In passing, fix erroneous volatility marking of ACL functions.
2004-03-22 02:38:18 +01:00
dt2 = timestamp2timestamptz(timestampVal);
PG_RETURN_BOOL(timestamp_cmp_internal(dt1, dt2) == 0);
}
Datum
timestamptz_ne_timestamp(PG_FUNCTION_ARGS)
{
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
TimestampTz dt1 = PG_GETARG_TIMESTAMPTZ(0);
Add timestamp-versus-timestamptz cross-type comparison functions, flesh out the index operator classes to include these. In passing, fix erroneous volatility marking of ACL functions.
2004-03-22 02:38:18 +01:00
Timestamp timestampVal = PG_GETARG_TIMESTAMP(1);
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
TimestampTz dt2;
Add timestamp-versus-timestamptz cross-type comparison functions, flesh out the index operator classes to include these. In passing, fix erroneous volatility marking of ACL functions.
2004-03-22 02:38:18 +01:00
dt2 = timestamp2timestamptz(timestampVal);
PG_RETURN_BOOL(timestamp_cmp_internal(dt1, dt2) != 0);
}
Datum
timestamptz_lt_timestamp(PG_FUNCTION_ARGS)
{
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
TimestampTz dt1 = PG_GETARG_TIMESTAMPTZ(0);
Add timestamp-versus-timestamptz cross-type comparison functions, flesh out the index operator classes to include these. In passing, fix erroneous volatility marking of ACL functions.
2004-03-22 02:38:18 +01:00
Timestamp timestampVal = PG_GETARG_TIMESTAMP(1);
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
TimestampTz dt2;
Add timestamp-versus-timestamptz cross-type comparison functions, flesh out the index operator classes to include these. In passing, fix erroneous volatility marking of ACL functions.
2004-03-22 02:38:18 +01:00
dt2 = timestamp2timestamptz(timestampVal);
PG_RETURN_BOOL(timestamp_cmp_internal(dt1, dt2) < 0);
}
Datum
timestamptz_gt_timestamp(PG_FUNCTION_ARGS)
{
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
TimestampTz dt1 = PG_GETARG_TIMESTAMPTZ(0);
Add timestamp-versus-timestamptz cross-type comparison functions, flesh out the index operator classes to include these. In passing, fix erroneous volatility marking of ACL functions.
2004-03-22 02:38:18 +01:00
Timestamp timestampVal = PG_GETARG_TIMESTAMP(1);
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
TimestampTz dt2;
Add timestamp-versus-timestamptz cross-type comparison functions, flesh out the index operator classes to include these. In passing, fix erroneous volatility marking of ACL functions.
2004-03-22 02:38:18 +01:00
dt2 = timestamp2timestamptz(timestampVal);
PG_RETURN_BOOL(timestamp_cmp_internal(dt1, dt2) > 0);
}
Datum
timestamptz_le_timestamp(PG_FUNCTION_ARGS)
{
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
TimestampTz dt1 = PG_GETARG_TIMESTAMPTZ(0);
Add timestamp-versus-timestamptz cross-type comparison functions, flesh out the index operator classes to include these. In passing, fix erroneous volatility marking of ACL functions.
2004-03-22 02:38:18 +01:00
Timestamp timestampVal = PG_GETARG_TIMESTAMP(1);
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
TimestampTz dt2;
Add timestamp-versus-timestamptz cross-type comparison functions, flesh out the index operator classes to include these. In passing, fix erroneous volatility marking of ACL functions.
2004-03-22 02:38:18 +01:00
dt2 = timestamp2timestamptz(timestampVal);
PG_RETURN_BOOL(timestamp_cmp_internal(dt1, dt2) <= 0);
}
Datum
timestamptz_ge_timestamp(PG_FUNCTION_ARGS)
{
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
TimestampTz dt1 = PG_GETARG_TIMESTAMPTZ(0);
Add timestamp-versus-timestamptz cross-type comparison functions, flesh out the index operator classes to include these. In passing, fix erroneous volatility marking of ACL functions.
2004-03-22 02:38:18 +01:00
Timestamp timestampVal = PG_GETARG_TIMESTAMP(1);
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
TimestampTz dt2;
Add timestamp-versus-timestamptz cross-type comparison functions, flesh out the index operator classes to include these. In passing, fix erroneous volatility marking of ACL functions.
2004-03-22 02:38:18 +01:00
dt2 = timestamp2timestamptz(timestampVal);
PG_RETURN_BOOL(timestamp_cmp_internal(dt1, dt2) >= 0);
}
Datum
timestamptz_cmp_timestamp(PG_FUNCTION_ARGS)
{
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
TimestampTz dt1 = PG_GETARG_TIMESTAMPTZ(0);
Add timestamp-versus-timestamptz cross-type comparison functions, flesh out the index operator classes to include these. In passing, fix erroneous volatility marking of ACL functions.
2004-03-22 02:38:18 +01:00
Timestamp timestampVal = PG_GETARG_TIMESTAMP(1);
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
TimestampTz dt2;
Add timestamp-versus-timestamptz cross-type comparison functions, flesh out the index operator classes to include these. In passing, fix erroneous volatility marking of ACL functions.
2004-03-22 02:38:18 +01:00
dt2 = timestamp2timestamptz(timestampVal);
PG_RETURN_INT32(timestamp_cmp_internal(dt1, dt2));
}
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
/*
* interval_relop - is interval1 relop interval2
Fix integer-overflow problems in interval comparison. When using integer timestamps, the interval-comparison functions tried to compute the overall magnitude of an interval as an int64 number of microseconds. As reported by Frazer McLean, this overflows for intervals exceeding about 296000 years, which is bad since we nominally allow intervals many times larger than that. That results in wrong comparison results, and possibly in corrupted btree indexes for columns containing such large interval values. To fix, compute the magnitude as int128 instead. Although some compilers have native support for int128 calculations, many don't, so create our own support functions that can do 128-bit addition and multiplication if the compiler support isn't there. These support functions are designed with an eye to allowing the int128 code paths in numeric.c to be rewritten for use on all platforms, although this patch doesn't do that, or even provide all the int128 primitives that will be needed for it. Back-patch as far as 9.4. Earlier releases did not guard against overflow of interval values at all (commit 146604ec4 fixed that), so it seems not very exciting to worry about overly-large intervals for them. Before 9.6, we did not assume that unreferenced "static inline" functions would not draw compiler warnings, so omit functions not directly referenced by timestamp.c, the only present consumer of int128.h. (We could have omitted these functions in HEAD too, but since they were written and debugged on the way to the present patch, and they look likely to be needed by numeric.c, let's keep them in HEAD.) I did not bother to try to prevent such warnings in a --disable-integer-datetimes build, though. Before 9.5, configure will never define HAVE_INT128, so the part of int128.h that exploits a native int128 implementation is dead code in the 9.4 branch. I didn't bother to remove it, thinking that keeping the file looking similar in different branches is more useful. In HEAD only, add a simple test harness for int128.h in src/tools/. In back branches, this does not change the float-timestamps code path. That's not subject to the same kind of overflow risk, since it computes the interval magnitude as float8. (No doubt, when this code was originally written, overflow was disregarded for exactly that reason.) There is a precision hazard instead :-(, but we'll avert our eyes from that question, since no complaints have been reported and that code's deprecated anyway. Kyotaro Horiguchi and Tom Lane Discussion: https://postgr.es/m/1490104629.422698.918452336.26FA96B7@webmail.messagingengine.com
2017-04-06 05:51:27 +02:00
*
* Interval comparison is based on converting interval values to a linear
* representation expressed in the units of the time field (microseconds,
* in the case of integer timestamps) with days assumed to be always 24 hours
* and months assumed to be always 30 days. To avoid overflow, we need a
* wider-than-int64 datatype for the linear representation, so use INT128.
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
*/
Fix integer-overflow problems in interval comparison. When using integer timestamps, the interval-comparison functions tried to compute the overall magnitude of an interval as an int64 number of microseconds. As reported by Frazer McLean, this overflows for intervals exceeding about 296000 years, which is bad since we nominally allow intervals many times larger than that. That results in wrong comparison results, and possibly in corrupted btree indexes for columns containing such large interval values. To fix, compute the magnitude as int128 instead. Although some compilers have native support for int128 calculations, many don't, so create our own support functions that can do 128-bit addition and multiplication if the compiler support isn't there. These support functions are designed with an eye to allowing the int128 code paths in numeric.c to be rewritten for use on all platforms, although this patch doesn't do that, or even provide all the int128 primitives that will be needed for it. Back-patch as far as 9.4. Earlier releases did not guard against overflow of interval values at all (commit 146604ec4 fixed that), so it seems not very exciting to worry about overly-large intervals for them. Before 9.6, we did not assume that unreferenced "static inline" functions would not draw compiler warnings, so omit functions not directly referenced by timestamp.c, the only present consumer of int128.h. (We could have omitted these functions in HEAD too, but since they were written and debugged on the way to the present patch, and they look likely to be needed by numeric.c, let's keep them in HEAD.) I did not bother to try to prevent such warnings in a --disable-integer-datetimes build, though. Before 9.5, configure will never define HAVE_INT128, so the part of int128.h that exploits a native int128 implementation is dead code in the 9.4 branch. I didn't bother to remove it, thinking that keeping the file looking similar in different branches is more useful. In HEAD only, add a simple test harness for int128.h in src/tools/. In back branches, this does not change the float-timestamps code path. That's not subject to the same kind of overflow risk, since it computes the interval magnitude as float8. (No doubt, when this code was originally written, overflow was disregarded for exactly that reason.) There is a precision hazard instead :-(, but we'll avert our eyes from that question, since no complaints have been reported and that code's deprecated anyway. Kyotaro Horiguchi and Tom Lane Discussion: https://postgr.es/m/1490104629.422698.918452336.26FA96B7@webmail.messagingengine.com
2017-04-06 05:51:27 +02:00
static inline INT128
Rewrite interval_hash() so that the hashcodes are equal for values that interval_eq() considers equal. I'm not sure how that fundamental requirement escaped us through multiple revisions of this hash function, but there it is; it's been wrong since interval_hash was first written for PG 7.1. Per bug #4748 from Roman Kononov. Backpatch to all supported releases. This patch changes the contents of hash indexes for interval columns. That's no particular problem for PG 8.4, since we've broken on-disk compatibility of hash indexes already; but it will require a migration warning note in the next minor releases of all existing branches: "if you have any hash indexes on columns of type interval, REINDEX them after updating".
2009-04-04 06:53:25 +02:00
interval_cmp_value(const Interval *interval)
Ensure that btree sort ordering functions and boolean comparison operators give consistent results for all datatypes. Types float4, float8, and numeric were broken for NaN values; abstime, timestamp, and interval were broken for INVALID values; timetz was just plain broken (some possible pairs of values were neither < nor = nor >). Also clean up text, bpchar, varchar, and bit/varbit to eliminate duplicate code and thereby reduce the probability of similar inconsistencies arising in the future.
2001-05-03 21:00:37 +02:00
{
Fix integer-overflow problems in interval comparison. When using integer timestamps, the interval-comparison functions tried to compute the overall magnitude of an interval as an int64 number of microseconds. As reported by Frazer McLean, this overflows for intervals exceeding about 296000 years, which is bad since we nominally allow intervals many times larger than that. That results in wrong comparison results, and possibly in corrupted btree indexes for columns containing such large interval values. To fix, compute the magnitude as int128 instead. Although some compilers have native support for int128 calculations, many don't, so create our own support functions that can do 128-bit addition and multiplication if the compiler support isn't there. These support functions are designed with an eye to allowing the int128 code paths in numeric.c to be rewritten for use on all platforms, although this patch doesn't do that, or even provide all the int128 primitives that will be needed for it. Back-patch as far as 9.4. Earlier releases did not guard against overflow of interval values at all (commit 146604ec4 fixed that), so it seems not very exciting to worry about overly-large intervals for them. Before 9.6, we did not assume that unreferenced "static inline" functions would not draw compiler warnings, so omit functions not directly referenced by timestamp.c, the only present consumer of int128.h. (We could have omitted these functions in HEAD too, but since they were written and debugged on the way to the present patch, and they look likely to be needed by numeric.c, let's keep them in HEAD.) I did not bother to try to prevent such warnings in a --disable-integer-datetimes build, though. Before 9.5, configure will never define HAVE_INT128, so the part of int128.h that exploits a native int128 implementation is dead code in the 9.4 branch. I didn't bother to remove it, thinking that keeping the file looking similar in different branches is more useful. In HEAD only, add a simple test harness for int128.h in src/tools/. In back branches, this does not change the float-timestamps code path. That's not subject to the same kind of overflow risk, since it computes the interval magnitude as float8. (No doubt, when this code was originally written, overflow was disregarded for exactly that reason.) There is a precision hazard instead :-(, but we'll avert our eyes from that question, since no complaints have been reported and that code's deprecated anyway. Kyotaro Horiguchi and Tom Lane Discussion: https://postgr.es/m/1490104629.422698.918452336.26FA96B7@webmail.messagingengine.com
2017-04-06 05:51:27 +02:00
INT128 span;
int64 dayfraction;
int64 days;
/*
* Separate time field into days and dayfraction, then add the month and
* day fields to the days part. We cannot overflow int64 days here.
*/
dayfraction = interval->time % USECS_PER_DAY;
days = interval->time / USECS_PER_DAY;
days += interval->month * INT64CONST(30);
days += interval->day;
Ensure that btree sort ordering functions and boolean comparison operators give consistent results for all datatypes. Types float4, float8, and numeric were broken for NaN values; abstime, timestamp, and interval were broken for INVALID values; timetz was just plain broken (some possible pairs of values were neither < nor = nor >). Also clean up text, bpchar, varchar, and bit/varbit to eliminate duplicate code and thereby reduce the probability of similar inconsistencies arising in the future.
2001-05-03 21:00:37 +02:00
Fix integer-overflow problems in interval comparison. When using integer timestamps, the interval-comparison functions tried to compute the overall magnitude of an interval as an int64 number of microseconds. As reported by Frazer McLean, this overflows for intervals exceeding about 296000 years, which is bad since we nominally allow intervals many times larger than that. That results in wrong comparison results, and possibly in corrupted btree indexes for columns containing such large interval values. To fix, compute the magnitude as int128 instead. Although some compilers have native support for int128 calculations, many don't, so create our own support functions that can do 128-bit addition and multiplication if the compiler support isn't there. These support functions are designed with an eye to allowing the int128 code paths in numeric.c to be rewritten for use on all platforms, although this patch doesn't do that, or even provide all the int128 primitives that will be needed for it. Back-patch as far as 9.4. Earlier releases did not guard against overflow of interval values at all (commit 146604ec4 fixed that), so it seems not very exciting to worry about overly-large intervals for them. Before 9.6, we did not assume that unreferenced "static inline" functions would not draw compiler warnings, so omit functions not directly referenced by timestamp.c, the only present consumer of int128.h. (We could have omitted these functions in HEAD too, but since they were written and debugged on the way to the present patch, and they look likely to be needed by numeric.c, let's keep them in HEAD.) I did not bother to try to prevent such warnings in a --disable-integer-datetimes build, though. Before 9.5, configure will never define HAVE_INT128, so the part of int128.h that exploits a native int128 implementation is dead code in the 9.4 branch. I didn't bother to remove it, thinking that keeping the file looking similar in different branches is more useful. In HEAD only, add a simple test harness for int128.h in src/tools/. In back branches, this does not change the float-timestamps code path. That's not subject to the same kind of overflow risk, since it computes the interval magnitude as float8. (No doubt, when this code was originally written, overflow was disregarded for exactly that reason.) There is a precision hazard instead :-(, but we'll avert our eyes from that question, since no complaints have been reported and that code's deprecated anyway. Kyotaro Horiguchi and Tom Lane Discussion: https://postgr.es/m/1490104629.422698.918452336.26FA96B7@webmail.messagingengine.com
2017-04-06 05:51:27 +02:00
/* Widen dayfraction to 128 bits */
span = int64_to_int128(dayfraction);
/* Scale up days to microseconds, forming a 128-bit product */
int128_add_int64_mul_int64(&span, days, USECS_PER_DAY);
Ensure that btree sort ordering functions and boolean comparison operators give consistent results for all datatypes. Types float4, float8, and numeric were broken for NaN values; abstime, timestamp, and interval were broken for INVALID values; timetz was just plain broken (some possible pairs of values were neither < nor = nor >). Also clean up text, bpchar, varchar, and bit/varbit to eliminate duplicate code and thereby reduce the probability of similar inconsistencies arising in the future.
2001-05-03 21:00:37 +02:00
Rewrite interval_hash() so that the hashcodes are equal for values that interval_eq() considers equal. I'm not sure how that fundamental requirement escaped us through multiple revisions of this hash function, but there it is; it's been wrong since interval_hash was first written for PG 7.1. Per bug #4748 from Roman Kononov. Backpatch to all supported releases. This patch changes the contents of hash indexes for interval columns. That's no particular problem for PG 8.4, since we've broken on-disk compatibility of hash indexes already; but it will require a migration warning note in the next minor releases of all existing branches: "if you have any hash indexes on columns of type interval, REINDEX them after updating".
2009-04-04 06:53:25 +02:00
return span;
}
static int
interval_cmp_internal(Interval *interval1, Interval *interval2)
{
Fix integer-overflow problems in interval comparison. When using integer timestamps, the interval-comparison functions tried to compute the overall magnitude of an interval as an int64 number of microseconds. As reported by Frazer McLean, this overflows for intervals exceeding about 296000 years, which is bad since we nominally allow intervals many times larger than that. That results in wrong comparison results, and possibly in corrupted btree indexes for columns containing such large interval values. To fix, compute the magnitude as int128 instead. Although some compilers have native support for int128 calculations, many don't, so create our own support functions that can do 128-bit addition and multiplication if the compiler support isn't there. These support functions are designed with an eye to allowing the int128 code paths in numeric.c to be rewritten for use on all platforms, although this patch doesn't do that, or even provide all the int128 primitives that will be needed for it. Back-patch as far as 9.4. Earlier releases did not guard against overflow of interval values at all (commit 146604ec4 fixed that), so it seems not very exciting to worry about overly-large intervals for them. Before 9.6, we did not assume that unreferenced "static inline" functions would not draw compiler warnings, so omit functions not directly referenced by timestamp.c, the only present consumer of int128.h. (We could have omitted these functions in HEAD too, but since they were written and debugged on the way to the present patch, and they look likely to be needed by numeric.c, let's keep them in HEAD.) I did not bother to try to prevent such warnings in a --disable-integer-datetimes build, though. Before 9.5, configure will never define HAVE_INT128, so the part of int128.h that exploits a native int128 implementation is dead code in the 9.4 branch. I didn't bother to remove it, thinking that keeping the file looking similar in different branches is more useful. In HEAD only, add a simple test harness for int128.h in src/tools/. In back branches, this does not change the float-timestamps code path. That's not subject to the same kind of overflow risk, since it computes the interval magnitude as float8. (No doubt, when this code was originally written, overflow was disregarded for exactly that reason.) There is a precision hazard instead :-(, but we'll avert our eyes from that question, since no complaints have been reported and that code's deprecated anyway. Kyotaro Horiguchi and Tom Lane Discussion: https://postgr.es/m/1490104629.422698.918452336.26FA96B7@webmail.messagingengine.com
2017-04-06 05:51:27 +02:00
INT128 span1 = interval_cmp_value(interval1);
INT128 span2 = interval_cmp_value(interval2);
Rewrite interval_hash() so that the hashcodes are equal for values that interval_eq() considers equal. I'm not sure how that fundamental requirement escaped us through multiple revisions of this hash function, but there it is; it's been wrong since interval_hash was first written for PG 7.1. Per bug #4748 from Roman Kononov. Backpatch to all supported releases. This patch changes the contents of hash indexes for interval columns. That's no particular problem for PG 8.4, since we've broken on-disk compatibility of hash indexes already; but it will require a migration warning note in the next minor releases of all existing branches: "if you have any hash indexes on columns of type interval, REINDEX them after updating".
2009-04-04 06:53:25 +02:00
Fix integer-overflow problems in interval comparison. When using integer timestamps, the interval-comparison functions tried to compute the overall magnitude of an interval as an int64 number of microseconds. As reported by Frazer McLean, this overflows for intervals exceeding about 296000 years, which is bad since we nominally allow intervals many times larger than that. That results in wrong comparison results, and possibly in corrupted btree indexes for columns containing such large interval values. To fix, compute the magnitude as int128 instead. Although some compilers have native support for int128 calculations, many don't, so create our own support functions that can do 128-bit addition and multiplication if the compiler support isn't there. These support functions are designed with an eye to allowing the int128 code paths in numeric.c to be rewritten for use on all platforms, although this patch doesn't do that, or even provide all the int128 primitives that will be needed for it. Back-patch as far as 9.4. Earlier releases did not guard against overflow of interval values at all (commit 146604ec4 fixed that), so it seems not very exciting to worry about overly-large intervals for them. Before 9.6, we did not assume that unreferenced "static inline" functions would not draw compiler warnings, so omit functions not directly referenced by timestamp.c, the only present consumer of int128.h. (We could have omitted these functions in HEAD too, but since they were written and debugged on the way to the present patch, and they look likely to be needed by numeric.c, let's keep them in HEAD.) I did not bother to try to prevent such warnings in a --disable-integer-datetimes build, though. Before 9.5, configure will never define HAVE_INT128, so the part of int128.h that exploits a native int128 implementation is dead code in the 9.4 branch. I didn't bother to remove it, thinking that keeping the file looking similar in different branches is more useful. In HEAD only, add a simple test harness for int128.h in src/tools/. In back branches, this does not change the float-timestamps code path. That's not subject to the same kind of overflow risk, since it computes the interval magnitude as float8. (No doubt, when this code was originally written, overflow was disregarded for exactly that reason.) There is a precision hazard instead :-(, but we'll avert our eyes from that question, since no complaints have been reported and that code's deprecated anyway. Kyotaro Horiguchi and Tom Lane Discussion: https://postgr.es/m/1490104629.422698.918452336.26FA96B7@webmail.messagingengine.com
2017-04-06 05:51:27 +02:00
return int128_compare(span1, span2);
Ensure that btree sort ordering functions and boolean comparison operators give consistent results for all datatypes. Types float4, float8, and numeric were broken for NaN values; abstime, timestamp, and interval were broken for INVALID values; timetz was just plain broken (some possible pairs of values were neither < nor = nor >). Also clean up text, bpchar, varchar, and bit/varbit to eliminate duplicate code and thereby reduce the probability of similar inconsistencies arising in the future.
2001-05-03 21:00:37 +02:00
}
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
Datum
interval_eq(PG_FUNCTION_ARGS)
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
{
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
Interval *interval1 = PG_GETARG_INTERVAL_P(0);
Interval *interval2 = PG_GETARG_INTERVAL_P(1);
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
Ensure that btree sort ordering functions and boolean comparison operators give consistent results for all datatypes. Types float4, float8, and numeric were broken for NaN values; abstime, timestamp, and interval were broken for INVALID values; timetz was just plain broken (some possible pairs of values were neither < nor = nor >). Also clean up text, bpchar, varchar, and bit/varbit to eliminate duplicate code and thereby reduce the probability of similar inconsistencies arising in the future.
2001-05-03 21:00:37 +02:00
PG_RETURN_BOOL(interval_cmp_internal(interval1, interval2) == 0);
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
}
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
Datum
interval_ne(PG_FUNCTION_ARGS)
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
{
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
Interval *interval1 = PG_GETARG_INTERVAL_P(0);
Interval *interval2 = PG_GETARG_INTERVAL_P(1);
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
Ensure that btree sort ordering functions and boolean comparison operators give consistent results for all datatypes. Types float4, float8, and numeric were broken for NaN values; abstime, timestamp, and interval were broken for INVALID values; timetz was just plain broken (some possible pairs of values were neither < nor = nor >). Also clean up text, bpchar, varchar, and bit/varbit to eliminate duplicate code and thereby reduce the probability of similar inconsistencies arising in the future.
2001-05-03 21:00:37 +02:00
PG_RETURN_BOOL(interval_cmp_internal(interval1, interval2) != 0);
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
}
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
Datum
interval_lt(PG_FUNCTION_ARGS)
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
{
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
Interval *interval1 = PG_GETARG_INTERVAL_P(0);
Interval *interval2 = PG_GETARG_INTERVAL_P(1);
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
Ensure that btree sort ordering functions and boolean comparison operators give consistent results for all datatypes. Types float4, float8, and numeric were broken for NaN values; abstime, timestamp, and interval were broken for INVALID values; timetz was just plain broken (some possible pairs of values were neither < nor = nor >). Also clean up text, bpchar, varchar, and bit/varbit to eliminate duplicate code and thereby reduce the probability of similar inconsistencies arising in the future.
2001-05-03 21:00:37 +02:00
PG_RETURN_BOOL(interval_cmp_internal(interval1, interval2) < 0);
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
}
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
Datum
interval_gt(PG_FUNCTION_ARGS)
Shorten routine names to <= 16 characters to fit in pg_proc table.
1997-10-25 07:22:58 +02:00
{
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
Interval *interval1 = PG_GETARG_INTERVAL_P(0);
Interval *interval2 = PG_GETARG_INTERVAL_P(1);
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
Ensure that btree sort ordering functions and boolean comparison operators give consistent results for all datatypes. Types float4, float8, and numeric were broken for NaN values; abstime, timestamp, and interval were broken for INVALID values; timetz was just plain broken (some possible pairs of values were neither < nor = nor >). Also clean up text, bpchar, varchar, and bit/varbit to eliminate duplicate code and thereby reduce the probability of similar inconsistencies arising in the future.
2001-05-03 21:00:37 +02:00
PG_RETURN_BOOL(interval_cmp_internal(interval1, interval2) > 0);
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
}
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
Datum
interval_le(PG_FUNCTION_ARGS)
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
{
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
Interval *interval1 = PG_GETARG_INTERVAL_P(0);
Interval *interval2 = PG_GETARG_INTERVAL_P(1);
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
Ensure that btree sort ordering functions and boolean comparison operators give consistent results for all datatypes. Types float4, float8, and numeric were broken for NaN values; abstime, timestamp, and interval were broken for INVALID values; timetz was just plain broken (some possible pairs of values were neither < nor = nor >). Also clean up text, bpchar, varchar, and bit/varbit to eliminate duplicate code and thereby reduce the probability of similar inconsistencies arising in the future.
2001-05-03 21:00:37 +02:00
PG_RETURN_BOOL(interval_cmp_internal(interval1, interval2) <= 0);
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
}
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
Datum
interval_ge(PG_FUNCTION_ARGS)
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
{
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
Interval *interval1 = PG_GETARG_INTERVAL_P(0);
Interval *interval2 = PG_GETARG_INTERVAL_P(1);
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
Ensure that btree sort ordering functions and boolean comparison operators give consistent results for all datatypes. Types float4, float8, and numeric were broken for NaN values; abstime, timestamp, and interval were broken for INVALID values; timetz was just plain broken (some possible pairs of values were neither < nor = nor >). Also clean up text, bpchar, varchar, and bit/varbit to eliminate duplicate code and thereby reduce the probability of similar inconsistencies arising in the future.
2001-05-03 21:00:37 +02:00
PG_RETURN_BOOL(interval_cmp_internal(interval1, interval2) >= 0);
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
}
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
Datum
interval_cmp(PG_FUNCTION_ARGS)
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
{
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
Interval *interval1 = PG_GETARG_INTERVAL_P(0);
Interval *interval2 = PG_GETARG_INTERVAL_P(1);
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
Ensure that btree sort ordering functions and boolean comparison operators give consistent results for all datatypes. Types float4, float8, and numeric were broken for NaN values; abstime, timestamp, and interval were broken for INVALID values; timetz was just plain broken (some possible pairs of values were neither < nor = nor >). Also clean up text, bpchar, varchar, and bit/varbit to eliminate duplicate code and thereby reduce the probability of similar inconsistencies arising in the future.
2001-05-03 21:00:37 +02:00
PG_RETURN_INT32(interval_cmp_internal(interval1, interval2));
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
}
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
Rewrite interval_hash() so that the hashcodes are equal for values that interval_eq() considers equal. I'm not sure how that fundamental requirement escaped us through multiple revisions of this hash function, but there it is; it's been wrong since interval_hash was first written for PG 7.1. Per bug #4748 from Roman Kononov. Backpatch to all supported releases. This patch changes the contents of hash indexes for interval columns. That's no particular problem for PG 8.4, since we've broken on-disk compatibility of hash indexes already; but it will require a migration warning note in the next minor releases of all existing branches: "if you have any hash indexes on columns of type interval, REINDEX them after updating".
2009-04-04 06:53:25 +02:00
/*
* Hashing for intervals
*
* We must produce equal hashvals for values that interval_cmp_internal()
* considers equal. So, compute the net span the same way it does,
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
* and then hash that.
Rewrite interval_hash() so that the hashcodes are equal for values that interval_eq() considers equal. I'm not sure how that fundamental requirement escaped us through multiple revisions of this hash function, but there it is; it's been wrong since interval_hash was first written for PG 7.1. Per bug #4748 from Roman Kononov. Backpatch to all supported releases. This patch changes the contents of hash indexes for interval columns. That's no particular problem for PG 8.4, since we've broken on-disk compatibility of hash indexes already; but it will require a migration warning note in the next minor releases of all existing branches: "if you have any hash indexes on columns of type interval, REINDEX them after updating".
2009-04-04 06:53:25 +02:00
*/
Clean up bogosities in pg_opclass, pg_amop, pg_amproc. There are amproc entries now for int8 and network hash indexes. int24_ops and int42_ops are gone. pg_opclass no longer contains multiple entries claiming to be the default opclass for the same datatype. opr_sanity regress test extended to catch errors like these in the future.
2000-06-19 05:55:01 +02:00
Datum
interval_hash(PG_FUNCTION_ARGS)
{
Rewrite interval_hash() so that the hashcodes are equal for values that interval_eq() considers equal. I'm not sure how that fundamental requirement escaped us through multiple revisions of this hash function, but there it is; it's been wrong since interval_hash was first written for PG 7.1. Per bug #4748 from Roman Kononov. Backpatch to all supported releases. This patch changes the contents of hash indexes for interval columns. That's no particular problem for PG 8.4, since we've broken on-disk compatibility of hash indexes already; but it will require a migration warning note in the next minor releases of all existing branches: "if you have any hash indexes on columns of type interval, REINDEX them after updating".
2009-04-04 06:53:25 +02:00
Interval *interval = PG_GETARG_INTERVAL_P(0);
Fix integer-overflow problems in interval comparison. When using integer timestamps, the interval-comparison functions tried to compute the overall magnitude of an interval as an int64 number of microseconds. As reported by Frazer McLean, this overflows for intervals exceeding about 296000 years, which is bad since we nominally allow intervals many times larger than that. That results in wrong comparison results, and possibly in corrupted btree indexes for columns containing such large interval values. To fix, compute the magnitude as int128 instead. Although some compilers have native support for int128 calculations, many don't, so create our own support functions that can do 128-bit addition and multiplication if the compiler support isn't there. These support functions are designed with an eye to allowing the int128 code paths in numeric.c to be rewritten for use on all platforms, although this patch doesn't do that, or even provide all the int128 primitives that will be needed for it. Back-patch as far as 9.4. Earlier releases did not guard against overflow of interval values at all (commit 146604ec4 fixed that), so it seems not very exciting to worry about overly-large intervals for them. Before 9.6, we did not assume that unreferenced "static inline" functions would not draw compiler warnings, so omit functions not directly referenced by timestamp.c, the only present consumer of int128.h. (We could have omitted these functions in HEAD too, but since they were written and debugged on the way to the present patch, and they look likely to be needed by numeric.c, let's keep them in HEAD.) I did not bother to try to prevent such warnings in a --disable-integer-datetimes build, though. Before 9.5, configure will never define HAVE_INT128, so the part of int128.h that exploits a native int128 implementation is dead code in the 9.4 branch. I didn't bother to remove it, thinking that keeping the file looking similar in different branches is more useful. In HEAD only, add a simple test harness for int128.h in src/tools/. In back branches, this does not change the float-timestamps code path. That's not subject to the same kind of overflow risk, since it computes the interval magnitude as float8. (No doubt, when this code was originally written, overflow was disregarded for exactly that reason.) There is a precision hazard instead :-(, but we'll avert our eyes from that question, since no complaints have been reported and that code's deprecated anyway. Kyotaro Horiguchi and Tom Lane Discussion: https://postgr.es/m/1490104629.422698.918452336.26FA96B7@webmail.messagingengine.com
2017-04-06 05:51:27 +02:00
INT128 span = interval_cmp_value(interval);
int64 span64;
/*
* Use only the least significant 64 bits for hashing. The upper 64 bits
* seldom add any useful information, and besides we must do it like this
* for compatibility with hashes calculated before use of INT128 was
* introduced.
*/
span64 = int128_to_int64(span);
Clean up bogosities in pg_opclass, pg_amop, pg_amproc. There are amproc entries now for int8 and network hash indexes. int24_ops and int42_ops are gone. pg_opclass no longer contains multiple entries claiming to be the default opclass for the same datatype. opr_sanity regress test extended to catch errors like these in the future.
2000-06-19 05:55:01 +02:00
Fix integer-overflow problems in interval comparison. When using integer timestamps, the interval-comparison functions tried to compute the overall magnitude of an interval as an int64 number of microseconds. As reported by Frazer McLean, this overflows for intervals exceeding about 296000 years, which is bad since we nominally allow intervals many times larger than that. That results in wrong comparison results, and possibly in corrupted btree indexes for columns containing such large interval values. To fix, compute the magnitude as int128 instead. Although some compilers have native support for int128 calculations, many don't, so create our own support functions that can do 128-bit addition and multiplication if the compiler support isn't there. These support functions are designed with an eye to allowing the int128 code paths in numeric.c to be rewritten for use on all platforms, although this patch doesn't do that, or even provide all the int128 primitives that will be needed for it. Back-patch as far as 9.4. Earlier releases did not guard against overflow of interval values at all (commit 146604ec4 fixed that), so it seems not very exciting to worry about overly-large intervals for them. Before 9.6, we did not assume that unreferenced "static inline" functions would not draw compiler warnings, so omit functions not directly referenced by timestamp.c, the only present consumer of int128.h. (We could have omitted these functions in HEAD too, but since they were written and debugged on the way to the present patch, and they look likely to be needed by numeric.c, let's keep them in HEAD.) I did not bother to try to prevent such warnings in a --disable-integer-datetimes build, though. Before 9.5, configure will never define HAVE_INT128, so the part of int128.h that exploits a native int128 implementation is dead code in the 9.4 branch. I didn't bother to remove it, thinking that keeping the file looking similar in different branches is more useful. In HEAD only, add a simple test harness for int128.h in src/tools/. In back branches, this does not change the float-timestamps code path. That's not subject to the same kind of overflow risk, since it computes the interval magnitude as float8. (No doubt, when this code was originally written, overflow was disregarded for exactly that reason.) There is a precision hazard instead :-(, but we'll avert our eyes from that question, since no complaints have been reported and that code's deprecated anyway. Kyotaro Horiguchi and Tom Lane Discussion: https://postgr.es/m/1490104629.422698.918452336.26FA96B7@webmail.messagingengine.com
2017-04-06 05:51:27 +02:00
return DirectFunctionCall1(hashint8, Int64GetDatumFast(span64));
Clean up bogosities in pg_opclass, pg_amop, pg_amproc. There are amproc entries now for int8 and network hash indexes. int24_ops and int42_ops are gone. pg_opclass no longer contains multiple entries claiming to be the default opclass for the same datatype. opr_sanity regress test extended to catch errors like these in the future.
2000-06-19 05:55:01 +02:00
}
Introduce 64-bit hash functions with a 64-bit seed. This will be useful for hash partitioning, which needs a way to seed the hash functions to avoid problems such as a hash index on a hash partitioned table clumping all values into a small portion of the bucket space; it's also useful for anything that wants a 64-bit hash value rather than a 32-bit hash value. Just in case somebody wants a 64-bit hash value that is compatible with the existing 32-bit hash values, make the low 32-bits of the 64-bit hash value match the 32-bit hash value when the seed is 0. Robert Haas and Amul Sul Discussion: http://postgr.es/m/CA+Tgmoafx2yoJuhCQQOL5CocEi-w_uG4S2xT0EtgiJnPGcHW3g@mail.gmail.com
2017-09-01 04:21:21 +02:00
Datum
interval_hash_extended(PG_FUNCTION_ARGS)
{
Interval *interval = PG_GETARG_INTERVAL_P(0);
INT128 span = interval_cmp_value(interval);
int64 span64;
/* Same approach as interval_hash */
span64 = int128_to_int64(span);
return DirectFunctionCall2(hashint8extended, Int64GetDatumFast(span64),
PG_GETARG_DATUM(1));
}
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
/* overlaps_timestamp() --- implements the SQL OVERLAPS operator.
Make OVERLAPS operators conform to SQL92 spec regarding NULL handling. As I read it, the spec requires a non-null result in some cases where one of the inputs is NULL: specifically, if the other endpoint of that interval is between the endpoints of the other interval, then the result is known TRUE despite the missing endpoint. The spec could've been a lot simpler if they did not intend this behavior. I did not force an initdb for this change, but if you don't do one you'll still see the old strict-function behavior.
2000-12-07 19:38:59 +01:00
*
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
* Algorithm is per SQL spec. This is much harder than you'd think
Make OVERLAPS operators conform to SQL92 spec regarding NULL handling. As I read it, the spec requires a non-null result in some cases where one of the inputs is NULL: specifically, if the other endpoint of that interval is between the endpoints of the other interval, then the result is known TRUE despite the missing endpoint. The spec could've been a lot simpler if they did not intend this behavior. I did not force an initdb for this change, but if you don't do one you'll still see the old strict-function behavior.
2000-12-07 19:38:59 +01:00
* because the spec requires us to deliver a non-null answer in some cases
* where some of the inputs are null.
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
*/
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
Datum
overlaps_timestamp(PG_FUNCTION_ARGS)
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. Make it all clean.
2001-03-22 05:01:46 +01:00
/*
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* The arguments are Timestamps, but we leave them as generic Datums to
* avoid unnecessary conversions between value and reference forms --- not
* to mention possible dereferences of null pointers.
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
*/
Datum ts1 = PG_GETARG_DATUM(0);
Datum te1 = PG_GETARG_DATUM(1);
Datum ts2 = PG_GETARG_DATUM(2);
Datum te2 = PG_GETARG_DATUM(3);
Make OVERLAPS operators conform to SQL92 spec regarding NULL handling. As I read it, the spec requires a non-null result in some cases where one of the inputs is NULL: specifically, if the other endpoint of that interval is between the endpoints of the other interval, then the result is known TRUE despite the missing endpoint. The spec could've been a lot simpler if they did not intend this behavior. I did not force an initdb for this change, but if you don't do one you'll still see the old strict-function behavior.
2000-12-07 19:38:59 +01:00
bool ts1IsNull = PG_ARGISNULL(0);
bool te1IsNull = PG_ARGISNULL(1);
bool ts2IsNull = PG_ARGISNULL(2);
bool te2IsNull = PG_ARGISNULL(3);
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
#define TIMESTAMP_GT(t1,t2) \
DatumGetBool(DirectFunctionCall2(timestamp_gt,t1,t2))
#define TIMESTAMP_LT(t1,t2) \
DatumGetBool(DirectFunctionCall2(timestamp_lt,t1,t2))
Make OVERLAPS operators conform to SQL92 spec regarding NULL handling. As I read it, the spec requires a non-null result in some cases where one of the inputs is NULL: specifically, if the other endpoint of that interval is between the endpoints of the other interval, then the result is known TRUE despite the missing endpoint. The spec could've been a lot simpler if they did not intend this behavior. I did not force an initdb for this change, but if you don't do one you'll still see the old strict-function behavior.
2000-12-07 19:38:59 +01:00
/*
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* If both endpoints of interval 1 are null, the result is null (unknown).
* If just one endpoint is null, take ts1 as the non-null one. Otherwise,
* take ts1 as the lesser endpoint.
Make OVERLAPS operators conform to SQL92 spec regarding NULL handling. As I read it, the spec requires a non-null result in some cases where one of the inputs is NULL: specifically, if the other endpoint of that interval is between the endpoints of the other interval, then the result is known TRUE despite the missing endpoint. The spec could've been a lot simpler if they did not intend this behavior. I did not force an initdb for this change, but if you don't do one you'll still see the old strict-function behavior.
2000-12-07 19:38:59 +01:00
*/
if (ts1IsNull)
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
{
Make OVERLAPS operators conform to SQL92 spec regarding NULL handling. As I read it, the spec requires a non-null result in some cases where one of the inputs is NULL: specifically, if the other endpoint of that interval is between the endpoints of the other interval, then the result is known TRUE despite the missing endpoint. The spec could've been a lot simpler if they did not intend this behavior. I did not force an initdb for this change, but if you don't do one you'll still see the old strict-function behavior.
2000-12-07 19:38:59 +01:00
if (te1IsNull)
PG_RETURN_NULL();
/* swap null for non-null */
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
ts1 = te1;
Make OVERLAPS operators conform to SQL92 spec regarding NULL handling. As I read it, the spec requires a non-null result in some cases where one of the inputs is NULL: specifically, if the other endpoint of that interval is between the endpoints of the other interval, then the result is known TRUE despite the missing endpoint. The spec could've been a lot simpler if they did not intend this behavior. I did not force an initdb for this change, but if you don't do one you'll still see the old strict-function behavior.
2000-12-07 19:38:59 +01:00
te1IsNull = true;
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
}
Make OVERLAPS operators conform to SQL92 spec regarding NULL handling. As I read it, the spec requires a non-null result in some cases where one of the inputs is NULL: specifically, if the other endpoint of that interval is between the endpoints of the other interval, then the result is known TRUE despite the missing endpoint. The spec could've been a lot simpler if they did not intend this behavior. I did not force an initdb for this change, but if you don't do one you'll still see the old strict-function behavior.
2000-12-07 19:38:59 +01:00
else if (!te1IsNull)
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
{
Make OVERLAPS operators conform to SQL92 spec regarding NULL handling. As I read it, the spec requires a non-null result in some cases where one of the inputs is NULL: specifically, if the other endpoint of that interval is between the endpoints of the other interval, then the result is known TRUE despite the missing endpoint. The spec could've been a lot simpler if they did not intend this behavior. I did not force an initdb for this change, but if you don't do one you'll still see the old strict-function behavior.
2000-12-07 19:38:59 +01:00
if (TIMESTAMP_GT(ts1, te1))
{
pgindent run. Make it all clean.
2001-03-22 05:01:46 +01:00
Datum tt = ts1;
Ye-old pgindent run. Same 4-space tabs.
2000-04-12 19:17:23 +02:00
Make OVERLAPS operators conform to SQL92 spec regarding NULL handling. As I read it, the spec requires a non-null result in some cases where one of the inputs is NULL: specifically, if the other endpoint of that interval is between the endpoints of the other interval, then the result is known TRUE despite the missing endpoint. The spec could've been a lot simpler if they did not intend this behavior. I did not force an initdb for this change, but if you don't do one you'll still see the old strict-function behavior.
2000-12-07 19:38:59 +01:00
ts1 = te1;
te1 = tt;
}
}
/* Likewise for interval 2. */
if (ts2IsNull)
{
if (te2IsNull)
PG_RETURN_NULL();
/* swap null for non-null */
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
ts2 = te2;
Make OVERLAPS operators conform to SQL92 spec regarding NULL handling. As I read it, the spec requires a non-null result in some cases where one of the inputs is NULL: specifically, if the other endpoint of that interval is between the endpoints of the other interval, then the result is known TRUE despite the missing endpoint. The spec could've been a lot simpler if they did not intend this behavior. I did not force an initdb for this change, but if you don't do one you'll still see the old strict-function behavior.
2000-12-07 19:38:59 +01:00
te2IsNull = true;
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
}
Make OVERLAPS operators conform to SQL92 spec regarding NULL handling. As I read it, the spec requires a non-null result in some cases where one of the inputs is NULL: specifically, if the other endpoint of that interval is between the endpoints of the other interval, then the result is known TRUE despite the missing endpoint. The spec could've been a lot simpler if they did not intend this behavior. I did not force an initdb for this change, but if you don't do one you'll still see the old strict-function behavior.
2000-12-07 19:38:59 +01:00
else if (!te2IsNull)
{
if (TIMESTAMP_GT(ts2, te2))
{
pgindent run. Make it all clean.
2001-03-22 05:01:46 +01:00
Datum tt = ts2;
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
Make OVERLAPS operators conform to SQL92 spec regarding NULL handling. As I read it, the spec requires a non-null result in some cases where one of the inputs is NULL: specifically, if the other endpoint of that interval is between the endpoints of the other interval, then the result is known TRUE despite the missing endpoint. The spec could've been a lot simpler if they did not intend this behavior. I did not force an initdb for this change, but if you don't do one you'll still see the old strict-function behavior.
2000-12-07 19:38:59 +01:00
ts2 = te2;
te2 = tt;
}
}
/*
* At this point neither ts1 nor ts2 is null, so we can consider three
* cases: ts1 > ts2, ts1 < ts2, ts1 = ts2
*/
if (TIMESTAMP_GT(ts1, ts2))
{
pgindent run. Make it all clean.
2001-03-22 05:01:46 +01:00
/*
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* This case is ts1 < te2 OR te1 < te2, which may look redundant but
* in the presence of nulls it's not quite completely so.
Make OVERLAPS operators conform to SQL92 spec regarding NULL handling. As I read it, the spec requires a non-null result in some cases where one of the inputs is NULL: specifically, if the other endpoint of that interval is between the endpoints of the other interval, then the result is known TRUE despite the missing endpoint. The spec could've been a lot simpler if they did not intend this behavior. I did not force an initdb for this change, but if you don't do one you'll still see the old strict-function behavior.
2000-12-07 19:38:59 +01:00
*/
if (te2IsNull)
PG_RETURN_NULL();
if (TIMESTAMP_LT(ts1, te2))
PG_RETURN_BOOL(true);
if (te1IsNull)
PG_RETURN_NULL();
pgindent run. Make it all clean.
2001-03-22 05:01:46 +01:00
/*
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* If te1 is not null then we had ts1 <= te1 above, and we just found
* ts1 >= te2, hence te1 >= te2.
Make OVERLAPS operators conform to SQL92 spec regarding NULL handling. As I read it, the spec requires a non-null result in some cases where one of the inputs is NULL: specifically, if the other endpoint of that interval is between the endpoints of the other interval, then the result is known TRUE despite the missing endpoint. The spec could've been a lot simpler if they did not intend this behavior. I did not force an initdb for this change, but if you don't do one you'll still see the old strict-function behavior.
2000-12-07 19:38:59 +01:00
*/
PG_RETURN_BOOL(false);
}
else if (TIMESTAMP_LT(ts1, ts2))
{
/* This case is ts2 < te1 OR te2 < te1 */
if (te1IsNull)
PG_RETURN_NULL();
if (TIMESTAMP_LT(ts2, te1))
PG_RETURN_BOOL(true);
if (te2IsNull)
PG_RETURN_NULL();
pgindent run. Make it all clean.
2001-03-22 05:01:46 +01:00
/*
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* If te2 is not null then we had ts2 <= te2 above, and we just found
* ts2 >= te1, hence te2 >= te1.
Make OVERLAPS operators conform to SQL92 spec regarding NULL handling. As I read it, the spec requires a non-null result in some cases where one of the inputs is NULL: specifically, if the other endpoint of that interval is between the endpoints of the other interval, then the result is known TRUE despite the missing endpoint. The spec could've been a lot simpler if they did not intend this behavior. I did not force an initdb for this change, but if you don't do one you'll still see the old strict-function behavior.
2000-12-07 19:38:59 +01:00
*/
PG_RETURN_BOOL(false);
}
else
{
pgindent run. Make it all clean.
2001-03-22 05:01:46 +01:00
/*
* For ts1 = ts2 the spec says te1 <> te2 OR te1 = te2, which is a
Various minor corrections of and improvements to comments. Aleksander Alekseev
2016-03-18 14:38:59 +01:00
* rather silly way of saying "true if both are non-null, else null".
Make OVERLAPS operators conform to SQL92 spec regarding NULL handling. As I read it, the spec requires a non-null result in some cases where one of the inputs is NULL: specifically, if the other endpoint of that interval is between the endpoints of the other interval, then the result is known TRUE despite the missing endpoint. The spec could've been a lot simpler if they did not intend this behavior. I did not force an initdb for this change, but if you don't do one you'll still see the old strict-function behavior.
2000-12-07 19:38:59 +01:00
*/
if (te1IsNull || te2IsNull)
PG_RETURN_NULL();
PG_RETURN_BOOL(true);
}
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
#undef TIMESTAMP_GT
#undef TIMESTAMP_LT
}
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
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
/*----------------------------------------------------------
* "Arithmetic" operators on date/times.
*---------------------------------------------------------*/
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
Datum
timestamp_smaller(PG_FUNCTION_ARGS)
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
{
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
Timestamp dt1 = PG_GETARG_TIMESTAMP(0);
Timestamp dt2 = PG_GETARG_TIMESTAMP(1);
Timestamp result;
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
Fix floating-point timestamp comparisons to not go nuts if NaN is encountered; per bug report from Christian van der Leeden 8/7/03. Also, adjust larger/smaller routines (MAX/MIN) to share code with comparisons for timestamp, interval, timetz.
2003-08-08 02:10:31 +02:00
/* use timestamp_cmp_internal to be sure this agrees with comparisons */
if (timestamp_cmp_internal(dt1, dt2) < 0)
result = dt1;
else
result = dt2;
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
PG_RETURN_TIMESTAMP(result);
}
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
Datum
timestamp_larger(PG_FUNCTION_ARGS)
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
{
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
Timestamp dt1 = PG_GETARG_TIMESTAMP(0);
Timestamp dt2 = PG_GETARG_TIMESTAMP(1);
Timestamp result;
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
Fix floating-point timestamp comparisons to not go nuts if NaN is encountered; per bug report from Christian van der Leeden 8/7/03. Also, adjust larger/smaller routines (MAX/MIN) to share code with comparisons for timestamp, interval, timetz.
2003-08-08 02:10:31 +02:00
if (timestamp_cmp_internal(dt1, dt2) > 0)
result = dt1;
else
result = dt2;
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
PG_RETURN_TIMESTAMP(result);
}
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
Datum
timestamp_mi(PG_FUNCTION_ARGS)
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
{
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
Timestamp dt1 = PG_GETARG_TIMESTAMP(0);
Timestamp dt2 = PG_GETARG_TIMESTAMP(1);
Back out pfrees for justify_hour function calls.
2005-08-25 05:53:22 +02:00
Interval *result;
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
result = (Interval *) palloc(sizeof(Interval));
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +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
if (TIMESTAMP_NOT_FINITE(dt1) || TIMESTAMP_NOT_FINITE(dt2))
Error message editing in utils/adt. Again thanks to Joe Conway for doing the bulk of the heavy lifting ...
2003-07-27 06:53:12 +02:00
ereport(ERROR,
(errcode(ERRCODE_DATETIME_VALUE_OUT_OF_RANGE),
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("cannot subtract infinite timestamps")));
Error message editing in utils/adt. Again thanks to Joe Conway for doing the bulk of the heavy lifting ...
2003-07-27 06:53:12 +02:00
timestamptz_izone should return the input, not NULL, when the input is a non-finite timestamp, for consistency with related functions. In other words: +infinity rotated to a different timezone is still +infinity.
2005-09-09 08:46:14 +02:00
result->time = dt1 - dt2;
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
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
result->month = 0;
Add 'day' field to INTERVAL so 1 day interval can be distinguished from 24 hours. This is very helpful for daylight savings time: select '2005-05-03 00:00:00 EST'::timestamp with time zone + '24 hours'; ?column? ---------------------- 2005-05-04 01:00:00-04 select '2005-05-03 00:00:00 EST'::timestamp with time zone + '1 day'; ?column? ---------------------- 2005-05-04 01:00:00-04 Michael Glaesemann
2005-07-20 18:42:32 +02:00
result->day = 0;
Fix up comment munged by pg_indent. (Not pg_indent's fault; should have protected comment with dashes the first time round.)
2005-11-22 23:30:33 +01:00
/*----------
* This is wrong, but removing it breaks a lot of regression tests.
* For example:
Add comment documenting actual failure case of using interval_justify_hours in timestamp subtraction. TODO already has text description.
2005-10-27 04:45:22 +02:00
*
Fix up comment munged by pg_indent. (Not pg_indent's fault; should have protected comment with dashes the first time round.)
2005-11-22 23:30:33 +01:00
* test=> SET timezone = 'EST5EDT';
* test=> SELECT
* test-> ('2005-10-30 13:22:00-05'::timestamptz -
pgindent run for 8.2.
2006-10-04 02:30:14 +02:00
* test(> '2005-10-29 13:22:00-04'::timestamptz);
Fix up comment munged by pg_indent. (Not pg_indent's fault; should have protected comment with dashes the first time round.)
2005-11-22 23:30:33 +01:00
* ?column?
* ----------------
* 1 day 01:00:00
* (1 row)
Add comment documenting actual failure case of using interval_justify_hours in timestamp subtraction. TODO already has text description.
2005-10-27 04:45:22 +02:00
*
Fix up comment munged by pg_indent. (Not pg_indent's fault; should have protected comment with dashes the first time round.)
2005-11-22 23:30:33 +01:00
* so adding that to the first timestamp gets:
Add comment documenting actual failure case of using interval_justify_hours in timestamp subtraction. TODO already has text description.
2005-10-27 04:45:22 +02:00
*
Fix up comment munged by pg_indent. (Not pg_indent's fault; should have protected comment with dashes the first time round.)
2005-11-22 23:30:33 +01:00
* test=> SELECT
* test-> ('2005-10-29 13:22:00-04'::timestamptz +
* test(> ('2005-10-30 13:22:00-05'::timestamptz -
* test(> '2005-10-29 13:22:00-04'::timestamptz)) at time zone 'EST';
pgindent run for 8.2.
2006-10-04 02:30:14 +02:00
* timezone
Fix up comment munged by pg_indent. (Not pg_indent's fault; should have protected comment with dashes the first time round.)
2005-11-22 23:30:33 +01:00
* --------------------
* 2005-10-30 14:22:00
* (1 row)
*----------
Add comment documenting actual failure case of using interval_justify_hours in timestamp subtraction. TODO already has text description.
2005-10-27 04:45:22 +02:00
*/
Re-run pgindent, fixing a problem where comment lines after a blank comment line where output as too long, and update typedefs for /lib directory. Also fix case where identifiers were used as variable names in the backend, but as typedefs in ecpg (favor the backend for indenting). Backpatch to 8.1.X.
2005-11-22 19:17:34 +01:00
result = DatumGetIntervalP(DirectFunctionCall1(interval_justify_hours,
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
IntervalPGetDatum(result)));
Remove justify_hours call from interval_mul and interval_div, and make some small stylistic improvements in these functions. Also fix several places where TMODULO() was being used with wrong-sized quotient argument, creating a risk of overflow --- interval2tm was actually capable of going into an infinite loop because of this.
2005-10-25 19:13:07 +02:00
Back out pfrees for justify_hour function calls.
2005-08-25 05:53:22 +02:00
PG_RETURN_INTERVAL_P(result);
Add 'day' field to INTERVAL so 1 day interval can be distinguished from 24 hours. This is very helpful for daylight savings time: select '2005-05-03 00:00:00 EST'::timestamp with time zone + '24 hours'; ?column? ---------------------- 2005-05-04 01:00:00-04 select '2005-05-03 00:00:00 EST'::timestamp with time zone + '1 day'; ?column? ---------------------- 2005-05-04 01:00:00-04 Michael Glaesemann
2005-07-20 18:42:32 +02:00
}
Attached is the new patch. To summarize: - new function justify_interval(interval) - modified function justify_hours(interval) - modified function justify_days(interval) These functions are defined to meet the requirements as discussed in this thread. Specifically: - justify_hours makes certain the sign bit on the hours matches the sign bit on the days. It only checks the sign bit on the days, and not the months, when determining if the hours should be positive or negative. After the call, -24 < hours < 24. - justify_days makes certain the sign bit on the days matches the sign bit on the months. It's behavior does not depend on the hours, nor does it modify the hours. After the call, -30 < days < 30. - justify_interval makes sure the sign bits on all three fields months, days, and hours are all the same. After the call, -24 < hours < 24 AND -30 < days < 30. Mark Dilger
2006-03-06 23:49:17 +01:00
/*
pgindent run for 8.2.
2006-10-04 02:30:14 +02:00
* interval_justify_interval()
Attached is the new patch. To summarize: - new function justify_interval(interval) - modified function justify_hours(interval) - modified function justify_days(interval) These functions are defined to meet the requirements as discussed in this thread. Specifically: - justify_hours makes certain the sign bit on the hours matches the sign bit on the days. It only checks the sign bit on the days, and not the months, when determining if the hours should be positive or negative. After the call, -24 < hours < 24. - justify_days makes certain the sign bit on the days matches the sign bit on the months. It's behavior does not depend on the hours, nor does it modify the hours. After the call, -30 < days < 30. - justify_interval makes sure the sign bits on all three fields months, days, and hours are all the same. After the call, -24 < hours < 24 AND -30 < days < 30. Mark Dilger
2006-03-06 23:49:17 +01:00
*
pgindent run for 8.2.
2006-10-04 02:30:14 +02:00
* Adjust interval so 'month', 'day', and 'time' portions are within
* customary bounds. Specifically:
Attached is the new patch. To summarize: - new function justify_interval(interval) - modified function justify_hours(interval) - modified function justify_days(interval) These functions are defined to meet the requirements as discussed in this thread. Specifically: - justify_hours makes certain the sign bit on the hours matches the sign bit on the days. It only checks the sign bit on the days, and not the months, when determining if the hours should be positive or negative. After the call, -24 < hours < 24. - justify_days makes certain the sign bit on the days matches the sign bit on the months. It's behavior does not depend on the hours, nor does it modify the hours. After the call, -30 < days < 30. - justify_interval makes sure the sign bits on all three fields months, days, and hours are all the same. After the call, -24 < hours < 24 AND -30 < days < 30. Mark Dilger
2006-03-06 23:49:17 +01:00
*
pgindent run for 8.2.
2006-10-04 02:30:14 +02:00
* 0 <= abs(time) < 24 hours
* 0 <= abs(day) < 30 days
Attached is the new patch. To summarize: - new function justify_interval(interval) - modified function justify_hours(interval) - modified function justify_days(interval) These functions are defined to meet the requirements as discussed in this thread. Specifically: - justify_hours makes certain the sign bit on the hours matches the sign bit on the days. It only checks the sign bit on the days, and not the months, when determining if the hours should be positive or negative. After the call, -24 < hours < 24. - justify_days makes certain the sign bit on the days matches the sign bit on the months. It's behavior does not depend on the hours, nor does it modify the hours. After the call, -30 < days < 30. - justify_interval makes sure the sign bits on all three fields months, days, and hours are all the same. After the call, -24 < hours < 24 AND -30 < days < 30. Mark Dilger
2006-03-06 23:49:17 +01:00
*
pgindent run for 8.2.
2006-10-04 02:30:14 +02:00
* Also, the sign bit on all three fields is made equal, so either
* all three fields are negative or all are positive.
Attached is the new patch. To summarize: - new function justify_interval(interval) - modified function justify_hours(interval) - modified function justify_days(interval) These functions are defined to meet the requirements as discussed in this thread. Specifically: - justify_hours makes certain the sign bit on the hours matches the sign bit on the days. It only checks the sign bit on the days, and not the months, when determining if the hours should be positive or negative. After the call, -24 < hours < 24. - justify_days makes certain the sign bit on the days matches the sign bit on the months. It's behavior does not depend on the hours, nor does it modify the hours. After the call, -30 < days < 30. - justify_interval makes sure the sign bits on all three fields months, days, and hours are all the same. After the call, -24 < hours < 24 AND -30 < days < 30. Mark Dilger
2006-03-06 23:49:17 +01:00
*/
Datum
interval_justify_interval(PG_FUNCTION_ARGS)
{
Interval *span = PG_GETARG_INTERVAL_P(0);
Interval *result;
Get rid of a bunch of #ifdef HAVE_INT64_TIMESTAMP conditionals by inventing a new typedef TimeOffset to represent an intermediate time value. It's either int64 or double as appropriate, and in most usages will be measured in microseconds or seconds the same as Timestamp. We don't call it Timestamp, though, since the value doesn't necessarily represent an absolute time instant. Warren Turkal
2008-03-21 02:31:43 +01:00
TimeOffset wholeday;
Attached is the new patch. To summarize: - new function justify_interval(interval) - modified function justify_hours(interval) - modified function justify_days(interval) These functions are defined to meet the requirements as discussed in this thread. Specifically: - justify_hours makes certain the sign bit on the hours matches the sign bit on the days. It only checks the sign bit on the days, and not the months, when determining if the hours should be positive or negative. After the call, -24 < hours < 24. - justify_days makes certain the sign bit on the days matches the sign bit on the months. It's behavior does not depend on the hours, nor does it modify the hours. After the call, -30 < days < 30. - justify_interval makes sure the sign bits on all three fields months, days, and hours are all the same. After the call, -24 < hours < 24 AND -30 < days < 30. Mark Dilger
2006-03-06 23:49:17 +01:00
int32 wholemonth;
result = (Interval *) palloc(sizeof(Interval));
result->month = span->month;
result->day = span->day;
result->time = span->time;
TMODULO(result->time, wholeday, USECS_PER_DAY);
result->day += wholeday; /* could overflow... */
wholemonth = result->day / DAYS_PER_MONTH;
result->day -= wholemonth * DAYS_PER_MONTH;
result->month += wholemonth;
if (result->month > 0 &&
(result->day < 0 || (result->day == 0 && result->time < 0)))
{
result->day += DAYS_PER_MONTH;
result->month--;
}
else if (result->month < 0 &&
pgindent run for 8.2.
2006-10-04 02:30:14 +02:00
(result->day > 0 || (result->day == 0 && result->time > 0)))
Attached is the new patch. To summarize: - new function justify_interval(interval) - modified function justify_hours(interval) - modified function justify_days(interval) These functions are defined to meet the requirements as discussed in this thread. Specifically: - justify_hours makes certain the sign bit on the hours matches the sign bit on the days. It only checks the sign bit on the days, and not the months, when determining if the hours should be positive or negative. After the call, -24 < hours < 24. - justify_days makes certain the sign bit on the days matches the sign bit on the months. It's behavior does not depend on the hours, nor does it modify the hours. After the call, -30 < days < 30. - justify_interval makes sure the sign bits on all three fields months, days, and hours are all the same. After the call, -24 < hours < 24 AND -30 < days < 30. Mark Dilger
2006-03-06 23:49:17 +01:00
{
result->day -= DAYS_PER_MONTH;
result->month++;
}
if (result->day > 0 && result->time < 0)
{
result->time += USECS_PER_DAY;
result->day--;
}
else if (result->day < 0 && result->time > 0)
pgindent run for 8.2.
2006-10-04 02:30:14 +02:00
{
Attached is the new patch. To summarize: - new function justify_interval(interval) - modified function justify_hours(interval) - modified function justify_days(interval) These functions are defined to meet the requirements as discussed in this thread. Specifically: - justify_hours makes certain the sign bit on the hours matches the sign bit on the days. It only checks the sign bit on the days, and not the months, when determining if the hours should be positive or negative. After the call, -24 < hours < 24. - justify_days makes certain the sign bit on the days matches the sign bit on the months. It's behavior does not depend on the hours, nor does it modify the hours. After the call, -30 < days < 30. - justify_interval makes sure the sign bits on all three fields months, days, and hours are all the same. After the call, -24 < hours < 24 AND -30 < days < 30. Mark Dilger
2006-03-06 23:49:17 +01:00
result->time -= USECS_PER_DAY;
result->day++;
}
PG_RETURN_INTERVAL_P(result);
}
Remove justify_hours call from interval_mul and interval_div, and make some small stylistic improvements in these functions. Also fix several places where TMODULO() was being used with wrong-sized quotient argument, creating a risk of overflow --- interval2tm was actually capable of going into an infinite loop because of this.
2005-10-25 19:13:07 +02:00
/*
* interval_justify_hours()
*
* Adjust interval so 'time' contains less than a whole day, adding
* the excess to 'day'. This is useful for
Add 'day' field to INTERVAL so 1 day interval can be distinguished from 24 hours. This is very helpful for daylight savings time: select '2005-05-03 00:00:00 EST'::timestamp with time zone + '24 hours'; ?column? ---------------------- 2005-05-04 01:00:00-04 select '2005-05-03 00:00:00 EST'::timestamp with time zone + '1 day'; ?column? ---------------------- 2005-05-04 01:00:00-04 Michael Glaesemann
2005-07-20 18:42:32 +02:00
* situations (such as non-TZ) where '1 day' = '24 hours' is valid,
Remove justify_hours call from interval_mul and interval_div, and make some small stylistic improvements in these functions. Also fix several places where TMODULO() was being used with wrong-sized quotient argument, creating a risk of overflow --- interval2tm was actually capable of going into an infinite loop because of this.
2005-10-25 19:13:07 +02:00
* e.g. interval subtraction and division.
Add 'day' field to INTERVAL so 1 day interval can be distinguished from 24 hours. This is very helpful for daylight savings time: select '2005-05-03 00:00:00 EST'::timestamp with time zone + '24 hours'; ?column? ---------------------- 2005-05-04 01:00:00-04 select '2005-05-03 00:00:00 EST'::timestamp with time zone + '1 day'; ?column? ---------------------- 2005-05-04 01:00:00-04 Michael Glaesemann
2005-07-20 18:42:32 +02:00
*/
Datum
interval_justify_hours(PG_FUNCTION_ARGS)
{
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
Interval *span = PG_GETARG_INTERVAL_P(0);
Interval *result;
Get rid of a bunch of #ifdef HAVE_INT64_TIMESTAMP conditionals by inventing a new typedef TimeOffset to represent an intermediate time value. It's either int64 or double as appropriate, and in most usages will be measured in microseconds or seconds the same as Timestamp. We don't call it Timestamp, though, since the value doesn't necessarily represent an absolute time instant. Warren Turkal
2008-03-21 02:31:43 +01:00
TimeOffset wholeday;
Add 'day' field to INTERVAL so 1 day interval can be distinguished from 24 hours. This is very helpful for daylight savings time: select '2005-05-03 00:00:00 EST'::timestamp with time zone + '24 hours'; ?column? ---------------------- 2005-05-04 01:00:00-04 select '2005-05-03 00:00:00 EST'::timestamp with time zone + '1 day'; ?column? ---------------------- 2005-05-04 01:00:00-04 Michael Glaesemann
2005-07-20 18:42:32 +02:00
result = (Interval *) palloc(sizeof(Interval));
result->month = span->month;
Remove justify_hours call from interval_mul and interval_div, and make some small stylistic improvements in these functions. Also fix several places where TMODULO() was being used with wrong-sized quotient argument, creating a risk of overflow --- interval2tm was actually capable of going into an infinite loop because of this.
2005-10-25 19:13:07 +02:00
result->day = span->day;
Add 'day' field to INTERVAL so 1 day interval can be distinguished from 24 hours. This is very helpful for daylight savings time: select '2005-05-03 00:00:00 EST'::timestamp with time zone + '24 hours'; ?column? ---------------------- 2005-05-04 01:00:00-04 select '2005-05-03 00:00:00 EST'::timestamp with time zone + '1 day'; ?column? ---------------------- 2005-05-04 01:00:00-04 Michael Glaesemann
2005-07-20 18:42:32 +02:00
result->time = span->time;
Remove justify_hours call from interval_mul and interval_div, and make some small stylistic improvements in these functions. Also fix several places where TMODULO() was being used with wrong-sized quotient argument, creating a risk of overflow --- interval2tm was actually capable of going into an infinite loop because of this.
2005-10-25 19:13:07 +02:00
TMODULO(result->time, wholeday, USECS_PER_DAY);
result->day += wholeday; /* could overflow... */
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
Attached is the new patch. To summarize: - new function justify_interval(interval) - modified function justify_hours(interval) - modified function justify_days(interval) These functions are defined to meet the requirements as discussed in this thread. Specifically: - justify_hours makes certain the sign bit on the hours matches the sign bit on the days. It only checks the sign bit on the days, and not the months, when determining if the hours should be positive or negative. After the call, -24 < hours < 24. - justify_days makes certain the sign bit on the days matches the sign bit on the months. It's behavior does not depend on the hours, nor does it modify the hours. After the call, -30 < days < 30. - justify_interval makes sure the sign bits on all three fields months, days, and hours are all the same. After the call, -24 < hours < 24 AND -30 < days < 30. Mark Dilger
2006-03-06 23:49:17 +01:00
if (result->day > 0 && result->time < 0)
{
result->time += USECS_PER_DAY;
result->day--;
}
else if (result->day < 0 && result->time > 0)
{
result->time -= USECS_PER_DAY;
result->day++;
}
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
PG_RETURN_INTERVAL_P(result);
}
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
Remove justify_hours call from interval_mul and interval_div, and make some small stylistic improvements in these functions. Also fix several places where TMODULO() was being used with wrong-sized quotient argument, creating a risk of overflow --- interval2tm was actually capable of going into an infinite loop because of this.
2005-10-25 19:13:07 +02:00
/*
* interval_justify_days()
*
* Adjust interval so 'day' contains less than 30 days, adding
* the excess to 'month'.
Add 'day' field to INTERVAL so 1 day interval can be distinguished from 24 hours. This is very helpful for daylight savings time: select '2005-05-03 00:00:00 EST'::timestamp with time zone + '24 hours'; ?column? ---------------------- 2005-05-04 01:00:00-04 select '2005-05-03 00:00:00 EST'::timestamp with time zone + '1 day'; ?column? ---------------------- 2005-05-04 01:00:00-04 Michael Glaesemann
2005-07-20 18:42:32 +02:00
*/
Datum
interval_justify_days(PG_FUNCTION_ARGS)
{
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
Interval *span = PG_GETARG_INTERVAL_P(0);
Interval *result;
Remove justify_hours call from interval_mul and interval_div, and make some small stylistic improvements in these functions. Also fix several places where TMODULO() was being used with wrong-sized quotient argument, creating a risk of overflow --- interval2tm was actually capable of going into an infinite loop because of this.
2005-10-25 19:13:07 +02:00
int32 wholemonth;
Add 'day' field to INTERVAL so 1 day interval can be distinguished from 24 hours. This is very helpful for daylight savings time: select '2005-05-03 00:00:00 EST'::timestamp with time zone + '24 hours'; ?column? ---------------------- 2005-05-04 01:00:00-04 select '2005-05-03 00:00:00 EST'::timestamp with time zone + '1 day'; ?column? ---------------------- 2005-05-04 01:00:00-04 Michael Glaesemann
2005-07-20 18:42:32 +02:00
result = (Interval *) palloc(sizeof(Interval));
Remove justify_hours call from interval_mul and interval_div, and make some small stylistic improvements in these functions. Also fix several places where TMODULO() was being used with wrong-sized quotient argument, creating a risk of overflow --- interval2tm was actually capable of going into an infinite loop because of this.
2005-10-25 19:13:07 +02:00
result->month = span->month;
Add 'day' field to INTERVAL so 1 day interval can be distinguished from 24 hours. This is very helpful for daylight savings time: select '2005-05-03 00:00:00 EST'::timestamp with time zone + '24 hours'; ?column? ---------------------- 2005-05-04 01:00:00-04 select '2005-05-03 00:00:00 EST'::timestamp with time zone + '1 day'; ?column? ---------------------- 2005-05-04 01:00:00-04 Michael Glaesemann
2005-07-20 18:42:32 +02:00
result->day = span->day;
result->time = span->time;
Remove justify_hours call from interval_mul and interval_div, and make some small stylistic improvements in these functions. Also fix several places where TMODULO() was being used with wrong-sized quotient argument, creating a risk of overflow --- interval2tm was actually capable of going into an infinite loop because of this.
2005-10-25 19:13:07 +02:00
wholemonth = result->day / DAYS_PER_MONTH;
result->day -= wholemonth * DAYS_PER_MONTH;
result->month += wholemonth;
Add 'day' field to INTERVAL so 1 day interval can be distinguished from 24 hours. This is very helpful for daylight savings time: select '2005-05-03 00:00:00 EST'::timestamp with time zone + '24 hours'; ?column? ---------------------- 2005-05-04 01:00:00-04 select '2005-05-03 00:00:00 EST'::timestamp with time zone + '1 day'; ?column? ---------------------- 2005-05-04 01:00:00-04 Michael Glaesemann
2005-07-20 18:42:32 +02:00
Attached is the new patch. To summarize: - new function justify_interval(interval) - modified function justify_hours(interval) - modified function justify_days(interval) These functions are defined to meet the requirements as discussed in this thread. Specifically: - justify_hours makes certain the sign bit on the hours matches the sign bit on the days. It only checks the sign bit on the days, and not the months, when determining if the hours should be positive or negative. After the call, -24 < hours < 24. - justify_days makes certain the sign bit on the days matches the sign bit on the months. It's behavior does not depend on the hours, nor does it modify the hours. After the call, -30 < days < 30. - justify_interval makes sure the sign bits on all three fields months, days, and hours are all the same. After the call, -24 < hours < 24 AND -30 < days < 30. Mark Dilger
2006-03-06 23:49:17 +01:00
if (result->month > 0 && result->day < 0)
{
result->day += DAYS_PER_MONTH;
result->month--;
}
else if (result->month < 0 && result->day > 0)
{
result->day -= DAYS_PER_MONTH;
result->month++;
}
Add 'day' field to INTERVAL so 1 day interval can be distinguished from 24 hours. This is very helpful for daylight savings time: select '2005-05-03 00:00:00 EST'::timestamp with time zone + '24 hours'; ?column? ---------------------- 2005-05-04 01:00:00-04 select '2005-05-03 00:00:00 EST'::timestamp with time zone + '1 day'; ?column? ---------------------- 2005-05-04 01:00:00-04 Michael Glaesemann
2005-07-20 18:42:32 +02:00
PG_RETURN_INTERVAL_P(result);
}
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
Create crosstype comparison operators for date vs. timestamp and date vs. timestamptz. This allows use of indexes for expressions like datecol >= date 'today' - interval '1 month' which were formerly not indexable without casting the righthand side down from timestamp to date.
2004-02-14 21:16:18 +01:00
/* timestamp_pl_interval()
Collection of typo fixes. Use "a" and "an" correctly, mostly in comments. Two error messages were also fixed (they were just elogs, so no translation work required). Two function comments in pg_proc.h were also fixed. Etsuro Fujita reported one of these, but I found a lot more with grep. Also fix a few other typos spotted while grepping for the a/an typos. For example, "consists out of ..." -> "consists of ...". Plus a "though"/ "through" mixup reported by Euler Taveira. Many of these typos were in old code, which would be nice to backpatch to make future backpatching easier. But much of the code was new, and I didn't feel like crafting separate patches for each branch. So no backpatching.
2015-05-20 15:18:11 +02:00
* Add an interval to a timestamp data type.
Add 'day' field to INTERVAL so 1 day interval can be distinguished from 24 hours. This is very helpful for daylight savings time: select '2005-05-03 00:00:00 EST'::timestamp with time zone + '24 hours'; ?column? ---------------------- 2005-05-04 01:00:00-04 select '2005-05-03 00:00:00 EST'::timestamp with time zone + '1 day'; ?column? ---------------------- 2005-05-04 01:00:00-04 Michael Glaesemann
2005-07-20 18:42:32 +02:00
* Note that interval has provisions for qualitative year/month and day
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
* units, so try to do the right thing with them.
* To add a month, increment the month, and use the same day of month.
* Then, if the next month has fewer days, set the day of month
* to the last day of month.
Add 'day' field to INTERVAL so 1 day interval can be distinguished from 24 hours. This is very helpful for daylight savings time: select '2005-05-03 00:00:00 EST'::timestamp with time zone + '24 hours'; ?column? ---------------------- 2005-05-04 01:00:00-04 select '2005-05-03 00:00:00 EST'::timestamp with time zone + '1 day'; ?column? ---------------------- 2005-05-04 01:00:00-04 Michael Glaesemann
2005-07-20 18:42:32 +02:00
* To add a day, increment the mday, and use the same time of day.
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
* Lastly, add in the "quantitative time".
*/
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
Datum
Create crosstype comparison operators for date vs. timestamp and date vs. timestamptz. This allows use of indexes for expressions like datecol >= date 'today' - interval '1 month' which were formerly not indexable without casting the righthand side down from timestamp to date.
2004-02-14 21:16:18 +01:00
timestamp_pl_interval(PG_FUNCTION_ARGS)
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
{
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
Timestamp timestamp = PG_GETARG_TIMESTAMP(0);
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
Interval *span = PG_GETARG_INTERVAL_P(1);
Timestamp result;
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 (TIMESTAMP_NOT_FINITE(timestamp))
result = timestamp;
else
{
if (span->month != 0)
{
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
struct pg_tm tt,
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 = &tt;
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;
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 unnecessary parentheses in assignments. Add spaces where needed. Reference time interval variables as tinterval.
2005-07-21 06:41:43 +02:00
if (timestamp2tm(timestamp, NULL, tm, &fsec, NULL, NULL) != 0)
Error message editing in utils/adt. Again thanks to Joe Conway for doing the bulk of the heavy lifting ...
2003-07-27 06:53:12 +02:00
ereport(ERROR,
(errcode(ERRCODE_DATETIME_VALUE_OUT_OF_RANGE),
errmsg("timestamp out of range")));
tm->tm_mon += span->month;
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
if (tm->tm_mon > MONTHS_PER_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
{
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_year += (tm->tm_mon - 1) / MONTHS_PER_YEAR;
tm->tm_mon = ((tm->tm_mon - 1) % MONTHS_PER_YEAR) + 1;
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
}
Error message editing in utils/adt. Again thanks to Joe Conway for doing the bulk of the heavy lifting ...
2003-07-27 06:53:12 +02:00
else if (tm->tm_mon < 1)
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
{
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_year += tm->tm_mon / MONTHS_PER_YEAR - 1;
tm->tm_mon = tm->tm_mon % MONTHS_PER_YEAR + MONTHS_PER_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
}
Error message editing in utils/adt. Again thanks to Joe Conway for doing the bulk of the heavy lifting ...
2003-07-27 06:53:12 +02:00
/* adjust for end of month boundary problems... */
if (tm->tm_mday > day_tab[isleap(tm->tm_year)][tm->tm_mon - 1])
tm->tm_mday = (day_tab[isleap(tm->tm_year)][tm->tm_mon - 1]);
Remove unnecessary parentheses in assignments. Add spaces where needed. Reference time interval variables as tinterval.
2005-07-21 06:41:43 +02:00
if (tm2timestamp(tm, fsec, NULL, &timestamp) != 0)
Error message editing in utils/adt. Again thanks to Joe Conway for doing the bulk of the heavy lifting ...
2003-07-27 06:53:12 +02:00
ereport(ERROR,
(errcode(ERRCODE_DATETIME_VALUE_OUT_OF_RANGE),
errmsg("timestamp out of range")));
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
}
Add 'day' field to INTERVAL so 1 day interval can be distinguished from 24 hours. This is very helpful for daylight savings time: select '2005-05-03 00:00:00 EST'::timestamp with time zone + '24 hours'; ?column? ---------------------- 2005-05-04 01:00:00-04 select '2005-05-03 00:00:00 EST'::timestamp with time zone + '1 day'; ?column? ---------------------- 2005-05-04 01:00:00-04 Michael Glaesemann
2005-07-20 18:42:32 +02:00
if (span->day != 0)
{
struct pg_tm tt,
*tm = &tt;
fsec_t fsec;
int julian;
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +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
if (timestamp2tm(timestamp, NULL, tm, &fsec, NULL, NULL) != 0)
Add 'day' field to INTERVAL so 1 day interval can be distinguished from 24 hours. This is very helpful for daylight savings time: select '2005-05-03 00:00:00 EST'::timestamp with time zone + '24 hours'; ?column? ---------------------- 2005-05-04 01:00:00-04 select '2005-05-03 00:00:00 EST'::timestamp with time zone + '1 day'; ?column? ---------------------- 2005-05-04 01:00:00-04 Michael Glaesemann
2005-07-20 18:42:32 +02:00
ereport(ERROR,
(errcode(ERRCODE_DATETIME_VALUE_OUT_OF_RANGE),
errmsg("timestamp out of range")));
Various minor corrections of and improvements to comments. Aleksander Alekseev
2016-03-18 14:38:59 +01:00
/* Add days by converting to and from Julian */
Add 'day' field to INTERVAL so 1 day interval can be distinguished from 24 hours. This is very helpful for daylight savings time: select '2005-05-03 00:00:00 EST'::timestamp with time zone + '24 hours'; ?column? ---------------------- 2005-05-04 01:00:00-04 select '2005-05-03 00:00:00 EST'::timestamp with time zone + '1 day'; ?column? ---------------------- 2005-05-04 01:00:00-04 Michael Glaesemann
2005-07-20 18:42:32 +02:00
julian = date2j(tm->tm_year, tm->tm_mon, tm->tm_mday) + span->day;
j2date(julian, &tm->tm_year, &tm->tm_mon, &tm->tm_mday);
Remove unnecessary parentheses in assignments. Add spaces where needed. Reference time interval variables as tinterval.
2005-07-21 06:41:43 +02:00
if (tm2timestamp(tm, fsec, NULL, &timestamp) != 0)
Add 'day' field to INTERVAL so 1 day interval can be distinguished from 24 hours. This is very helpful for daylight savings time: select '2005-05-03 00:00:00 EST'::timestamp with time zone + '24 hours'; ?column? ---------------------- 2005-05-04 01:00:00-04 select '2005-05-03 00:00:00 EST'::timestamp with time zone + '1 day'; ?column? ---------------------- 2005-05-04 01:00:00-04 Michael Glaesemann
2005-07-20 18:42:32 +02:00
ereport(ERROR,
(errcode(ERRCODE_DATETIME_VALUE_OUT_OF_RANGE),
errmsg("timestamp out of range")));
}
timestamp += span->time;
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
if (!IS_VALID_TIMESTAMP(timestamp))
ereport(ERROR,
(errcode(ERRCODE_DATETIME_VALUE_OUT_OF_RANGE),
errmsg("timestamp out of range")));
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
result = timestamp;
}
PG_RETURN_TIMESTAMP(result);
}
Datum
Create crosstype comparison operators for date vs. timestamp and date vs. timestamptz. This allows use of indexes for expressions like datecol >= date 'today' - interval '1 month' which were formerly not indexable without casting the righthand side down from timestamp to date.
2004-02-14 21:16:18 +01:00
timestamp_mi_interval(PG_FUNCTION_ARGS)
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
{
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
Timestamp timestamp = PG_GETARG_TIMESTAMP(0);
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
Interval *span = PG_GETARG_INTERVAL_P(1);
Interval tspan;
tspan.month = -span->month;
Add 'day' field to INTERVAL so 1 day interval can be distinguished from 24 hours. This is very helpful for daylight savings time: select '2005-05-03 00:00:00 EST'::timestamp with time zone + '24 hours'; ?column? ---------------------- 2005-05-04 01:00:00-04 select '2005-05-03 00:00:00 EST'::timestamp with time zone + '1 day'; ?column? ---------------------- 2005-05-04 01:00:00-04 Michael Glaesemann
2005-07-20 18:42:32 +02:00
tspan.day = -span->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
tspan.time = -span->time;
Create crosstype comparison operators for date vs. timestamp and date vs. timestamptz. This allows use of indexes for expressions like datecol >= date 'today' - interval '1 month' which were formerly not indexable without casting the righthand side down from timestamp to date.
2004-02-14 21:16:18 +01:00
return DirectFunctionCall2(timestamp_pl_interval,
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
TimestampGetDatum(timestamp),
PointerGetDatum(&tspan));
}
Create crosstype comparison operators for date vs. timestamp and date vs. timestamptz. This allows use of indexes for expressions like datecol >= date 'today' - interval '1 month' which were formerly not indexable without casting the righthand side down from timestamp to date.
2004-02-14 21:16:18 +01:00
/* timestamptz_pl_interval()
Collection of typo fixes. Use "a" and "an" correctly, mostly in comments. Two error messages were also fixed (they were just elogs, so no translation work required). Two function comments in pg_proc.h were also fixed. Etsuro Fujita reported one of these, but I found a lot more with grep. Also fix a few other typos spotted while grepping for the a/an typos. For example, "consists out of ..." -> "consists of ...". Plus a "though"/ "through" mixup reported by Euler Taveira. Many of these typos were in old code, which would be nice to backpatch to make future backpatching easier. But much of the code was new, and I didn't feel like crafting separate patches for each branch. So no backpatching.
2015-05-20 15:18:11 +02:00
* Add an interval to a timestamp with time zone data type.
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
* Note that interval has provisions for qualitative year/month
* units, so try to do the right thing with them.
* To add a month, increment the month, and use the same day of month.
* Then, if the next month has fewer days, set the day of month
* to the last day of month.
* Lastly, add in the "quantitative time".
*/
Datum
Create crosstype comparison operators for date vs. timestamp and date vs. timestamptz. This allows use of indexes for expressions like datecol >= date 'today' - interval '1 month' which were formerly not indexable without casting the righthand side down from timestamp to date.
2004-02-14 21:16:18 +01:00
timestamptz_pl_interval(PG_FUNCTION_ARGS)
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
{
Add binary I/O routines for a bunch more datatypes. Still a few to go, but that was enough tedium for one day. Along the way, move the few support routines for types xid and cid into a more logical place.
2003-05-13 01:08:52 +02:00
TimestampTz timestamp = PG_GETARG_TIMESTAMPTZ(0);
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
Interval *span = PG_GETARG_INTERVAL_P(1);
pgindent run on all C files. Java run to follow. initdb/regression tests pass.
2001-10-25 07:50:21 +02:00
TimestampTz result;
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
int tz;
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
if (TIMESTAMP_NOT_FINITE(timestamp))
result = timestamp;
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
else
{
if (span->month != 0)
{
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
struct pg_tm tt,
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
*tm = &tt;
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;
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
Remove unused tzn arguments for timestamp2tm()
2012-03-15 20:13:35 +01:00
if (timestamp2tm(timestamp, &tz, tm, &fsec, NULL, NULL) != 0)
Error message editing in utils/adt. Again thanks to Joe Conway for doing the bulk of the heavy lifting ...
2003-07-27 06:53:12 +02:00
ereport(ERROR,
(errcode(ERRCODE_DATETIME_VALUE_OUT_OF_RANGE),
errmsg("timestamp out of range")));
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
Error message editing in utils/adt. Again thanks to Joe Conway for doing the bulk of the heavy lifting ...
2003-07-27 06:53:12 +02:00
tm->tm_mon += span->month;
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
if (tm->tm_mon > MONTHS_PER_YEAR)
Error message editing in utils/adt. Again thanks to Joe Conway for doing the bulk of the heavy lifting ...
2003-07-27 06:53:12 +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
tm->tm_year += (tm->tm_mon - 1) / MONTHS_PER_YEAR;
tm->tm_mon = ((tm->tm_mon - 1) % MONTHS_PER_YEAR) + 1;
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
}
Error message editing in utils/adt. Again thanks to Joe Conway for doing the bulk of the heavy lifting ...
2003-07-27 06:53:12 +02:00
else if (tm->tm_mon < 1)
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
{
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_year += tm->tm_mon / MONTHS_PER_YEAR - 1;
tm->tm_mon = tm->tm_mon % MONTHS_PER_YEAR + MONTHS_PER_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
}
Error message editing in utils/adt. Again thanks to Joe Conway for doing the bulk of the heavy lifting ...
2003-07-27 06:53:12 +02:00
/* adjust for end of month boundary problems... */
if (tm->tm_mday > day_tab[isleap(tm->tm_year)][tm->tm_mon - 1])
tm->tm_mday = (day_tab[isleap(tm->tm_year)][tm->tm_mon - 1]);
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
tz = DetermineTimeZoneOffset(tm, session_timezone);
Error message editing in utils/adt. Again thanks to Joe Conway for doing the bulk of the heavy lifting ...
2003-07-27 06:53:12 +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
if (tm2timestamp(tm, fsec, &tz, &timestamp) != 0)
Error message editing in utils/adt. Again thanks to Joe Conway for doing the bulk of the heavy lifting ...
2003-07-27 06:53:12 +02:00
ereport(ERROR,
(errcode(ERRCODE_DATETIME_VALUE_OUT_OF_RANGE),
errmsg("timestamp out of range")));
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
}
Add 'day' field to INTERVAL so 1 day interval can be distinguished from 24 hours. This is very helpful for daylight savings time: select '2005-05-03 00:00:00 EST'::timestamp with time zone + '24 hours'; ?column? ---------------------- 2005-05-04 01:00:00-04 select '2005-05-03 00:00:00 EST'::timestamp with time zone + '1 day'; ?column? ---------------------- 2005-05-04 01:00:00-04 Michael Glaesemann
2005-07-20 18:42:32 +02:00
if (span->day != 0)
{
struct pg_tm tt,
*tm = &tt;
fsec_t fsec;
int julian;
Remove unused tzn arguments for timestamp2tm()
2012-03-15 20:13:35 +01:00
if (timestamp2tm(timestamp, &tz, tm, &fsec, NULL, NULL) != 0)
Add 'day' field to INTERVAL so 1 day interval can be distinguished from 24 hours. This is very helpful for daylight savings time: select '2005-05-03 00:00:00 EST'::timestamp with time zone + '24 hours'; ?column? ---------------------- 2005-05-04 01:00:00-04 select '2005-05-03 00:00:00 EST'::timestamp with time zone + '1 day'; ?column? ---------------------- 2005-05-04 01:00:00-04 Michael Glaesemann
2005-07-20 18:42:32 +02:00
ereport(ERROR,
(errcode(ERRCODE_DATETIME_VALUE_OUT_OF_RANGE),
errmsg("timestamp out of range")));
Various minor corrections of and improvements to comments. Aleksander Alekseev
2016-03-18 14:38:59 +01:00
/* Add days by converting to and from Julian */
Add 'day' field to INTERVAL so 1 day interval can be distinguished from 24 hours. This is very helpful for daylight savings time: select '2005-05-03 00:00:00 EST'::timestamp with time zone + '24 hours'; ?column? ---------------------- 2005-05-04 01:00:00-04 select '2005-05-03 00:00:00 EST'::timestamp with time zone + '1 day'; ?column? ---------------------- 2005-05-04 01:00:00-04 Michael Glaesemann
2005-07-20 18:42:32 +02:00
julian = date2j(tm->tm_year, tm->tm_mon, tm->tm_mday) + span->day;
j2date(julian, &tm->tm_year, &tm->tm_mon, &tm->tm_mday);
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
tz = DetermineTimeZoneOffset(tm, session_timezone);
Add 'day' field to INTERVAL so 1 day interval can be distinguished from 24 hours. This is very helpful for daylight savings time: select '2005-05-03 00:00:00 EST'::timestamp with time zone + '24 hours'; ?column? ---------------------- 2005-05-04 01:00:00-04 select '2005-05-03 00:00:00 EST'::timestamp with time zone + '1 day'; ?column? ---------------------- 2005-05-04 01:00:00-04 Michael Glaesemann
2005-07-20 18:42:32 +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
if (tm2timestamp(tm, fsec, &tz, &timestamp) != 0)
Add 'day' field to INTERVAL so 1 day interval can be distinguished from 24 hours. This is very helpful for daylight savings time: select '2005-05-03 00:00:00 EST'::timestamp with time zone + '24 hours'; ?column? ---------------------- 2005-05-04 01:00:00-04 select '2005-05-03 00:00:00 EST'::timestamp with time zone + '1 day'; ?column? ---------------------- 2005-05-04 01:00:00-04 Michael Glaesemann
2005-07-20 18:42:32 +02:00
ereport(ERROR,
(errcode(ERRCODE_DATETIME_VALUE_OUT_OF_RANGE),
errmsg("timestamp out of range")));
}
timestamp += span->time;
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
if (!IS_VALID_TIMESTAMP(timestamp))
ereport(ERROR,
(errcode(ERRCODE_DATETIME_VALUE_OUT_OF_RANGE),
errmsg("timestamp out of range")));
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
result = timestamp;
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
}
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
PG_RETURN_TIMESTAMP(result);
}
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
Datum
Create crosstype comparison operators for date vs. timestamp and date vs. timestamptz. This allows use of indexes for expressions like datecol >= date 'today' - interval '1 month' which were formerly not indexable without casting the righthand side down from timestamp to date.
2004-02-14 21:16:18 +01:00
timestamptz_mi_interval(PG_FUNCTION_ARGS)
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
{
Add binary I/O routines for a bunch more datatypes. Still a few to go, but that was enough tedium for one day. Along the way, move the few support routines for types xid and cid into a more logical place.
2003-05-13 01:08:52 +02:00
TimestampTz timestamp = PG_GETARG_TIMESTAMPTZ(0);
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
Interval *span = PG_GETARG_INTERVAL_P(1);
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
Interval tspan;
pgindent run. Make it all clean.
2001-03-22 05:01:46 +01:00
tspan.month = -span->month;
Add 'day' field to INTERVAL so 1 day interval can be distinguished from 24 hours. This is very helpful for daylight savings time: select '2005-05-03 00:00:00 EST'::timestamp with time zone + '24 hours'; ?column? ---------------------- 2005-05-04 01:00:00-04 select '2005-05-03 00:00:00 EST'::timestamp with time zone + '1 day'; ?column? ---------------------- 2005-05-04 01:00:00-04 Michael Glaesemann
2005-07-20 18:42:32 +02:00
tspan.day = -span->day;
pgindent run. Make it all clean.
2001-03-22 05:01:46 +01:00
tspan.time = -span->time;
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
Create crosstype comparison operators for date vs. timestamp and date vs. timestamptz. This allows use of indexes for expressions like datecol >= date 'today' - interval '1 month' which were formerly not indexable without casting the righthand side down from timestamp to date.
2004-02-14 21:16:18 +01:00
return DirectFunctionCall2(timestamptz_pl_interval,
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
TimestampGetDatum(timestamp),
PointerGetDatum(&tspan));
}
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
Datum
interval_um(PG_FUNCTION_ARGS)
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
{
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
Interval *interval = PG_GETARG_INTERVAL_P(0);
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
Interval *result;
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
result = (Interval *) palloc(sizeof(Interval));
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
Remove unnecessary parentheses in assignments. Add spaces where needed. Reference time interval variables as tinterval.
2005-07-21 06:41:43 +02:00
result->time = -interval->time;
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
/* overflow check copied from int4um */
if (interval->time != 0 && SAMESIGN(result->time, interval->time))
ereport(ERROR,
(errcode(ERRCODE_DATETIME_VALUE_OUT_OF_RANGE),
errmsg("interval out of range")));
Remove unnecessary parentheses in assignments. Add spaces where needed. Reference time interval variables as tinterval.
2005-07-21 06:41:43 +02:00
result->day = -interval->day;
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
if (interval->day != 0 && SAMESIGN(result->day, interval->day))
ereport(ERROR,
(errcode(ERRCODE_DATETIME_VALUE_OUT_OF_RANGE),
errmsg("interval out of range")));
Remove unnecessary parentheses in assignments. Add spaces where needed. Reference time interval variables as tinterval.
2005-07-21 06:41:43 +02:00
result->month = -interval->month;
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
if (interval->month != 0 && SAMESIGN(result->month, interval->month))
ereport(ERROR,
(errcode(ERRCODE_DATETIME_VALUE_OUT_OF_RANGE),
errmsg("interval out of range")));
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
PG_RETURN_INTERVAL_P(result);
}
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
Datum
interval_smaller(PG_FUNCTION_ARGS)
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
{
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
Interval *interval1 = PG_GETARG_INTERVAL_P(0);
Interval *interval2 = PG_GETARG_INTERVAL_P(1);
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
Interval *result;
pgindent run.
2002-09-04 22:31:48 +02:00
Fix floating-point timestamp comparisons to not go nuts if NaN is encountered; per bug report from Christian van der Leeden 8/7/03. Also, adjust larger/smaller routines (MAX/MIN) to share code with comparisons for timestamp, interval, timetz.
2003-08-08 02:10:31 +02:00
/* use interval_cmp_internal to be sure this agrees with comparisons */
if (interval_cmp_internal(interval1, interval2) < 0)
result = interval1;
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
Fix floating-point timestamp comparisons to not go nuts if NaN is encountered; per bug report from Christian van der Leeden 8/7/03. Also, adjust larger/smaller routines (MAX/MIN) to share code with comparisons for timestamp, interval, timetz.
2003-08-08 02:10:31 +02:00
result = interval2;
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
PG_RETURN_INTERVAL_P(result);
}
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
Datum
interval_larger(PG_FUNCTION_ARGS)
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
{
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
Interval *interval1 = PG_GETARG_INTERVAL_P(0);
Interval *interval2 = PG_GETARG_INTERVAL_P(1);
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
Interval *result;
pgindent run.
2002-09-04 22:31:48 +02:00
Fix floating-point timestamp comparisons to not go nuts if NaN is encountered; per bug report from Christian van der Leeden 8/7/03. Also, adjust larger/smaller routines (MAX/MIN) to share code with comparisons for timestamp, interval, timetz.
2003-08-08 02:10:31 +02:00
if (interval_cmp_internal(interval1, interval2) > 0)
result = interval1;
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
Fix floating-point timestamp comparisons to not go nuts if NaN is encountered; per bug report from Christian van der Leeden 8/7/03. Also, adjust larger/smaller routines (MAX/MIN) to share code with comparisons for timestamp, interval, timetz.
2003-08-08 02:10:31 +02:00
result = interval2;
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
PG_RETURN_INTERVAL_P(result);
}
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
Datum
interval_pl(PG_FUNCTION_ARGS)
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
{
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
Interval *span1 = PG_GETARG_INTERVAL_P(0);
Interval *span2 = PG_GETARG_INTERVAL_P(1);
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
Interval *result;
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
result = (Interval *) palloc(sizeof(Interval));
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
Remove unnecessary parentheses in assignments. Add spaces where needed. Reference time interval variables as tinterval.
2005-07-21 06:41:43 +02:00
result->month = span1->month + span2->month;
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
/* overflow check copied from int4pl */
if (SAMESIGN(span1->month, span2->month) &&
!SAMESIGN(result->month, span1->month))
ereport(ERROR,
(errcode(ERRCODE_DATETIME_VALUE_OUT_OF_RANGE),
errmsg("interval out of range")));
Remove unnecessary parentheses in assignments. Add spaces where needed. Reference time interval variables as tinterval.
2005-07-21 06:41:43 +02:00
result->day = span1->day + span2->day;
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
if (SAMESIGN(span1->day, span2->day) &&
!SAMESIGN(result->day, span1->day))
ereport(ERROR,
(errcode(ERRCODE_DATETIME_VALUE_OUT_OF_RANGE),
errmsg("interval out of range")));
Remove unnecessary parentheses in assignments. Add spaces where needed. Reference time interval variables as tinterval.
2005-07-21 06:41:43 +02:00
result->time = span1->time + span2->time;
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
if (SAMESIGN(span1->time, span2->time) &&
!SAMESIGN(result->time, span1->time))
ereport(ERROR,
(errcode(ERRCODE_DATETIME_VALUE_OUT_OF_RANGE),
errmsg("interval out of range")));
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
PG_RETURN_INTERVAL_P(result);
}
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
Datum
interval_mi(PG_FUNCTION_ARGS)
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
{
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
Interval *span1 = PG_GETARG_INTERVAL_P(0);
Interval *span2 = PG_GETARG_INTERVAL_P(1);
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
Interval *result;
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
result = (Interval *) palloc(sizeof(Interval));
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
Remove unnecessary parentheses in assignments. Add spaces where needed. Reference time interval variables as tinterval.
2005-07-21 06:41:43 +02:00
result->month = span1->month - span2->month;
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
/* overflow check copied from int4mi */
if (!SAMESIGN(span1->month, span2->month) &&
!SAMESIGN(result->month, span1->month))
ereport(ERROR,
(errcode(ERRCODE_DATETIME_VALUE_OUT_OF_RANGE),
errmsg("interval out of range")));
Remove unnecessary parentheses in assignments. Add spaces where needed. Reference time interval variables as tinterval.
2005-07-21 06:41:43 +02:00
result->day = span1->day - span2->day;
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
if (!SAMESIGN(span1->day, span2->day) &&
!SAMESIGN(result->day, span1->day))
ereport(ERROR,
(errcode(ERRCODE_DATETIME_VALUE_OUT_OF_RANGE),
errmsg("interval out of range")));
Remove unnecessary parentheses in assignments. Add spaces where needed. Reference time interval variables as tinterval.
2005-07-21 06:41:43 +02:00
result->time = span1->time - span2->time;
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
if (!SAMESIGN(span1->time, span2->time) &&
!SAMESIGN(result->time, span1->time))
ereport(ERROR,
(errcode(ERRCODE_DATETIME_VALUE_OUT_OF_RANGE),
errmsg("interval out of range")));
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
PG_RETURN_INTERVAL_P(result);
}
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
interval_abs(): Add C comment about why there is no interval_abs(): it is unclear what value to return: http://archives.postgresql.org/pgsql-general/2009-10/msg01031.php http://archives.postgresql.org/pgsql-general/2009-11/msg00041.php
2009-11-10 19:41:24 +01:00
/*
pgindent run for 9.0
2010-02-26 03:01:40 +01:00
* There is no interval_abs(): it is unclear what value to return:
* http://archives.postgresql.org/pgsql-general/2009-10/msg01031.php
* http://archives.postgresql.org/pgsql-general/2009-11/msg00041.php
interval_abs(): Add C comment about why there is no interval_abs(): it is unclear what value to return: http://archives.postgresql.org/pgsql-general/2009-10/msg01031.php http://archives.postgresql.org/pgsql-general/2009-11/msg00041.php
2009-11-10 19:41:24 +01:00
*/
pgindent run for 9.0
2010-02-26 03:01:40 +01:00
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
Datum
interval_mul(PG_FUNCTION_ARGS)
Add transcendental math functions (sine, cosine, etc) Add a random number generator and seed setter (random(), SET SEED) Fix up the interval*float8 math to carry partial months into the time field. Add float8*interval so we have symmetry in the available math. Fix the parser and define.c to accept SQL92 types as field arguments. Fix the parser to accept SQL92 types for CREATE TYPE, etc. This is necessary to allow... Bit/varbit support in contrib/bit cleaned up to compile and load cleanly. Still needs some work before final release. Implement the "SOME" keyword as a synonym for "ANY" per SQL92. Implement ascii(text), ichar(int4), repeat(text,int4) to help support the ODBC driver. Enable the TRUNCATE() function mapping in the ODBC driver.
2000-04-07 15:40:45 +02:00
{
Fix rounding problem in interval_div by using rint(), and improve interval_mul function.
2005-07-24 06:37:07 +02:00
Interval *span = PG_GETARG_INTERVAL_P(0);
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
float8 factor = PG_GETARG_FLOAT8(1);
pgindent run for 8.2.
2006-10-04 02:30:14 +02:00
double month_remainder_days,
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
sec_remainder,
result_double;
pgindent run for 8.2.
2006-10-04 02:30:14 +02:00
int32 orig_month = span->month,
orig_day = span->day;
Back out pfrees for justify_hour function calls.
2005-08-25 05:53:22 +02:00
Interval *result;
pgindent run.
2002-09-04 22:31:48 +02:00
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
result = (Interval *) palloc(sizeof(Interval));
Add transcendental math functions (sine, cosine, etc) Add a random number generator and seed setter (random(), SET SEED) Fix up the interval*float8 math to carry partial months into the time field. Add float8*interval so we have symmetry in the available math. Fix the parser and define.c to accept SQL92 types as field arguments. Fix the parser to accept SQL92 types for CREATE TYPE, etc. This is necessary to allow... Bit/varbit support in contrib/bit cleaned up to compile and load cleanly. Still needs some work before final release. Implement the "SOME" keyword as a synonym for "ANY" per SQL92. Implement ascii(text), ichar(int4), repeat(text,int4) to help support the ODBC driver. Enable the TRUNCATE() function mapping in the ODBC driver.
2000-04-07 15:40:45 +02:00
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
result_double = span->month * factor;
Fix interval_mul() to not produce insane results. interval_mul() attempts to prevent its calculations from producing silly results, but it forgot that zero times infinity yields NaN in IEEE arithmetic. Hence, a case like '1 second'::interval * 'infinity'::float8 produced a NaN for the months product, which didn't trigger the range check, resulting in bogus and possibly platform-dependent output. This isn't terribly obvious to the naked eye because if you try that exact case, you get "interval out of range" which is what you expect --- but if you look closer, the error is coming from interval_out not interval_mul. interval_mul has allowed a bogus value into the system. Fix by adding isnan tests. Noted while testing Vitaly Burovoy's fix for infinity input to to_timestamp(). Given the lack of field complaints, I doubt this is worth a back-patch.
2016-03-29 23:21:12 +02:00
if (isnan(result_double) ||
result_double > INT_MAX || result_double < 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
ereport(ERROR,
(errcode(ERRCODE_DATETIME_VALUE_OUT_OF_RANGE),
errmsg("interval out of range")));
result->month = (int32) result_double;
result_double = span->day * factor;
Fix interval_mul() to not produce insane results. interval_mul() attempts to prevent its calculations from producing silly results, but it forgot that zero times infinity yields NaN in IEEE arithmetic. Hence, a case like '1 second'::interval * 'infinity'::float8 produced a NaN for the months product, which didn't trigger the range check, resulting in bogus and possibly platform-dependent output. This isn't terribly obvious to the naked eye because if you try that exact case, you get "interval out of range" which is what you expect --- but if you look closer, the error is coming from interval_out not interval_mul. interval_mul has allowed a bogus value into the system. Fix by adding isnan tests. Noted while testing Vitaly Burovoy's fix for infinity input to to_timestamp(). Given the lack of field complaints, I doubt this is worth a back-patch.
2016-03-29 23:21:12 +02:00
if (isnan(result_double) ||
result_double > INT_MAX || result_double < 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
ereport(ERROR,
(errcode(ERRCODE_DATETIME_VALUE_OUT_OF_RANGE),
errmsg("interval out of range")));
result->day = (int32) result_double;
Fix rounding problem in interval_div by using rint(), and improve interval_mul function.
2005-07-24 06:37:07 +02:00
Remove justify_hours call from interval_mul and interval_div, and make some small stylistic improvements in these functions. Also fix several places where TMODULO() was being used with wrong-sized quotient argument, creating a risk of overflow --- interval2tm was actually capable of going into an infinite loop because of this.
2005-10-25 19:13:07 +02:00
/*
Re-run pgindent, fixing a problem where comment lines after a blank comment line where output as too long, and update typedefs for /lib directory. Also fix case where identifiers were used as variable names in the backend, but as typedefs in ecpg (favor the backend for indenting). Backpatch to 8.1.X.
2005-11-22 19:17:34 +01:00
* The above correctly handles the whole-number part of the month and day
* products, but we have to do something with any fractional part
Various minor corrections of and improvements to comments. Aleksander Alekseev
2016-03-18 14:38:59 +01:00
* resulting when the factor is non-integral. We cascade the fractions
Remove justify_hours call from interval_mul and interval_div, and make some small stylistic improvements in these functions. Also fix several places where TMODULO() was being used with wrong-sized quotient argument, creating a risk of overflow --- interval2tm was actually capable of going into an infinite loop because of this.
2005-10-25 19:13:07 +02:00
* down to lower units using the conversion factors DAYS_PER_MONTH and
Re-run pgindent, fixing a problem where comment lines after a blank comment line where output as too long, and update typedefs for /lib directory. Also fix case where identifiers were used as variable names in the backend, but as typedefs in ecpg (favor the backend for indenting). Backpatch to 8.1.X.
2005-11-22 19:17:34 +01:00
* SECS_PER_DAY. Note we do NOT cascade up, since we are not forced to do
* so by the representation. The user can choose to cascade up later,
Remove justify_hours call from interval_mul and interval_div, and make some small stylistic improvements in these functions. Also fix several places where TMODULO() was being used with wrong-sized quotient argument, creating a risk of overflow --- interval2tm was actually capable of going into an infinite loop because of this.
2005-10-25 19:13:07 +02:00
* using justify_hours and/or justify_days.
*/
Fix rounding problem in interval_div by using rint(), and improve interval_mul function.
2005-07-24 06:37:07 +02:00
Properly round months into days and into seconds for interval multiplication/division queries like select '41 mon 10:00:00'::interval / 10 as "pos". Report from Michael Glaesemann
2006-09-03 05:34:04 +02:00
/*
pgindent run for 8.2.
2006-10-04 02:30:14 +02:00
* Fractional months full days into days.
Properly round months into days and into seconds for interval multiplication/division queries like select '41 mon 10:00:00'::interval / 10 as "pos". Report from Michael Glaesemann
2006-09-03 05:34:04 +02:00
*
Various minor corrections of and improvements to comments. Aleksander Alekseev
2016-03-18 14:38:59 +01:00
* Floating point calculation are inherently imprecise, so these
pgindent run for 8.2.
2006-10-04 02:30:14 +02:00
* calculations are crafted to produce the most reliable result possible.
* TSROUND() is needed to more accurately produce whole numbers where
* appropriate.
Properly round months into days and into seconds for interval multiplication/division queries like select '41 mon 10:00:00'::interval / 10 as "pos". Report from Michael Glaesemann
2006-09-03 05:34:04 +02:00
*/
Fix imprecision from interval rounding of multiplication/division. Bruce, Michael Glaesemann
2006-09-05 03:13:40 +02:00
month_remainder_days = (orig_month * factor - result->month) * DAYS_PER_MONTH;
month_remainder_days = TSROUND(month_remainder_days);
sec_remainder = (orig_day * factor - result->day +
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
month_remainder_days - (int) month_remainder_days) * SECS_PER_DAY;
Fix imprecision from interval rounding of multiplication/division. Bruce, Michael Glaesemann
2006-09-05 03:13:40 +02:00
sec_remainder = TSROUND(sec_remainder);
/*
pgindent run for 8.2.
2006-10-04 02:30:14 +02:00
* Might have 24:00:00 hours due to rounding, or >24 hours because of time
* cascade from months and days. It might still be >24 if the combination
* of cascade and the seconds factor operation itself.
Fix imprecision from interval rounding of multiplication/division. Bruce, Michael Glaesemann
2006-09-05 03:13:40 +02:00
*/
if (Abs(sec_remainder) >= SECS_PER_DAY)
{
pgindent run for 8.2.
2006-10-04 02:30:14 +02:00
result->day += (int) (sec_remainder / SECS_PER_DAY);
sec_remainder -= (int) (sec_remainder / SECS_PER_DAY) * SECS_PER_DAY;
Fix imprecision from interval rounding of multiplication/division. Bruce, Michael Glaesemann
2006-09-05 03:13:40 +02:00
}
Fix rounding problem in interval_div by using rint(), and improve interval_mul function.
2005-07-24 06:37:07 +02:00
Properly round months into days and into seconds for interval multiplication/division queries like select '41 mon 10:00:00'::interval / 10 as "pos". Report from Michael Glaesemann
2006-09-03 05:34:04 +02:00
/* cascade units down */
result->day += (int32) month_remainder_days;
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
result_double = rint(span->time * factor + sec_remainder * USECS_PER_SEC);
Fix integer-overflow edge case detection in interval_mul and pgbench. This patch adopts the overflow check logic introduced by commit cbdb8b4c0 into two more places. interval_mul() failed to notice if it computed a new microseconds value that was one more than INT64_MAX, and pgbench's double-to-int64 logic had the same sorts of edge-case problems that cbdb8b4c0 fixed in the core code. To make this easier to get right in future, put the guts of the checks into new macros in c.h, and add commentary about how to use the macros correctly. Back-patch to all supported branches, as we did with the previous fix. Yuya Watari Discussion: https://postgr.es/m/CAJ2pMkbkkFw2hb9Qb1Zj8d06EhWAQXFLy73St4qWv6aX=vqnjw@mail.gmail.com
2019-11-07 17:22:52 +01:00
if (isnan(result_double) || !FLOAT8_FITS_IN_INT64(result_double))
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
ereport(ERROR,
(errcode(ERRCODE_DATETIME_VALUE_OUT_OF_RANGE),
errmsg("interval out of range")));
result->time = (int64) result_double;
Add transcendental math functions (sine, cosine, etc) Add a random number generator and seed setter (random(), SET SEED) Fix up the interval*float8 math to carry partial months into the time field. Add float8*interval so we have symmetry in the available math. Fix the parser and define.c to accept SQL92 types as field arguments. Fix the parser to accept SQL92 types for CREATE TYPE, etc. This is necessary to allow... Bit/varbit support in contrib/bit cleaned up to compile and load cleanly. Still needs some work before final release. Implement the "SOME" keyword as a synonym for "ANY" per SQL92. Implement ascii(text), ichar(int4), repeat(text,int4) to help support the ODBC driver. Enable the TRUNCATE() function mapping in the ODBC driver.
2000-04-07 15:40:45 +02:00
Back out pfrees for justify_hour function calls.
2005-08-25 05:53:22 +02:00
PG_RETURN_INTERVAL_P(result);
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
}
Add transcendental math functions (sine, cosine, etc) Add a random number generator and seed setter (random(), SET SEED) Fix up the interval*float8 math to carry partial months into the time field. Add float8*interval so we have symmetry in the available math. Fix the parser and define.c to accept SQL92 types as field arguments. Fix the parser to accept SQL92 types for CREATE TYPE, etc. This is necessary to allow... Bit/varbit support in contrib/bit cleaned up to compile and load cleanly. Still needs some work before final release. Implement the "SOME" keyword as a synonym for "ANY" per SQL92. Implement ascii(text), ichar(int4), repeat(text,int4) to help support the ODBC driver. Enable the TRUNCATE() function mapping in the ODBC driver.
2000-04-07 15:40:45 +02:00
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
Datum
mul_d_interval(PG_FUNCTION_ARGS)
Add transcendental math functions (sine, cosine, etc) Add a random number generator and seed setter (random(), SET SEED) Fix up the interval*float8 math to carry partial months into the time field. Add float8*interval so we have symmetry in the available math. Fix the parser and define.c to accept SQL92 types as field arguments. Fix the parser to accept SQL92 types for CREATE TYPE, etc. This is necessary to allow... Bit/varbit support in contrib/bit cleaned up to compile and load cleanly. Still needs some work before final release. Implement the "SOME" keyword as a synonym for "ANY" per SQL92. Implement ascii(text), ichar(int4), repeat(text,int4) to help support the ODBC driver. Enable the TRUNCATE() function mapping in the ODBC driver.
2000-04-07 15:40:45 +02:00
{
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
/* Args are float8 and Interval *, but leave them as generic Datum */
Datum factor = PG_GETARG_DATUM(0);
Fix rounding problem in interval_div by using rint(), and improve interval_mul function.
2005-07-24 06:37:07 +02:00
Datum span = PG_GETARG_DATUM(1);
Add transcendental math functions (sine, cosine, etc) Add a random number generator and seed setter (random(), SET SEED) Fix up the interval*float8 math to carry partial months into the time field. Add float8*interval so we have symmetry in the available math. Fix the parser and define.c to accept SQL92 types as field arguments. Fix the parser to accept SQL92 types for CREATE TYPE, etc. This is necessary to allow... Bit/varbit support in contrib/bit cleaned up to compile and load cleanly. Still needs some work before final release. Implement the "SOME" keyword as a synonym for "ANY" per SQL92. Implement ascii(text), ichar(int4), repeat(text,int4) to help support the ODBC driver. Enable the TRUNCATE() function mapping in the ODBC driver.
2000-04-07 15:40:45 +02:00
Fix rounding problem in interval_div by using rint(), and improve interval_mul function.
2005-07-24 06:37:07 +02:00
return DirectFunctionCall2(interval_mul, span, factor);
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
}
Datum
interval_div(PG_FUNCTION_ARGS)
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +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
Interval *span = PG_GETARG_INTERVAL_P(0);
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
float8 factor = PG_GETARG_FLOAT8(1);
pgindent run for 8.2.
2006-10-04 02:30:14 +02:00
double month_remainder_days,
sec_remainder;
int32 orig_month = span->month,
orig_day = span->day;
Back out pfrees for justify_hour function calls.
2005-08-25 05:53:22 +02:00
Interval *result;
pgindent run for 8.2.
2006-10-04 02:30:14 +02:00
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
result = (Interval *) palloc(sizeof(Interval));
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
if (factor == 0.0)
Error message editing in utils/adt. Again thanks to Joe Conway for doing the bulk of the heavy lifting ...
2003-07-27 06:53:12 +02:00
ereport(ERROR,
(errcode(ERRCODE_DIVISION_BY_ZERO),
errmsg("division by zero")));
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
Properly round months into days and into seconds for interval multiplication/division queries like select '41 mon 10:00:00'::interval / 10 as "pos". Report from Michael Glaesemann
2006-09-03 05:34:04 +02:00
result->month = (int32) (span->month / factor);
result->day = (int32) (span->day / factor);
Add 'day' field to INTERVAL so 1 day interval can be distinguished from 24 hours. This is very helpful for daylight savings time: select '2005-05-03 00:00:00 EST'::timestamp with time zone + '24 hours'; ?column? ---------------------- 2005-05-04 01:00:00-04 select '2005-05-03 00:00:00 EST'::timestamp with time zone + '1 day'; ?column? ---------------------- 2005-05-04 01:00:00-04 Michael Glaesemann
2005-07-20 18:42:32 +02:00
Remove justify_hours call from interval_mul and interval_div, and make some small stylistic improvements in these functions. Also fix several places where TMODULO() was being used with wrong-sized quotient argument, creating a risk of overflow --- interval2tm was actually capable of going into an infinite loop because of this.
2005-10-25 19:13:07 +02:00
/*
pgindent run for 8.2.
2006-10-04 02:30:14 +02:00
* Fractional months full days into days. See comment in interval_mul().
Remove justify_hours call from interval_mul and interval_div, and make some small stylistic improvements in these functions. Also fix several places where TMODULO() was being used with wrong-sized quotient argument, creating a risk of overflow --- interval2tm was actually capable of going into an infinite loop because of this.
2005-10-25 19:13:07 +02:00
*/
Fix imprecision from interval rounding of multiplication/division. Bruce, Michael Glaesemann
2006-09-05 03:13:40 +02:00
month_remainder_days = (orig_month / factor - result->month) * DAYS_PER_MONTH;
month_remainder_days = TSROUND(month_remainder_days);
sec_remainder = (orig_day / factor - result->day +
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
month_remainder_days - (int) month_remainder_days) * SECS_PER_DAY;
Fix imprecision from interval rounding of multiplication/division. Bruce, Michael Glaesemann
2006-09-05 03:13:40 +02:00
sec_remainder = TSROUND(sec_remainder);
if (Abs(sec_remainder) >= SECS_PER_DAY)
{
pgindent run for 8.2.
2006-10-04 02:30:14 +02:00
result->day += (int) (sec_remainder / SECS_PER_DAY);
sec_remainder -= (int) (sec_remainder / SECS_PER_DAY) * SECS_PER_DAY;
Fix imprecision from interval rounding of multiplication/division. Bruce, Michael Glaesemann
2006-09-05 03:13:40 +02:00
}
Add 'day' field to INTERVAL so 1 day interval can be distinguished from 24 hours. This is very helpful for daylight savings time: select '2005-05-03 00:00:00 EST'::timestamp with time zone + '24 hours'; ?column? ---------------------- 2005-05-04 01:00:00-04 select '2005-05-03 00:00:00 EST'::timestamp with time zone + '1 day'; ?column? ---------------------- 2005-05-04 01:00:00-04 Michael Glaesemann
2005-07-20 18:42:32 +02:00
Properly round months into days and into seconds for interval multiplication/division queries like select '41 mon 10:00:00'::interval / 10 as "pos". Report from Michael Glaesemann
2006-09-03 05:34:04 +02:00
/* cascade units down */
Remove justify_hours call from interval_mul and interval_div, and make some small stylistic improvements in these functions. Also fix several places where TMODULO() was being used with wrong-sized quotient argument, creating a risk of overflow --- interval2tm was actually capable of going into an infinite loop because of this.
2005-10-25 19:13:07 +02:00
result->day += (int32) month_remainder_days;
Properly round months into days and into seconds for interval multiplication/division queries like select '41 mon 10:00:00'::interval / 10 as "pos". Report from Michael Glaesemann
2006-09-03 05:34:04 +02:00
result->time = rint(span->time / factor + sec_remainder * USECS_PER_SEC);
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
Back out pfrees for justify_hour function calls.
2005-08-25 05:53:22 +02:00
PG_RETURN_INTERVAL_P(result);
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
}
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
Support all SQL:2011 options for window frame clauses. This patch adds the ability to use "RANGE offset PRECEDING/FOLLOWING" frame boundaries in window functions. We'd punted on that back in the original patch to add window functions, because it was not clear how to do it in a reasonably data-type-extensible fashion. That problem is resolved here by adding the ability for btree operator classes to provide an "in_range" support function that defines how to add or subtract the RANGE offset value. Factoring it this way also allows the operator class to avoid overflow problems near the ends of the datatype's range, if it wishes to expend effort on that. (In the committed patch, the integer opclasses handle that issue, but it did not seem worth the trouble to avoid overflow failures for datetime types.) The patch includes in_range support for the integer_ops opfamily (int2/int4/int8) as well as the standard datetime types. Support for other numeric types has been requested, but that seems like suitable material for a follow-on patch. In addition, the patch adds GROUPS mode which counts the offset in ORDER-BY peer groups rather than rows, and it adds the frame_exclusion options specified by SQL:2011. As far as I can see, we are now fully up to spec on window framing options. Existing behaviors remain unchanged, except that I changed the errcode for a couple of existing error reports to meet the SQL spec's expectation that negative "offset" values should be reported as SQLSTATE 22013. Internally and in relevant parts of the documentation, we now consistently use the terminology "offset PRECEDING/FOLLOWING" rather than "value PRECEDING/FOLLOWING", since the term "value" is confusingly vague. Oliver Ford, reviewed and whacked around some by me Discussion: https://postgr.es/m/CAGMVOdu9sivPAxbNN0X+q19Sfv9edEPv=HibOJhB14TJv_RCQg@mail.gmail.com
2018-02-07 06:06:50 +01:00
/*
* in_range support functions for timestamps and intervals.
*
* Per SQL spec, we support these with interval as the offset type.
* The spec's restriction that the offset not be negative is a bit hard to
* decipher for intervals, but we choose to interpret it the same as our
* interval comparison operators would.
*/
Datum
in_range_timestamptz_interval(PG_FUNCTION_ARGS)
{
TimestampTz val = PG_GETARG_TIMESTAMPTZ(0);
TimestampTz base = PG_GETARG_TIMESTAMPTZ(1);
Interval *offset = PG_GETARG_INTERVAL_P(2);
bool sub = PG_GETARG_BOOL(3);
bool less = PG_GETARG_BOOL(4);
TimestampTz sum;
if (int128_compare(interval_cmp_value(offset), int64_to_int128(0)) < 0)
ereport(ERROR,
Make new error code name match SQL standard more closely Discussion: https://www.postgresql.org/message-id/dff3d555-bea4-ac24-29b2-29521b9d08e8%402ndquadrant.com
2018-06-11 17:15:28 +02:00
(errcode(ERRCODE_INVALID_PRECEDING_OR_FOLLOWING_SIZE),
Support all SQL:2011 options for window frame clauses. This patch adds the ability to use "RANGE offset PRECEDING/FOLLOWING" frame boundaries in window functions. We'd punted on that back in the original patch to add window functions, because it was not clear how to do it in a reasonably data-type-extensible fashion. That problem is resolved here by adding the ability for btree operator classes to provide an "in_range" support function that defines how to add or subtract the RANGE offset value. Factoring it this way also allows the operator class to avoid overflow problems near the ends of the datatype's range, if it wishes to expend effort on that. (In the committed patch, the integer opclasses handle that issue, but it did not seem worth the trouble to avoid overflow failures for datetime types.) The patch includes in_range support for the integer_ops opfamily (int2/int4/int8) as well as the standard datetime types. Support for other numeric types has been requested, but that seems like suitable material for a follow-on patch. In addition, the patch adds GROUPS mode which counts the offset in ORDER-BY peer groups rather than rows, and it adds the frame_exclusion options specified by SQL:2011. As far as I can see, we are now fully up to spec on window framing options. Existing behaviors remain unchanged, except that I changed the errcode for a couple of existing error reports to meet the SQL spec's expectation that negative "offset" values should be reported as SQLSTATE 22013. Internally and in relevant parts of the documentation, we now consistently use the terminology "offset PRECEDING/FOLLOWING" rather than "value PRECEDING/FOLLOWING", since the term "value" is confusingly vague. Oliver Ford, reviewed and whacked around some by me Discussion: https://postgr.es/m/CAGMVOdu9sivPAxbNN0X+q19Sfv9edEPv=HibOJhB14TJv_RCQg@mail.gmail.com
2018-02-07 06:06:50 +01:00
errmsg("invalid preceding or following size in window function")));
/* We don't currently bother to avoid overflow hazards here */
if (sub)
sum = DatumGetTimestampTz(DirectFunctionCall2(timestamptz_mi_interval,
TimestampTzGetDatum(base),
IntervalPGetDatum(offset)));
else
sum = DatumGetTimestampTz(DirectFunctionCall2(timestamptz_pl_interval,
TimestampTzGetDatum(base),
IntervalPGetDatum(offset)));
if (less)
PG_RETURN_BOOL(val <= sum);
else
PG_RETURN_BOOL(val >= sum);
}
Datum
in_range_timestamp_interval(PG_FUNCTION_ARGS)
{
Timestamp val = PG_GETARG_TIMESTAMP(0);
Timestamp base = PG_GETARG_TIMESTAMP(1);
Interval *offset = PG_GETARG_INTERVAL_P(2);
bool sub = PG_GETARG_BOOL(3);
bool less = PG_GETARG_BOOL(4);
Timestamp sum;
if (int128_compare(interval_cmp_value(offset), int64_to_int128(0)) < 0)
ereport(ERROR,
Make new error code name match SQL standard more closely Discussion: https://www.postgresql.org/message-id/dff3d555-bea4-ac24-29b2-29521b9d08e8%402ndquadrant.com
2018-06-11 17:15:28 +02:00
(errcode(ERRCODE_INVALID_PRECEDING_OR_FOLLOWING_SIZE),
Support all SQL:2011 options for window frame clauses. This patch adds the ability to use "RANGE offset PRECEDING/FOLLOWING" frame boundaries in window functions. We'd punted on that back in the original patch to add window functions, because it was not clear how to do it in a reasonably data-type-extensible fashion. That problem is resolved here by adding the ability for btree operator classes to provide an "in_range" support function that defines how to add or subtract the RANGE offset value. Factoring it this way also allows the operator class to avoid overflow problems near the ends of the datatype's range, if it wishes to expend effort on that. (In the committed patch, the integer opclasses handle that issue, but it did not seem worth the trouble to avoid overflow failures for datetime types.) The patch includes in_range support for the integer_ops opfamily (int2/int4/int8) as well as the standard datetime types. Support for other numeric types has been requested, but that seems like suitable material for a follow-on patch. In addition, the patch adds GROUPS mode which counts the offset in ORDER-BY peer groups rather than rows, and it adds the frame_exclusion options specified by SQL:2011. As far as I can see, we are now fully up to spec on window framing options. Existing behaviors remain unchanged, except that I changed the errcode for a couple of existing error reports to meet the SQL spec's expectation that negative "offset" values should be reported as SQLSTATE 22013. Internally and in relevant parts of the documentation, we now consistently use the terminology "offset PRECEDING/FOLLOWING" rather than "value PRECEDING/FOLLOWING", since the term "value" is confusingly vague. Oliver Ford, reviewed and whacked around some by me Discussion: https://postgr.es/m/CAGMVOdu9sivPAxbNN0X+q19Sfv9edEPv=HibOJhB14TJv_RCQg@mail.gmail.com
2018-02-07 06:06:50 +01:00
errmsg("invalid preceding or following size in window function")));
/* We don't currently bother to avoid overflow hazards here */
if (sub)
sum = DatumGetTimestamp(DirectFunctionCall2(timestamp_mi_interval,
TimestampGetDatum(base),
IntervalPGetDatum(offset)));
else
sum = DatumGetTimestamp(DirectFunctionCall2(timestamp_pl_interval,
TimestampGetDatum(base),
IntervalPGetDatum(offset)));
if (less)
PG_RETURN_BOOL(val <= sum);
else
PG_RETURN_BOOL(val >= sum);
}
Datum
in_range_interval_interval(PG_FUNCTION_ARGS)
{
Interval *val = PG_GETARG_INTERVAL_P(0);
Interval *base = PG_GETARG_INTERVAL_P(1);
Interval *offset = PG_GETARG_INTERVAL_P(2);
bool sub = PG_GETARG_BOOL(3);
bool less = PG_GETARG_BOOL(4);
Interval *sum;
if (int128_compare(interval_cmp_value(offset), int64_to_int128(0)) < 0)
ereport(ERROR,
Make new error code name match SQL standard more closely Discussion: https://www.postgresql.org/message-id/dff3d555-bea4-ac24-29b2-29521b9d08e8%402ndquadrant.com
2018-06-11 17:15:28 +02:00
(errcode(ERRCODE_INVALID_PRECEDING_OR_FOLLOWING_SIZE),
Support all SQL:2011 options for window frame clauses. This patch adds the ability to use "RANGE offset PRECEDING/FOLLOWING" frame boundaries in window functions. We'd punted on that back in the original patch to add window functions, because it was not clear how to do it in a reasonably data-type-extensible fashion. That problem is resolved here by adding the ability for btree operator classes to provide an "in_range" support function that defines how to add or subtract the RANGE offset value. Factoring it this way also allows the operator class to avoid overflow problems near the ends of the datatype's range, if it wishes to expend effort on that. (In the committed patch, the integer opclasses handle that issue, but it did not seem worth the trouble to avoid overflow failures for datetime types.) The patch includes in_range support for the integer_ops opfamily (int2/int4/int8) as well as the standard datetime types. Support for other numeric types has been requested, but that seems like suitable material for a follow-on patch. In addition, the patch adds GROUPS mode which counts the offset in ORDER-BY peer groups rather than rows, and it adds the frame_exclusion options specified by SQL:2011. As far as I can see, we are now fully up to spec on window framing options. Existing behaviors remain unchanged, except that I changed the errcode for a couple of existing error reports to meet the SQL spec's expectation that negative "offset" values should be reported as SQLSTATE 22013. Internally and in relevant parts of the documentation, we now consistently use the terminology "offset PRECEDING/FOLLOWING" rather than "value PRECEDING/FOLLOWING", since the term "value" is confusingly vague. Oliver Ford, reviewed and whacked around some by me Discussion: https://postgr.es/m/CAGMVOdu9sivPAxbNN0X+q19Sfv9edEPv=HibOJhB14TJv_RCQg@mail.gmail.com
2018-02-07 06:06:50 +01:00
errmsg("invalid preceding or following size in window function")));
/* We don't currently bother to avoid overflow hazards here */
if (sub)
sum = DatumGetIntervalP(DirectFunctionCall2(interval_mi,
IntervalPGetDatum(base),
IntervalPGetDatum(offset)));
else
sum = DatumGetIntervalP(DirectFunctionCall2(interval_pl,
IntervalPGetDatum(base),
IntervalPGetDatum(offset)));
if (less)
PG_RETURN_BOOL(interval_cmp_internal(val, sum) <= 0);
else
PG_RETURN_BOOL(interval_cmp_internal(val, sum) >= 0);
}
Revise aggregate functions per earlier discussions in pghackers. There's now only one transition value and transition function. NULL handling in aggregates is a lot cleaner. Also, use Numeric accumulators instead of integer accumulators for sum/avg on integer datatypes --- this avoids overflow at the cost of being a little slower. Implement VARIANCE() and STDDEV() aggregates in the standard backend. Also, enable new LIKE selectivity estimators by default. Unrelated change, but as long as I had to force initdb anyway...
2000-07-17 05:05:41 +02:00
/*
Provide moving-aggregate support for a bunch of numerical aggregates. First installment of the promised moving-aggregate support in built-in aggregates: count(), sum(), avg(), stddev() and variance() for assorted datatypes, though not for float4/float8. In passing, remove a 2001-vintage kluge in interval_accum(): interval array elements have been properly aligned since around 2003, but nobody remembered to take out this workaround. Also, fix a thinko in the opr_sanity tests for moving-aggregate catalog entries. David Rowley and Florian Pflug, reviewed by Dean Rasheed
2014-04-13 02:33:09 +02:00
* interval_accum, interval_accum_inv, and interval_avg implement the
* AVG(interval) aggregate.
Revise aggregate functions per earlier discussions in pghackers. There's now only one transition value and transition function. NULL handling in aggregates is a lot cleaner. Also, use Numeric accumulators instead of integer accumulators for sum/avg on integer datatypes --- this avoids overflow at the cost of being a little slower. Implement VARIANCE() and STDDEV() aggregates in the standard backend. Also, enable new LIKE selectivity estimators by default. Unrelated change, but as long as I had to force initdb anyway...
2000-07-17 05:05:41 +02:00
*
* The transition datatype for this aggregate is a 2-element array of
* intervals, where the first is the running sum and the second contains
* the number of values so far in its 'time' field. This is a bit ugly
* but it beats inventing a specialized datatype for the purpose.
*/
Datum
interval_accum(PG_FUNCTION_ARGS)
{
ArrayType *transarray = PG_GETARG_ARRAYTYPE_P(0);
Interval *newval = PG_GETARG_INTERVAL_P(1);
Datum *transdatums;
int ndatums;
Interval sumX,
N;
Interval *newsum;
ArrayType *result;
deconstruct_array(transarray,
Introduce macros for typalign and typstorage constants. Our usual practice for "poor man's enum" catalog columns is to define macros for the possible values and use those, not literal constants, in C code. But for some reason lost in the mists of time, this was never done for typalign/attalign or typstorage/attstorage. It's never too late to make it better though, so let's do that. The reason I got interested in this right now is the need to duplicate some uses of the TYPSTORAGE constants in an upcoming ALTER TYPE patch. But in general, this sort of change aids greppability and readability, so it's a good idea even without any specific motivation. I may have missed a few places that could be converted, and it's even more likely that pending patches will re-introduce some hard-coded references. But that's not fatal --- there's no expectation that we'd actually change any of these values. We can clean up stragglers over time. Discussion: https://postgr.es/m/16457.1583189537@sss.pgh.pa.us
2020-03-04 16:34:25 +01:00
INTERVALOID, sizeof(Interval), false, TYPALIGN_DOUBLE,
Make SQL arrays support null elements. This commit fixes the core array functionality, but I still need to make another pass looking at places that incidentally use arrays (such as ACL manipulation) to make sure they are null-safe. Contrib needs work too. I have not changed the behaviors that are still under discussion about array comparison and what to do with lower bounds.
2005-11-17 23:14:56 +01:00
&transdatums, NULL, &ndatums);
Revise aggregate functions per earlier discussions in pghackers. There's now only one transition value and transition function. NULL handling in aggregates is a lot cleaner. Also, use Numeric accumulators instead of integer accumulators for sum/avg on integer datatypes --- this avoids overflow at the cost of being a little slower. Implement VARIANCE() and STDDEV() aggregates in the standard backend. Also, enable new LIKE selectivity estimators by default. Unrelated change, but as long as I had to force initdb anyway...
2000-07-17 05:05:41 +02:00
if (ndatums != 2)
Error message editing in utils/adt. Again thanks to Joe Conway for doing the bulk of the heavy lifting ...
2003-07-27 06:53:12 +02:00
elog(ERROR, "expected 2-element interval array");
pgindent run. Make it all clean.
2001-03-22 05:01:46 +01:00
Provide moving-aggregate support for a bunch of numerical aggregates. First installment of the promised moving-aggregate support in built-in aggregates: count(), sum(), avg(), stddev() and variance() for assorted datatypes, though not for float4/float8. In passing, remove a 2001-vintage kluge in interval_accum(): interval array elements have been properly aligned since around 2003, but nobody remembered to take out this workaround. Also, fix a thinko in the opr_sanity tests for moving-aggregate catalog entries. David Rowley and Florian Pflug, reviewed by Dean Rasheed
2014-04-13 02:33:09 +02:00
sumX = *(DatumGetIntervalP(transdatums[0]));
N = *(DatumGetIntervalP(transdatums[1]));
Revise aggregate functions per earlier discussions in pghackers. There's now only one transition value and transition function. NULL handling in aggregates is a lot cleaner. Also, use Numeric accumulators instead of integer accumulators for sum/avg on integer datatypes --- this avoids overflow at the cost of being a little slower. Implement VARIANCE() and STDDEV() aggregates in the standard backend. Also, enable new LIKE selectivity estimators by default. Unrelated change, but as long as I had to force initdb anyway...
2000-07-17 05:05:41 +02:00
newsum = DatumGetIntervalP(DirectFunctionCall2(interval_pl,
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
IntervalPGetDatum(&sumX),
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
IntervalPGetDatum(newval)));
Revise aggregate functions per earlier discussions in pghackers. There's now only one transition value and transition function. NULL handling in aggregates is a lot cleaner. Also, use Numeric accumulators instead of integer accumulators for sum/avg on integer datatypes --- this avoids overflow at the cost of being a little slower. Implement VARIANCE() and STDDEV() aggregates in the standard backend. Also, enable new LIKE selectivity estimators by default. Unrelated change, but as long as I had to force initdb anyway...
2000-07-17 05:05:41 +02:00
N.time += 1;
transdatums[0] = IntervalPGetDatum(newsum);
transdatums[1] = IntervalPGetDatum(&N);
result = construct_array(transdatums, 2,
Introduce macros for typalign and typstorage constants. Our usual practice for "poor man's enum" catalog columns is to define macros for the possible values and use those, not literal constants, in C code. But for some reason lost in the mists of time, this was never done for typalign/attalign or typstorage/attstorage. It's never too late to make it better though, so let's do that. The reason I got interested in this right now is the need to duplicate some uses of the TYPSTORAGE constants in an upcoming ALTER TYPE patch. But in general, this sort of change aids greppability and readability, so it's a good idea even without any specific motivation. I may have missed a few places that could be converted, and it's even more likely that pending patches will re-introduce some hard-coded references. But that's not fatal --- there's no expectation that we'd actually change any of these values. We can clean up stragglers over time. Discussion: https://postgr.es/m/16457.1583189537@sss.pgh.pa.us
2020-03-04 16:34:25 +01:00
INTERVALOID, sizeof(Interval), false, TYPALIGN_DOUBLE);
Revise aggregate functions per earlier discussions in pghackers. There's now only one transition value and transition function. NULL handling in aggregates is a lot cleaner. Also, use Numeric accumulators instead of integer accumulators for sum/avg on integer datatypes --- this avoids overflow at the cost of being a little slower. Implement VARIANCE() and STDDEV() aggregates in the standard backend. Also, enable new LIKE selectivity estimators by default. Unrelated change, but as long as I had to force initdb anyway...
2000-07-17 05:05:41 +02:00
PG_RETURN_ARRAYTYPE_P(result);
}
Add parallel query support functions for assorted aggregates. This lets us use parallel aggregate for a variety of useful cases that didn't work before, like sum(int8), sum(numeric), several versions of avg(), and various other functions. Add some regression tests, as well, testing the general sanity of these and future catalog entries. David Rowley, reviewed by Tomas Vondra, with a few further changes by me.
2016-04-05 20:24:59 +02:00
Datum
interval_combine(PG_FUNCTION_ARGS)
{
ArrayType *transarray1 = PG_GETARG_ARRAYTYPE_P(0);
ArrayType *transarray2 = PG_GETARG_ARRAYTYPE_P(1);
Datum *transdatums1;
Datum *transdatums2;
int ndatums1;
int ndatums2;
Interval sum1,
N1;
Interval sum2,
N2;
Interval *newsum;
ArrayType *result;
deconstruct_array(transarray1,
Introduce macros for typalign and typstorage constants. Our usual practice for "poor man's enum" catalog columns is to define macros for the possible values and use those, not literal constants, in C code. But for some reason lost in the mists of time, this was never done for typalign/attalign or typstorage/attstorage. It's never too late to make it better though, so let's do that. The reason I got interested in this right now is the need to duplicate some uses of the TYPSTORAGE constants in an upcoming ALTER TYPE patch. But in general, this sort of change aids greppability and readability, so it's a good idea even without any specific motivation. I may have missed a few places that could be converted, and it's even more likely that pending patches will re-introduce some hard-coded references. But that's not fatal --- there's no expectation that we'd actually change any of these values. We can clean up stragglers over time. Discussion: https://postgr.es/m/16457.1583189537@sss.pgh.pa.us
2020-03-04 16:34:25 +01:00
INTERVALOID, sizeof(Interval), false, TYPALIGN_DOUBLE,
Add parallel query support functions for assorted aggregates. This lets us use parallel aggregate for a variety of useful cases that didn't work before, like sum(int8), sum(numeric), several versions of avg(), and various other functions. Add some regression tests, as well, testing the general sanity of these and future catalog entries. David Rowley, reviewed by Tomas Vondra, with a few further changes by me.
2016-04-05 20:24:59 +02:00
&transdatums1, NULL, &ndatums1);
if (ndatums1 != 2)
elog(ERROR, "expected 2-element interval array");
sum1 = *(DatumGetIntervalP(transdatums1[0]));
N1 = *(DatumGetIntervalP(transdatums1[1]));
deconstruct_array(transarray2,
Introduce macros for typalign and typstorage constants. Our usual practice for "poor man's enum" catalog columns is to define macros for the possible values and use those, not literal constants, in C code. But for some reason lost in the mists of time, this was never done for typalign/attalign or typstorage/attstorage. It's never too late to make it better though, so let's do that. The reason I got interested in this right now is the need to duplicate some uses of the TYPSTORAGE constants in an upcoming ALTER TYPE patch. But in general, this sort of change aids greppability and readability, so it's a good idea even without any specific motivation. I may have missed a few places that could be converted, and it's even more likely that pending patches will re-introduce some hard-coded references. But that's not fatal --- there's no expectation that we'd actually change any of these values. We can clean up stragglers over time. Discussion: https://postgr.es/m/16457.1583189537@sss.pgh.pa.us
2020-03-04 16:34:25 +01:00
INTERVALOID, sizeof(Interval), false, TYPALIGN_DOUBLE,
Add parallel query support functions for assorted aggregates. This lets us use parallel aggregate for a variety of useful cases that didn't work before, like sum(int8), sum(numeric), several versions of avg(), and various other functions. Add some regression tests, as well, testing the general sanity of these and future catalog entries. David Rowley, reviewed by Tomas Vondra, with a few further changes by me.
2016-04-05 20:24:59 +02:00
&transdatums2, NULL, &ndatums2);
if (ndatums2 != 2)
elog(ERROR, "expected 2-element interval array");
sum2 = *(DatumGetIntervalP(transdatums2[0]));
N2 = *(DatumGetIntervalP(transdatums2[1]));
newsum = DatumGetIntervalP(DirectFunctionCall2(interval_pl,
IntervalPGetDatum(&sum1),
IntervalPGetDatum(&sum2)));
N1.time += N2.time;
transdatums1[0] = IntervalPGetDatum(newsum);
transdatums1[1] = IntervalPGetDatum(&N1);
result = construct_array(transdatums1, 2,
Introduce macros for typalign and typstorage constants. Our usual practice for "poor man's enum" catalog columns is to define macros for the possible values and use those, not literal constants, in C code. But for some reason lost in the mists of time, this was never done for typalign/attalign or typstorage/attstorage. It's never too late to make it better though, so let's do that. The reason I got interested in this right now is the need to duplicate some uses of the TYPSTORAGE constants in an upcoming ALTER TYPE patch. But in general, this sort of change aids greppability and readability, so it's a good idea even without any specific motivation. I may have missed a few places that could be converted, and it's even more likely that pending patches will re-introduce some hard-coded references. But that's not fatal --- there's no expectation that we'd actually change any of these values. We can clean up stragglers over time. Discussion: https://postgr.es/m/16457.1583189537@sss.pgh.pa.us
2020-03-04 16:34:25 +01:00
INTERVALOID, sizeof(Interval), false, TYPALIGN_DOUBLE);
Add parallel query support functions for assorted aggregates. This lets us use parallel aggregate for a variety of useful cases that didn't work before, like sum(int8), sum(numeric), several versions of avg(), and various other functions. Add some regression tests, as well, testing the general sanity of these and future catalog entries. David Rowley, reviewed by Tomas Vondra, with a few further changes by me.
2016-04-05 20:24:59 +02:00
PG_RETURN_ARRAYTYPE_P(result);
}
Provide moving-aggregate support for a bunch of numerical aggregates. First installment of the promised moving-aggregate support in built-in aggregates: count(), sum(), avg(), stddev() and variance() for assorted datatypes, though not for float4/float8. In passing, remove a 2001-vintage kluge in interval_accum(): interval array elements have been properly aligned since around 2003, but nobody remembered to take out this workaround. Also, fix a thinko in the opr_sanity tests for moving-aggregate catalog entries. David Rowley and Florian Pflug, reviewed by Dean Rasheed
2014-04-13 02:33:09 +02:00
Datum
interval_accum_inv(PG_FUNCTION_ARGS)
{
ArrayType *transarray = PG_GETARG_ARRAYTYPE_P(0);
Interval *newval = PG_GETARG_INTERVAL_P(1);
Datum *transdatums;
int ndatums;
Interval sumX,
N;
Interval *newsum;
ArrayType *result;
deconstruct_array(transarray,
Introduce macros for typalign and typstorage constants. Our usual practice for "poor man's enum" catalog columns is to define macros for the possible values and use those, not literal constants, in C code. But for some reason lost in the mists of time, this was never done for typalign/attalign or typstorage/attstorage. It's never too late to make it better though, so let's do that. The reason I got interested in this right now is the need to duplicate some uses of the TYPSTORAGE constants in an upcoming ALTER TYPE patch. But in general, this sort of change aids greppability and readability, so it's a good idea even without any specific motivation. I may have missed a few places that could be converted, and it's even more likely that pending patches will re-introduce some hard-coded references. But that's not fatal --- there's no expectation that we'd actually change any of these values. We can clean up stragglers over time. Discussion: https://postgr.es/m/16457.1583189537@sss.pgh.pa.us
2020-03-04 16:34:25 +01:00
INTERVALOID, sizeof(Interval), false, TYPALIGN_DOUBLE,
Provide moving-aggregate support for a bunch of numerical aggregates. First installment of the promised moving-aggregate support in built-in aggregates: count(), sum(), avg(), stddev() and variance() for assorted datatypes, though not for float4/float8. In passing, remove a 2001-vintage kluge in interval_accum(): interval array elements have been properly aligned since around 2003, but nobody remembered to take out this workaround. Also, fix a thinko in the opr_sanity tests for moving-aggregate catalog entries. David Rowley and Florian Pflug, reviewed by Dean Rasheed
2014-04-13 02:33:09 +02:00
&transdatums, NULL, &ndatums);
if (ndatums != 2)
elog(ERROR, "expected 2-element interval array");
sumX = *(DatumGetIntervalP(transdatums[0]));
N = *(DatumGetIntervalP(transdatums[1]));
newsum = DatumGetIntervalP(DirectFunctionCall2(interval_mi,
IntervalPGetDatum(&sumX),
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
IntervalPGetDatum(newval)));
Provide moving-aggregate support for a bunch of numerical aggregates. First installment of the promised moving-aggregate support in built-in aggregates: count(), sum(), avg(), stddev() and variance() for assorted datatypes, though not for float4/float8. In passing, remove a 2001-vintage kluge in interval_accum(): interval array elements have been properly aligned since around 2003, but nobody remembered to take out this workaround. Also, fix a thinko in the opr_sanity tests for moving-aggregate catalog entries. David Rowley and Florian Pflug, reviewed by Dean Rasheed
2014-04-13 02:33:09 +02:00
N.time -= 1;
transdatums[0] = IntervalPGetDatum(newsum);
transdatums[1] = IntervalPGetDatum(&N);
result = construct_array(transdatums, 2,
Introduce macros for typalign and typstorage constants. Our usual practice for "poor man's enum" catalog columns is to define macros for the possible values and use those, not literal constants, in C code. But for some reason lost in the mists of time, this was never done for typalign/attalign or typstorage/attstorage. It's never too late to make it better though, so let's do that. The reason I got interested in this right now is the need to duplicate some uses of the TYPSTORAGE constants in an upcoming ALTER TYPE patch. But in general, this sort of change aids greppability and readability, so it's a good idea even without any specific motivation. I may have missed a few places that could be converted, and it's even more likely that pending patches will re-introduce some hard-coded references. But that's not fatal --- there's no expectation that we'd actually change any of these values. We can clean up stragglers over time. Discussion: https://postgr.es/m/16457.1583189537@sss.pgh.pa.us
2020-03-04 16:34:25 +01:00
INTERVALOID, sizeof(Interval), false, TYPALIGN_DOUBLE);
Provide moving-aggregate support for a bunch of numerical aggregates. First installment of the promised moving-aggregate support in built-in aggregates: count(), sum(), avg(), stddev() and variance() for assorted datatypes, though not for float4/float8. In passing, remove a 2001-vintage kluge in interval_accum(): interval array elements have been properly aligned since around 2003, but nobody remembered to take out this workaround. Also, fix a thinko in the opr_sanity tests for moving-aggregate catalog entries. David Rowley and Florian Pflug, reviewed by Dean Rasheed
2014-04-13 02:33:09 +02:00
PG_RETURN_ARRAYTYPE_P(result);
}
Revise aggregate functions per earlier discussions in pghackers. There's now only one transition value and transition function. NULL handling in aggregates is a lot cleaner. Also, use Numeric accumulators instead of integer accumulators for sum/avg on integer datatypes --- this avoids overflow at the cost of being a little slower. Implement VARIANCE() and STDDEV() aggregates in the standard backend. Also, enable new LIKE selectivity estimators by default. Unrelated change, but as long as I had to force initdb anyway...
2000-07-17 05:05:41 +02:00
Datum
interval_avg(PG_FUNCTION_ARGS)
{
ArrayType *transarray = PG_GETARG_ARRAYTYPE_P(0);
Datum *transdatums;
int ndatums;
Interval sumX,
N;
deconstruct_array(transarray,
Introduce macros for typalign and typstorage constants. Our usual practice for "poor man's enum" catalog columns is to define macros for the possible values and use those, not literal constants, in C code. But for some reason lost in the mists of time, this was never done for typalign/attalign or typstorage/attstorage. It's never too late to make it better though, so let's do that. The reason I got interested in this right now is the need to duplicate some uses of the TYPSTORAGE constants in an upcoming ALTER TYPE patch. But in general, this sort of change aids greppability and readability, so it's a good idea even without any specific motivation. I may have missed a few places that could be converted, and it's even more likely that pending patches will re-introduce some hard-coded references. But that's not fatal --- there's no expectation that we'd actually change any of these values. We can clean up stragglers over time. Discussion: https://postgr.es/m/16457.1583189537@sss.pgh.pa.us
2020-03-04 16:34:25 +01:00
INTERVALOID, sizeof(Interval), false, TYPALIGN_DOUBLE,
Make SQL arrays support null elements. This commit fixes the core array functionality, but I still need to make another pass looking at places that incidentally use arrays (such as ACL manipulation) to make sure they are null-safe. Contrib needs work too. I have not changed the behaviors that are still under discussion about array comparison and what to do with lower bounds.
2005-11-17 23:14:56 +01:00
&transdatums, NULL, &ndatums);
Revise aggregate functions per earlier discussions in pghackers. There's now only one transition value and transition function. NULL handling in aggregates is a lot cleaner. Also, use Numeric accumulators instead of integer accumulators for sum/avg on integer datatypes --- this avoids overflow at the cost of being a little slower. Implement VARIANCE() and STDDEV() aggregates in the standard backend. Also, enable new LIKE selectivity estimators by default. Unrelated change, but as long as I had to force initdb anyway...
2000-07-17 05:05:41 +02:00
if (ndatums != 2)
Error message editing in utils/adt. Again thanks to Joe Conway for doing the bulk of the heavy lifting ...
2003-07-27 06:53:12 +02:00
elog(ERROR, "expected 2-element interval array");
pgindent run. Make it all clean.
2001-03-22 05:01:46 +01:00
Provide moving-aggregate support for a bunch of numerical aggregates. First installment of the promised moving-aggregate support in built-in aggregates: count(), sum(), avg(), stddev() and variance() for assorted datatypes, though not for float4/float8. In passing, remove a 2001-vintage kluge in interval_accum(): interval array elements have been properly aligned since around 2003, but nobody remembered to take out this workaround. Also, fix a thinko in the opr_sanity tests for moving-aggregate catalog entries. David Rowley and Florian Pflug, reviewed by Dean Rasheed
2014-04-13 02:33:09 +02:00
sumX = *(DatumGetIntervalP(transdatums[0]));
N = *(DatumGetIntervalP(transdatums[1]));
Revise aggregate functions per earlier discussions in pghackers. There's now only one transition value and transition function. NULL handling in aggregates is a lot cleaner. Also, use Numeric accumulators instead of integer accumulators for sum/avg on integer datatypes --- this avoids overflow at the cost of being a little slower. Implement VARIANCE() and STDDEV() aggregates in the standard backend. Also, enable new LIKE selectivity estimators by default. Unrelated change, but as long as I had to force initdb anyway...
2000-07-17 05:05:41 +02:00
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 defines AVG of no values to be NULL */
Revise aggregate functions per earlier discussions in pghackers. There's now only one transition value and transition function. NULL handling in aggregates is a lot cleaner. Also, use Numeric accumulators instead of integer accumulators for sum/avg on integer datatypes --- this avoids overflow at the cost of being a little slower. Implement VARIANCE() and STDDEV() aggregates in the standard backend. Also, enable new LIKE selectivity estimators by default. Unrelated change, but as long as I had to force initdb anyway...
2000-07-17 05:05:41 +02:00
if (N.time == 0)
PG_RETURN_NULL();
return DirectFunctionCall2(interval_div,
IntervalPGetDatum(&sumX),
Provide moving-aggregate support for a bunch of numerical aggregates. First installment of the promised moving-aggregate support in built-in aggregates: count(), sum(), avg(), stddev() and variance() for assorted datatypes, though not for float4/float8. In passing, remove a 2001-vintage kluge in interval_accum(): interval array elements have been properly aligned since around 2003, but nobody remembered to take out this workaround. Also, fix a thinko in the opr_sanity tests for moving-aggregate catalog entries. David Rowley and Florian Pflug, reviewed by Dean Rasheed
2014-04-13 02:33:09 +02:00
Float8GetDatum((double) N.time));
Revise aggregate functions per earlier discussions in pghackers. There's now only one transition value and transition function. NULL handling in aggregates is a lot cleaner. Also, use Numeric accumulators instead of integer accumulators for sum/avg on integer datatypes --- this avoids overflow at the cost of being a little slower. Implement VARIANCE() and STDDEV() aggregates in the standard backend. Also, enable new LIKE selectivity estimators by default. Unrelated change, but as long as I had to force initdb anyway...
2000-07-17 05:05:41 +02:00
}
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
/* timestamp_age()
* Calculate time difference while retaining year/month fields.
* Note that this does not result in an accurate absolute time span
* since year and month are out of context once the arithmetic
* is done.
*/
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
Datum
timestamp_age(PG_FUNCTION_ARGS)
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
{
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
Timestamp dt1 = PG_GETARG_TIMESTAMP(0);
Timestamp dt2 = PG_GETARG_TIMESTAMP(1);
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
Interval *result;
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,
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
fsec1,
fsec2;
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
struct pg_tm tt,
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
*tm = &tt;
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
struct pg_tm tt1,
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
*tm1 = &tt1;
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
struct pg_tm tt2,
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
*tm2 = &tt2;
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
result = (Interval *) palloc(sizeof(Interval));
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
This patch makes it possible to use the full set of timezones when doing "AT TIME ZONE", and not just the shorlist previously available. For example: SELECT CURRENT_TIMESTAMP AT TIME ZONE 'Europe/London'; works fine now. It will also obey whatever DST rules were in effect at just that date, which the previous implementation did not. It also supports the AT TIME ZONE on the timetz datatype. The whole handling of DST is a bit bogus there, so I chose to make it use whatever DST rules are in effect at the time of executig the query. not sure if anybody is actuallyi *using* timetz though, it seems pretty unpredictable just because of this... Magnus Hagander
2005-06-15 02:34:11 +02:00
if (timestamp2tm(dt1, NULL, tm1, &fsec1, NULL, NULL) == 0 &&
timestamp2tm(dt2, NULL, tm2, &fsec2, NULL, NULL) == 0)
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
{
Properly adjust age() seconds to match the sign of the larger units. Patch from Tom.
2007-07-18 05:13:13 +02:00
/* form the symbolic difference */
fsec = fsec1 - fsec2;
Remove more extraneous parentheses in date/time functions.
2005-05-24 04:09:45 +02:00
tm->tm_sec = tm1->tm_sec - tm2->tm_sec;
tm->tm_min = tm1->tm_min - tm2->tm_min;
tm->tm_hour = tm1->tm_hour - tm2->tm_hour;
tm->tm_mday = tm1->tm_mday - tm2->tm_mday;
tm->tm_mon = tm1->tm_mon - tm2->tm_mon;
tm->tm_year = tm1->tm_year - tm2->tm_year;
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
/* flip sign if necessary... */
if (dt1 < dt2)
{
fsec = -fsec;
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;
}
Properly adjust age() seconds to match the sign of the larger units. Patch from Tom.
2007-07-18 05:13:13 +02:00
/* propagate any negative fields into the next higher field */
while (fsec < 0)
{
fsec += USECS_PER_SEC;
tm->tm_sec--;
}
Fix timestamptz_age() to do calculation in local timezone not GMT, per bug 1332.
2004-12-01 20:57:49 +01:00
while (tm->tm_sec < 0)
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
{
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
tm->tm_sec += SECS_PER_MINUTE;
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
tm->tm_min--;
}
Fix timestamptz_age() to do calculation in local timezone not GMT, per bug 1332.
2004-12-01 20:57:49 +01:00
while (tm->tm_min < 0)
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
{
Update DAYS_PER_MONTH comment. Add SECS_PER_YEAR and MINS_PER_HOUR macros.
2005-07-21 20:06:13 +02:00
tm->tm_min += MINS_PER_HOUR;
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
tm->tm_hour--;
}
Fix timestamptz_age() to do calculation in local timezone not GMT, per bug 1332.
2004-12-01 20:57:49 +01:00
while (tm->tm_hour < 0)
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
{
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
tm->tm_hour += HOURS_PER_DAY;
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
tm->tm_mday--;
}
Fix timestamptz_age() to do calculation in local timezone not GMT, per bug 1332.
2004-12-01 20:57:49 +01:00
while (tm->tm_mday < 0)
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
{
if (dt1 < dt2)
{
tm->tm_mday += day_tab[isleap(tm1->tm_year)][tm1->tm_mon - 1];
tm->tm_mon--;
}
else
{
tm->tm_mday += day_tab[isleap(tm2->tm_year)][tm2->tm_mon - 1];
tm->tm_mon--;
}
}
Fix timestamptz_age() to do calculation in local timezone not GMT, per bug 1332.
2004-12-01 20:57:49 +01:00
while (tm->tm_mon < 0)
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
{
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
tm->tm_mon += MONTHS_PER_YEAR;
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
tm->tm_year--;
}
/* recover sign if necessary... */
if (dt1 < dt2)
{
fsec = -fsec;
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;
}
if (tm2interval(tm, fsec, result) != 0)
Error message editing in utils/adt. Again thanks to Joe Conway for doing the bulk of the heavy lifting ...
2003-07-27 06:53:12 +02:00
ereport(ERROR,
(errcode(ERRCODE_DATETIME_VALUE_OUT_OF_RANGE),
errmsg("interval out of range")));
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
}
else
Error message editing in utils/adt. Again thanks to Joe Conway for doing the bulk of the heavy lifting ...
2003-07-27 06:53:12 +02:00
ereport(ERROR,
(errcode(ERRCODE_DATETIME_VALUE_OUT_OF_RANGE),
errmsg("timestamp out of range")));
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
PG_RETURN_INTERVAL_P(result);
}
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +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
/* timestamptz_age()
* Calculate time difference while retaining year/month fields.
* Note that this does not result in an accurate absolute time span
* since year and month are out of context once the arithmetic
* is done.
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
*/
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
Datum
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
timestamptz_age(PG_FUNCTION_ARGS)
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
{
Add binary I/O routines for a bunch more datatypes. Still a few to go, but that was enough tedium for one day. Along the way, move the few support routines for types xid and cid into a more logical place.
2003-05-13 01:08:52 +02:00
TimestampTz dt1 = PG_GETARG_TIMESTAMPTZ(0);
TimestampTz dt2 = PG_GETARG_TIMESTAMPTZ(1);
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
Interval *result;
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,
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
fsec1,
fsec2;
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
struct pg_tm tt,
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 = &tt;
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
struct pg_tm tt1,
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
*tm1 = &tt1;
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
struct pg_tm tt2,
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
*tm2 = &tt2;
Fix timestamptz_age() to do calculation in local timezone not GMT, per bug 1332.
2004-12-01 20:57:49 +01:00
int tz1;
int tz2;
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
result = (Interval *) palloc(sizeof(Interval));
Remove unused tzn arguments for timestamp2tm()
2012-03-15 20:13:35 +01:00
if (timestamp2tm(dt1, &tz1, tm1, &fsec1, NULL, NULL) == 0 &&
timestamp2tm(dt2, &tz2, tm2, &fsec2, NULL, NULL) == 0)
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
{
Properly adjust age() seconds to match the sign of the larger units. Patch from Tom.
2007-07-18 05:13:13 +02:00
/* form the symbolic difference */
Remove more extraneous parentheses in date/time functions.
2005-05-24 04:09:45 +02:00
fsec = fsec1 - fsec2;
tm->tm_sec = tm1->tm_sec - tm2->tm_sec;
tm->tm_min = tm1->tm_min - tm2->tm_min;
tm->tm_hour = tm1->tm_hour - tm2->tm_hour;
tm->tm_mday = tm1->tm_mday - tm2->tm_mday;
tm->tm_mon = tm1->tm_mon - tm2->tm_mon;
tm->tm_year = tm1->tm_year - tm2->tm_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
/* flip sign if necessary... */
if (dt1 < dt2)
{
fsec = -fsec;
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;
}
Properly adjust age() seconds to match the sign of the larger units. Patch from Tom.
2007-07-18 05:13:13 +02:00
/* propagate any negative fields into the next higher field */
while (fsec < 0)
{
fsec += USECS_PER_SEC;
tm->tm_sec--;
}
Fix timestamptz_age() to do calculation in local timezone not GMT, per bug 1332.
2004-12-01 20:57:49 +01:00
while (tm->tm_sec < 0)
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
{
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_sec += SECS_PER_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--;
}
Fix timestamptz_age() to do calculation in local timezone not GMT, per bug 1332.
2004-12-01 20:57:49 +01:00
while (tm->tm_min < 0)
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
{
Update DAYS_PER_MONTH comment. Add SECS_PER_YEAR and MINS_PER_HOUR macros.
2005-07-21 20:06:13 +02:00
tm->tm_min += MINS_PER_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--;
}
Fix timestamptz_age() to do calculation in local timezone not GMT, per bug 1332.
2004-12-01 20:57:49 +01:00
while (tm->tm_hour < 0)
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
{
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_hour += HOURS_PER_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--;
}
Fix timestamptz_age() to do calculation in local timezone not GMT, per bug 1332.
2004-12-01 20:57:49 +01:00
while (tm->tm_mday < 0)
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 (dt1 < dt2)
{
tm->tm_mday += day_tab[isleap(tm1->tm_year)][tm1->tm_mon - 1];
tm->tm_mon--;
}
else
{
tm->tm_mday += day_tab[isleap(tm2->tm_year)][tm2->tm_mon - 1];
tm->tm_mon--;
}
}
Fix timestamptz_age() to do calculation in local timezone not GMT, per bug 1332.
2004-12-01 20:57:49 +01:00
while (tm->tm_mon < 0)
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
{
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 += MONTHS_PER_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--;
}
Fix timestamptz_age() to do calculation in local timezone not GMT, per bug 1332.
2004-12-01 20:57:49 +01:00
/*
* Note: we deliberately ignore any difference between tz1 and tz2.
*/
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
/* recover sign if necessary... */
if (dt1 < dt2)
{
fsec = -fsec;
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;
}
if (tm2interval(tm, fsec, result) != 0)
Error message editing in utils/adt. Again thanks to Joe Conway for doing the bulk of the heavy lifting ...
2003-07-27 06:53:12 +02:00
ereport(ERROR,
(errcode(ERRCODE_DATETIME_VALUE_OUT_OF_RANGE),
errmsg("interval out of range")));
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
Error message editing in utils/adt. Again thanks to Joe Conway for doing the bulk of the heavy lifting ...
2003-07-27 06:53:12 +02:00
ereport(ERROR,
(errcode(ERRCODE_DATETIME_VALUE_OUT_OF_RANGE),
errmsg("timestamp out of range")));
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
PG_RETURN_INTERVAL_P(result);
}
/*----------------------------------------------------------
* Conversion operators.
*---------------------------------------------------------*/
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
/* timestamp_trunc()
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
* Truncate timestamp to specified units.
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
*/
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
Datum
timestamp_trunc(PG_FUNCTION_ARGS)
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
{
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
text *units = PG_GETARG_TEXT_PP(0);
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
Timestamp timestamp = PG_GETARG_TIMESTAMP(1);
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
Timestamp result;
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
int type,
val;
Solve the 'Turkish problem' with undesirable locale behavior for case conversion of basic ASCII letters. Remove all uses of strcasecmp and strncasecmp in favor of new functions pg_strcasecmp and pg_strncasecmp; remove most but not all direct uses of toupper and tolower in favor of pg_toupper and pg_tolower. These functions use the same notions of case folding already developed for identifier case conversion. I left the straight locale-based folding in place for situations where we are just manipulating user data and not trying to match it to built-in strings --- for example, the SQL upper() function is still locale dependent. Perhaps this will prove not to be what's wanted, but at the moment we can initdb and pass regression tests in Turkish locale.
2004-05-07 02:24:59 +02:00
char *lowunits;
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;
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
struct pg_tm tt,
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
*tm = &tt;
Solve the 'Turkish problem' with undesirable locale behavior for case conversion of basic ASCII letters. Remove all uses of strcasecmp and strncasecmp in favor of new functions pg_strcasecmp and pg_strncasecmp; remove most but not all direct uses of toupper and tolower in favor of pg_toupper and pg_tolower. These functions use the same notions of case folding already developed for identifier case conversion. I left the straight locale-based folding in place for situations where we are just manipulating user data and not trying to match it to built-in strings --- for example, the SQL upper() function is still locale dependent. Perhaps this will prove not to be what's wanted, but at the moment we can initdb and pass regression tests in Turkish locale.
2004-05-07 02:24:59 +02:00
if (TIMESTAMP_NOT_FINITE(timestamp))
PG_RETURN_TIMESTAMP(timestamp);
Error message editing in utils/adt. Again thanks to Joe Conway for doing the bulk of the heavy lifting ...
2003-07-27 06:53:12 +02:00
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
lowunits = downcase_truncate_identifier(VARDATA_ANY(units),
VARSIZE_ANY_EXHDR(units),
Solve the 'Turkish problem' with undesirable locale behavior for case conversion of basic ASCII letters. Remove all uses of strcasecmp and strncasecmp in favor of new functions pg_strcasecmp and pg_strncasecmp; remove most but not all direct uses of toupper and tolower in favor of pg_toupper and pg_tolower. These functions use the same notions of case folding already developed for identifier case conversion. I left the straight locale-based folding in place for situations where we are just manipulating user data and not trying to match it to built-in strings --- for example, the SQL upper() function is still locale dependent. Perhaps this will prove not to be what's wanted, but at the moment we can initdb and pass regression tests in Turkish locale.
2004-05-07 02:24:59 +02:00
false);
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
type = DecodeUnits(0, lowunits, &val);
Error message editing in utils/adt. Again thanks to Joe Conway for doing the bulk of the heavy lifting ...
2003-07-27 06:53:12 +02:00
if (type == UNITS)
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 unnecessary parentheses in assignments. Add spaces where needed. Reference time interval variables as tinterval.
2005-07-21 06:41:43 +02:00
if (timestamp2tm(timestamp, NULL, tm, &fsec, NULL, NULL) != 0)
Error message editing in utils/adt. Again thanks to Joe Conway for doing the bulk of the heavy lifting ...
2003-07-27 06:53:12 +02:00
ereport(ERROR,
(errcode(ERRCODE_DATETIME_VALUE_OUT_OF_RANGE),
errmsg("timestamp out of range")));
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
switch (val)
{
Per a brief conversation with Tom, I've created a patch for adding support for 'week' within the date_trunc function. Within the patch I added a couple of test cases and associated target output, and changed the documentation to add 'week' appropriately. Robert Creager
2004-03-05 03:41:14 +01:00
case DTK_WEEK:
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
{
int woy;
woy = date2isoweek(tm->tm_year, tm->tm_mon, tm->tm_mday);
/*
* If it is week 52/53 and the month is January, then the
* week must belong to the previous year. Also, some
* December dates belong to the next year.
*/
if (woy >= 52 && tm->tm_mon == 1)
--tm->tm_year;
if (woy <= 1 && tm->tm_mon == MONTHS_PER_YEAR)
++tm->tm_year;
isoweek2date(woy, &(tm->tm_year), &(tm->tm_mon), &(tm->tm_mday));
tm->tm_hour = 0;
tm->tm_min = 0;
tm->tm_sec = 0;
fsec = 0;
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 DTK_MILLENNIUM:
> After all that about numbering centuries and millenia correctly, > why does CVS tip still give me > > regression=# select extract(century from now()); > date_part > ----------- > 20 > (1 row) > [ ... looks in code ... ] > > Apparently it's because you fixed only timestamp_part, and not > timestamptz_part. I'm not too sure about what timestamp_trunc or > timestamptz_trunc should do, but they may be wrong as well. Sigh... as usual, what is not tested does not work:-( > Could we have a more complete patch? Please find a submission attached. I hope it really fixes all decade, century and millenium issues for extract and *_trunc functions on interval and other timestamp types. If someone could check that the results are reasonnable, it would be great. I indeed overlooked the fact that there were two functions. The patch fixes the code so that both variants agree. I added comments to interval extractions, because it relies on the C division to have a negative remainder: -7/10 = 0 and remains -7. As for *_trunc functions, I have chosen to put the first year of the century or millennium: -100, 1, 101... 1001 2001 etc. Indeed, I don't think it would make sense to put 2000 (last year of the 2nd millennium) for rounding all years of the third millenium. I also fixed the code so that all decades last 10 years and decade 199 means the 1990's. I have added some tests that are relevant to deal with tricky cases. The formula may be simplified, but all these cases must pass. Please keep them. Fabien Coelho
2004-08-20 05:45:14 +02:00
/* see comments in timestamptz_trunc */
if (tm->tm_year > 0)
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
tm->tm_year = ((tm->tm_year + 999) / 1000) * 1000 - 999;
> After all that about numbering centuries and millenia correctly, > why does CVS tip still give me > > regression=# select extract(century from now()); > date_part > ----------- > 20 > (1 row) > [ ... looks in code ... ] > > Apparently it's because you fixed only timestamp_part, and not > timestamptz_part. I'm not too sure about what timestamp_trunc or > timestamptz_trunc should do, but they may be wrong as well. Sigh... as usual, what is not tested does not work:-( > Could we have a more complete patch? Please find a submission attached. I hope it really fixes all decade, century and millenium issues for extract and *_trunc functions on interval and other timestamp types. If someone could check that the results are reasonnable, it would be great. I indeed overlooked the fact that there were two functions. The patch fixes the code so that both variants agree. I added comments to interval extractions, because it relies on the C division to have a negative remainder: -7/10 = 0 and remains -7. As for *_trunc functions, I have chosen to put the first year of the century or millennium: -100, 1, 101... 1001 2001 etc. Indeed, I don't think it would make sense to put 2000 (last year of the 2nd millennium) for rounding all years of the third millenium. I also fixed the code so that all decades last 10 years and decade 199 means the 1990's. I have added some tests that are relevant to deal with tricky cases. The formula may be simplified, but all these cases must pass. Please keep them. Fabien Coelho
2004-08-20 05:45:14 +02:00
else
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
tm->tm_year = -((999 - (tm->tm_year - 1)) / 1000) * 1000 + 1;
Add -Wimplicit-fallthrough to CFLAGS and CXXFLAGS Use it at level 4, a bit more restrictive than the default level, and tweak our commanding comments to FALLTHROUGH. (However, leave zic.c alone, since it's external code; to avoid the warnings that would appear there, change CFLAGS for that file in the Makefile.) Author: Julien Rouhaud <rjuju123@gmail.com> Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Tom Lane <tgl@sss.pgh.pa.us> Discussion: https://postgr.es/m/20200412081825.qyo5vwwco3fv4gdo@nol Discussion: https://postgr.es/m/flat/E1fDenm-0000C8-IJ@gemulon.postgresql.org
2020-05-12 22:01:52 +02:00
/* FALLTHROUGH */
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 DTK_CENTURY:
> After all that about numbering centuries and millenia correctly, > why does CVS tip still give me > > regression=# select extract(century from now()); > date_part > ----------- > 20 > (1 row) > [ ... looks in code ... ] > > Apparently it's because you fixed only timestamp_part, and not > timestamptz_part. I'm not too sure about what timestamp_trunc or > timestamptz_trunc should do, but they may be wrong as well. Sigh... as usual, what is not tested does not work:-( > Could we have a more complete patch? Please find a submission attached. I hope it really fixes all decade, century and millenium issues for extract and *_trunc functions on interval and other timestamp types. If someone could check that the results are reasonnable, it would be great. I indeed overlooked the fact that there were two functions. The patch fixes the code so that both variants agree. I added comments to interval extractions, because it relies on the C division to have a negative remainder: -7/10 = 0 and remains -7. As for *_trunc functions, I have chosen to put the first year of the century or millennium: -100, 1, 101... 1001 2001 etc. Indeed, I don't think it would make sense to put 2000 (last year of the 2nd millennium) for rounding all years of the third millenium. I also fixed the code so that all decades last 10 years and decade 199 means the 1990's. I have added some tests that are relevant to deal with tricky cases. The formula may be simplified, but all these cases must pass. Please keep them. Fabien Coelho
2004-08-20 05:45:14 +02:00
/* see comments in timestamptz_trunc */
if (tm->tm_year > 0)
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
tm->tm_year = ((tm->tm_year + 99) / 100) * 100 - 99;
> After all that about numbering centuries and millenia correctly, > why does CVS tip still give me > > regression=# select extract(century from now()); > date_part > ----------- > 20 > (1 row) > [ ... looks in code ... ] > > Apparently it's because you fixed only timestamp_part, and not > timestamptz_part. I'm not too sure about what timestamp_trunc or > timestamptz_trunc should do, but they may be wrong as well. Sigh... as usual, what is not tested does not work:-( > Could we have a more complete patch? Please find a submission attached. I hope it really fixes all decade, century and millenium issues for extract and *_trunc functions on interval and other timestamp types. If someone could check that the results are reasonnable, it would be great. I indeed overlooked the fact that there were two functions. The patch fixes the code so that both variants agree. I added comments to interval extractions, because it relies on the C division to have a negative remainder: -7/10 = 0 and remains -7. As for *_trunc functions, I have chosen to put the first year of the century or millennium: -100, 1, 101... 1001 2001 etc. Indeed, I don't think it would make sense to put 2000 (last year of the 2nd millennium) for rounding all years of the third millenium. I also fixed the code so that all decades last 10 years and decade 199 means the 1990's. I have added some tests that are relevant to deal with tricky cases. The formula may be simplified, but all these cases must pass. Please keep them. Fabien Coelho
2004-08-20 05:45:14 +02:00
else
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
tm->tm_year = -((99 - (tm->tm_year - 1)) / 100) * 100 + 1;
Add -Wimplicit-fallthrough to CFLAGS and CXXFLAGS Use it at level 4, a bit more restrictive than the default level, and tweak our commanding comments to FALLTHROUGH. (However, leave zic.c alone, since it's external code; to avoid the warnings that would appear there, change CFLAGS for that file in the Makefile.) Author: Julien Rouhaud <rjuju123@gmail.com> Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Tom Lane <tgl@sss.pgh.pa.us> Discussion: https://postgr.es/m/20200412081825.qyo5vwwco3fv4gdo@nol Discussion: https://postgr.es/m/flat/E1fDenm-0000C8-IJ@gemulon.postgresql.org
2020-05-12 22:01:52 +02:00
/* FALLTHROUGH */
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 DTK_DECADE:
> After all that about numbering centuries and millenia correctly, > why does CVS tip still give me > > regression=# select extract(century from now()); > date_part > ----------- > 20 > (1 row) > [ ... looks in code ... ] > > Apparently it's because you fixed only timestamp_part, and not > timestamptz_part. I'm not too sure about what timestamp_trunc or > timestamptz_trunc should do, but they may be wrong as well. Sigh... as usual, what is not tested does not work:-( > Could we have a more complete patch? Please find a submission attached. I hope it really fixes all decade, century and millenium issues for extract and *_trunc functions on interval and other timestamp types. If someone could check that the results are reasonnable, it would be great. I indeed overlooked the fact that there were two functions. The patch fixes the code so that both variants agree. I added comments to interval extractions, because it relies on the C division to have a negative remainder: -7/10 = 0 and remains -7. As for *_trunc functions, I have chosen to put the first year of the century or millennium: -100, 1, 101... 1001 2001 etc. Indeed, I don't think it would make sense to put 2000 (last year of the 2nd millennium) for rounding all years of the third millenium. I also fixed the code so that all decades last 10 years and decade 199 means the 1990's. I have added some tests that are relevant to deal with tricky cases. The formula may be simplified, but all these cases must pass. Please keep them. Fabien Coelho
2004-08-20 05:45:14 +02:00
/* see comments in timestamptz_trunc */
if (val != DTK_MILLENNIUM && val != DTK_CENTURY)
{
if (tm->tm_year > 0)
tm->tm_year = (tm->tm_year / 10) * 10;
else
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
tm->tm_year = -((8 - (tm->tm_year - 1)) / 10) * 10;
> After all that about numbering centuries and millenia correctly, > why does CVS tip still give me > > regression=# select extract(century from now()); > date_part > ----------- > 20 > (1 row) > [ ... looks in code ... ] > > Apparently it's because you fixed only timestamp_part, and not > timestamptz_part. I'm not too sure about what timestamp_trunc or > timestamptz_trunc should do, but they may be wrong as well. Sigh... as usual, what is not tested does not work:-( > Could we have a more complete patch? Please find a submission attached. I hope it really fixes all decade, century and millenium issues for extract and *_trunc functions on interval and other timestamp types. If someone could check that the results are reasonnable, it would be great. I indeed overlooked the fact that there were two functions. The patch fixes the code so that both variants agree. I added comments to interval extractions, because it relies on the C division to have a negative remainder: -7/10 = 0 and remains -7. As for *_trunc functions, I have chosen to put the first year of the century or millennium: -100, 1, 101... 1001 2001 etc. Indeed, I don't think it would make sense to put 2000 (last year of the 2nd millennium) for rounding all years of the third millenium. I also fixed the code so that all decades last 10 years and decade 199 means the 1990's. I have added some tests that are relevant to deal with tricky cases. The formula may be simplified, but all these cases must pass. Please keep them. Fabien Coelho
2004-08-20 05:45:14 +02:00
}
Add -Wimplicit-fallthrough to CFLAGS and CXXFLAGS Use it at level 4, a bit more restrictive than the default level, and tweak our commanding comments to FALLTHROUGH. (However, leave zic.c alone, since it's external code; to avoid the warnings that would appear there, change CFLAGS for that file in the Makefile.) Author: Julien Rouhaud <rjuju123@gmail.com> Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Tom Lane <tgl@sss.pgh.pa.us> Discussion: https://postgr.es/m/20200412081825.qyo5vwwco3fv4gdo@nol Discussion: https://postgr.es/m/flat/E1fDenm-0000C8-IJ@gemulon.postgresql.org
2020-05-12 22:01:52 +02:00
/* FALLTHROUGH */
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 DTK_YEAR:
tm->tm_mon = 1;
Add -Wimplicit-fallthrough to CFLAGS and CXXFLAGS Use it at level 4, a bit more restrictive than the default level, and tweak our commanding comments to FALLTHROUGH. (However, leave zic.c alone, since it's external code; to avoid the warnings that would appear there, change CFLAGS for that file in the Makefile.) Author: Julien Rouhaud <rjuju123@gmail.com> Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Tom Lane <tgl@sss.pgh.pa.us> Discussion: https://postgr.es/m/20200412081825.qyo5vwwco3fv4gdo@nol Discussion: https://postgr.es/m/flat/E1fDenm-0000C8-IJ@gemulon.postgresql.org
2020-05-12 22:01:52 +02:00
/* FALLTHROUGH */
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 DTK_QUARTER:
I corecting date_trunc('quarter',...) and friends because orig version doing '2003-07-30' -> '2003-04-01', '2003-11-30' ->'2003-07-01' B?jthe Zolt?n
2003-07-26 17:17:36 +02:00
tm->tm_mon = (3 * ((tm->tm_mon - 1) / 3)) + 1;
Add -Wimplicit-fallthrough to CFLAGS and CXXFLAGS Use it at level 4, a bit more restrictive than the default level, and tweak our commanding comments to FALLTHROUGH. (However, leave zic.c alone, since it's external code; to avoid the warnings that would appear there, change CFLAGS for that file in the Makefile.) Author: Julien Rouhaud <rjuju123@gmail.com> Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Tom Lane <tgl@sss.pgh.pa.us> Discussion: https://postgr.es/m/20200412081825.qyo5vwwco3fv4gdo@nol Discussion: https://postgr.es/m/flat/E1fDenm-0000C8-IJ@gemulon.postgresql.org
2020-05-12 22:01:52 +02:00
/* FALLTHROUGH */
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 DTK_MONTH:
tm->tm_mday = 1;
Add -Wimplicit-fallthrough to CFLAGS and CXXFLAGS Use it at level 4, a bit more restrictive than the default level, and tweak our commanding comments to FALLTHROUGH. (However, leave zic.c alone, since it's external code; to avoid the warnings that would appear there, change CFLAGS for that file in the Makefile.) Author: Julien Rouhaud <rjuju123@gmail.com> Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Tom Lane <tgl@sss.pgh.pa.us> Discussion: https://postgr.es/m/20200412081825.qyo5vwwco3fv4gdo@nol Discussion: https://postgr.es/m/flat/E1fDenm-0000C8-IJ@gemulon.postgresql.org
2020-05-12 22:01:52 +02:00
/* FALLTHROUGH */
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 DTK_DAY:
tm->tm_hour = 0;
Add -Wimplicit-fallthrough to CFLAGS and CXXFLAGS Use it at level 4, a bit more restrictive than the default level, and tweak our commanding comments to FALLTHROUGH. (However, leave zic.c alone, since it's external code; to avoid the warnings that would appear there, change CFLAGS for that file in the Makefile.) Author: Julien Rouhaud <rjuju123@gmail.com> Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Tom Lane <tgl@sss.pgh.pa.us> Discussion: https://postgr.es/m/20200412081825.qyo5vwwco3fv4gdo@nol Discussion: https://postgr.es/m/flat/E1fDenm-0000C8-IJ@gemulon.postgresql.org
2020-05-12 22:01:52 +02:00
/* FALLTHROUGH */
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 DTK_HOUR:
tm->tm_min = 0;
Add -Wimplicit-fallthrough to CFLAGS and CXXFLAGS Use it at level 4, a bit more restrictive than the default level, and tweak our commanding comments to FALLTHROUGH. (However, leave zic.c alone, since it's external code; to avoid the warnings that would appear there, change CFLAGS for that file in the Makefile.) Author: Julien Rouhaud <rjuju123@gmail.com> Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Tom Lane <tgl@sss.pgh.pa.us> Discussion: https://postgr.es/m/20200412081825.qyo5vwwco3fv4gdo@nol Discussion: https://postgr.es/m/flat/E1fDenm-0000C8-IJ@gemulon.postgresql.org
2020-05-12 22:01:52 +02:00
/* FALLTHROUGH */
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 DTK_MINUTE:
tm->tm_sec = 0;
Add -Wimplicit-fallthrough to CFLAGS and CXXFLAGS Use it at level 4, a bit more restrictive than the default level, and tweak our commanding comments to FALLTHROUGH. (However, leave zic.c alone, since it's external code; to avoid the warnings that would appear there, change CFLAGS for that file in the Makefile.) Author: Julien Rouhaud <rjuju123@gmail.com> Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Tom Lane <tgl@sss.pgh.pa.us> Discussion: https://postgr.es/m/20200412081825.qyo5vwwco3fv4gdo@nol Discussion: https://postgr.es/m/flat/E1fDenm-0000C8-IJ@gemulon.postgresql.org
2020-05-12 22:01:52 +02:00
/* FALLTHROUGH */
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 DTK_SECOND:
fsec = 0;
break;
case DTK_MILLISEC:
Remove unnecessary parentheses in datetime/timestamp code.
2005-05-23 19:13:14 +02:00
fsec = (fsec / 1000) * 1000;
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;
case DTK_MICROSEC:
break;
default:
Error message editing in utils/adt. Again thanks to Joe Conway for doing the bulk of the heavy lifting ...
2003-07-27 06:53:12 +02:00
ereport(ERROR,
(errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
errmsg("timestamp units \"%s\" not supported",
lowunits)));
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
result = 0;
}
if (tm2timestamp(tm, fsec, NULL, &result) != 0)
Error message editing in utils/adt. Again thanks to Joe Conway for doing the bulk of the heavy lifting ...
2003-07-27 06:53:12 +02:00
ereport(ERROR,
(errcode(ERRCODE_DATETIME_VALUE_OUT_OF_RANGE),
errmsg("timestamp out of range")));
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
}
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
else
{
Error message editing in utils/adt. Again thanks to Joe Conway for doing the bulk of the heavy lifting ...
2003-07-27 06:53:12 +02:00
ereport(ERROR,
(errcode(ERRCODE_INVALID_PARAMETER_VALUE),
errmsg("timestamp units \"%s\" not recognized",
lowunits)));
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
result = 0;
}
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +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
PG_RETURN_TIMESTAMP(result);
}
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
Add a timezone-specific variant of date_trunc(). date_trunc(field, timestamptz, zone_name) performs truncation using the named time zone as reference, rather than working in the session time zone as is the default behavior. It's equivalent to date_trunc(field, timestamptz at time zone zone_name) at time zone zone_name but it's faster, easier to type, and arguably easier to understand. Vik Fearing and Tom Lane Discussion: https://postgr.es/m/6249ffc4-2b22-4c1b-4e7d-7af84fedd7c6@2ndquadrant.com
2018-11-14 21:41:07 +01:00
/*
* Common code for timestamptz_trunc() and timestamptz_trunc_zone().
*
* tzp identifies the zone to truncate with respect to. We assume
* infinite timestamps have already been rejected.
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
*/
Add a timezone-specific variant of date_trunc(). date_trunc(field, timestamptz, zone_name) performs truncation using the named time zone as reference, rather than working in the session time zone as is the default behavior. It's equivalent to date_trunc(field, timestamptz at time zone zone_name) at time zone zone_name but it's faster, easier to type, and arguably easier to understand. Vik Fearing and Tom Lane Discussion: https://postgr.es/m/6249ffc4-2b22-4c1b-4e7d-7af84fedd7c6@2ndquadrant.com
2018-11-14 21:41:07 +01:00
static TimestampTz
timestamptz_trunc_internal(text *units, TimestampTz timestamp, pg_tz *tzp)
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
TimestampTz result;
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
int tz;
int type,
val;
timestamptz_trunc() should only recalculate the timezone when truncating to DAY precision or coarser; leave the timezone alone when precision is HOUR or less. This avoids surprises for inputs near a DST transition time, as per example from Matthew Gabeler-Lee. (The only reason we recalculate at all is so that outputs that are supposed to represent days will come out as local midnight, and that's not relevant for sub-day precision.)
2004-11-01 23:00:30 +01:00
bool redotz = false;
Solve the 'Turkish problem' with undesirable locale behavior for case conversion of basic ASCII letters. Remove all uses of strcasecmp and strncasecmp in favor of new functions pg_strcasecmp and pg_strncasecmp; remove most but not all direct uses of toupper and tolower in favor of pg_toupper and pg_tolower. These functions use the same notions of case folding already developed for identifier case conversion. I left the straight locale-based folding in place for situations where we are just manipulating user data and not trying to match it to built-in strings --- for example, the SQL upper() function is still locale dependent. Perhaps this will prove not to be what's wanted, but at the moment we can initdb and pass regression tests in Turkish locale.
2004-05-07 02:24:59 +02:00
char *lowunits;
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;
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
struct pg_tm tt,
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 = &tt;
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
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
lowunits = downcase_truncate_identifier(VARDATA_ANY(units),
VARSIZE_ANY_EXHDR(units),
Solve the 'Turkish problem' with undesirable locale behavior for case conversion of basic ASCII letters. Remove all uses of strcasecmp and strncasecmp in favor of new functions pg_strcasecmp and pg_strncasecmp; remove most but not all direct uses of toupper and tolower in favor of pg_toupper and pg_tolower. These functions use the same notions of case folding already developed for identifier case conversion. I left the straight locale-based folding in place for situations where we are just manipulating user data and not trying to match it to built-in strings --- for example, the SQL upper() function is still locale dependent. Perhaps this will prove not to be what's wanted, but at the moment we can initdb and pass regression tests in Turkish locale.
2004-05-07 02:24:59 +02:00
false);
type = DecodeUnits(0, lowunits, &val);
Error message editing in utils/adt. Again thanks to Joe Conway for doing the bulk of the heavy lifting ...
2003-07-27 06:53:12 +02:00
if (type == UNITS)
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
{
Add a timezone-specific variant of date_trunc(). date_trunc(field, timestamptz, zone_name) performs truncation using the named time zone as reference, rather than working in the session time zone as is the default behavior. It's equivalent to date_trunc(field, timestamptz at time zone zone_name) at time zone zone_name but it's faster, easier to type, and arguably easier to understand. Vik Fearing and Tom Lane Discussion: https://postgr.es/m/6249ffc4-2b22-4c1b-4e7d-7af84fedd7c6@2ndquadrant.com
2018-11-14 21:41:07 +01:00
if (timestamp2tm(timestamp, &tz, tm, &fsec, NULL, tzp) != 0)
Error message editing in utils/adt. Again thanks to Joe Conway for doing the bulk of the heavy lifting ...
2003-07-27 06:53:12 +02:00
ereport(ERROR,
(errcode(ERRCODE_DATETIME_VALUE_OUT_OF_RANGE),
errmsg("timestamp out of range")));
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
switch (val)
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
{
Per a brief conversation with Tom, I've created a patch for adding support for 'week' within the date_trunc function. Within the patch I added a couple of test cases and associated target output, and changed the documentation to add 'week' appropriately. Robert Creager
2004-03-05 03:41:14 +01:00
case DTK_WEEK:
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
{
int woy;
woy = date2isoweek(tm->tm_year, tm->tm_mon, tm->tm_mday);
/*
* If it is week 52/53 and the month is January, then the
* week must belong to the previous year. Also, some
* December dates belong to the next year.
*/
if (woy >= 52 && tm->tm_mon == 1)
--tm->tm_year;
if (woy <= 1 && tm->tm_mon == MONTHS_PER_YEAR)
++tm->tm_year;
isoweek2date(woy, &(tm->tm_year), &(tm->tm_mon), &(tm->tm_mday));
tm->tm_hour = 0;
tm->tm_min = 0;
tm->tm_sec = 0;
fsec = 0;
redotz = true;
break;
}
> After all that about numbering centuries and millenia correctly, > why does CVS tip still give me > > regression=# select extract(century from now()); > date_part > ----------- > 20 > (1 row) > [ ... looks in code ... ] > > Apparently it's because you fixed only timestamp_part, and not > timestamptz_part. I'm not too sure about what timestamp_trunc or > timestamptz_trunc should do, but they may be wrong as well. Sigh... as usual, what is not tested does not work:-( > Could we have a more complete patch? Please find a submission attached. I hope it really fixes all decade, century and millenium issues for extract and *_trunc functions on interval and other timestamp types. If someone could check that the results are reasonnable, it would be great. I indeed overlooked the fact that there were two functions. The patch fixes the code so that both variants agree. I added comments to interval extractions, because it relies on the C division to have a negative remainder: -7/10 = 0 and remains -7. As for *_trunc functions, I have chosen to put the first year of the century or millennium: -100, 1, 101... 1001 2001 etc. Indeed, I don't think it would make sense to put 2000 (last year of the 2nd millennium) for rounding all years of the third millenium. I also fixed the code so that all decades last 10 years and decade 199 means the 1990's. I have added some tests that are relevant to deal with tricky cases. The formula may be simplified, but all these cases must pass. Please keep them. Fabien Coelho
2004-08-20 05:45:14 +02:00
/* one may consider DTK_THOUSAND and DTK_HUNDRED... */
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 DTK_MILLENNIUM:
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
/*
* truncating to the millennium? what is this supposed to
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* mean? let us put the first year of the millennium... i.e.
* -1000, 1, 1001, 2001...
> After all that about numbering centuries and millenia correctly, > why does CVS tip still give me > > regression=# select extract(century from now()); > date_part > ----------- > 20 > (1 row) > [ ... looks in code ... ] > > Apparently it's because you fixed only timestamp_part, and not > timestamptz_part. I'm not too sure about what timestamp_trunc or > timestamptz_trunc should do, but they may be wrong as well. Sigh... as usual, what is not tested does not work:-( > Could we have a more complete patch? Please find a submission attached. I hope it really fixes all decade, century and millenium issues for extract and *_trunc functions on interval and other timestamp types. If someone could check that the results are reasonnable, it would be great. I indeed overlooked the fact that there were two functions. The patch fixes the code so that both variants agree. I added comments to interval extractions, because it relies on the C division to have a negative remainder: -7/10 = 0 and remains -7. As for *_trunc functions, I have chosen to put the first year of the century or millennium: -100, 1, 101... 1001 2001 etc. Indeed, I don't think it would make sense to put 2000 (last year of the 2nd millennium) for rounding all years of the third millenium. I also fixed the code so that all decades last 10 years and decade 199 means the 1990's. I have added some tests that are relevant to deal with tricky cases. The formula may be simplified, but all these cases must pass. Please keep them. Fabien Coelho
2004-08-20 05:45:14 +02:00
*/
if (tm->tm_year > 0)
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
tm->tm_year = ((tm->tm_year + 999) / 1000) * 1000 - 999;
> After all that about numbering centuries and millenia correctly, > why does CVS tip still give me > > regression=# select extract(century from now()); > date_part > ----------- > 20 > (1 row) > [ ... looks in code ... ] > > Apparently it's because you fixed only timestamp_part, and not > timestamptz_part. I'm not too sure about what timestamp_trunc or > timestamptz_trunc should do, but they may be wrong as well. Sigh... as usual, what is not tested does not work:-( > Could we have a more complete patch? Please find a submission attached. I hope it really fixes all decade, century and millenium issues for extract and *_trunc functions on interval and other timestamp types. If someone could check that the results are reasonnable, it would be great. I indeed overlooked the fact that there were two functions. The patch fixes the code so that both variants agree. I added comments to interval extractions, because it relies on the C division to have a negative remainder: -7/10 = 0 and remains -7. As for *_trunc functions, I have chosen to put the first year of the century or millennium: -100, 1, 101... 1001 2001 etc. Indeed, I don't think it would make sense to put 2000 (last year of the 2nd millennium) for rounding all years of the third millenium. I also fixed the code so that all decades last 10 years and decade 199 means the 1990's. I have added some tests that are relevant to deal with tricky cases. The formula may be simplified, but all these cases must pass. Please keep them. Fabien Coelho
2004-08-20 05:45:14 +02:00
else
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
tm->tm_year = -((999 - (tm->tm_year - 1)) / 1000) * 1000 + 1;
Add -Wimplicit-fallthrough to CFLAGS and CXXFLAGS Use it at level 4, a bit more restrictive than the default level, and tweak our commanding comments to FALLTHROUGH. (However, leave zic.c alone, since it's external code; to avoid the warnings that would appear there, change CFLAGS for that file in the Makefile.) Author: Julien Rouhaud <rjuju123@gmail.com> Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Tom Lane <tgl@sss.pgh.pa.us> Discussion: https://postgr.es/m/20200412081825.qyo5vwwco3fv4gdo@nol Discussion: https://postgr.es/m/flat/E1fDenm-0000C8-IJ@gemulon.postgresql.org
2020-05-12 22:01:52 +02:00
/* FALLTHROUGH */
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 DTK_CENTURY:
> After all that about numbering centuries and millenia correctly, > why does CVS tip still give me > > regression=# select extract(century from now()); > date_part > ----------- > 20 > (1 row) > [ ... looks in code ... ] > > Apparently it's because you fixed only timestamp_part, and not > timestamptz_part. I'm not too sure about what timestamp_trunc or > timestamptz_trunc should do, but they may be wrong as well. Sigh... as usual, what is not tested does not work:-( > Could we have a more complete patch? Please find a submission attached. I hope it really fixes all decade, century and millenium issues for extract and *_trunc functions on interval and other timestamp types. If someone could check that the results are reasonnable, it would be great. I indeed overlooked the fact that there were two functions. The patch fixes the code so that both variants agree. I added comments to interval extractions, because it relies on the C division to have a negative remainder: -7/10 = 0 and remains -7. As for *_trunc functions, I have chosen to put the first year of the century or millennium: -100, 1, 101... 1001 2001 etc. Indeed, I don't think it would make sense to put 2000 (last year of the 2nd millennium) for rounding all years of the third millenium. I also fixed the code so that all decades last 10 years and decade 199 means the 1990's. I have added some tests that are relevant to deal with tricky cases. The formula may be simplified, but all these cases must pass. Please keep them. Fabien Coelho
2004-08-20 05:45:14 +02:00
/* truncating to the century? as above: -100, 1, 101... */
if (tm->tm_year > 0)
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
tm->tm_year = ((tm->tm_year + 99) / 100) * 100 - 99;
> After all that about numbering centuries and millenia correctly, > why does CVS tip still give me > > regression=# select extract(century from now()); > date_part > ----------- > 20 > (1 row) > [ ... looks in code ... ] > > Apparently it's because you fixed only timestamp_part, and not > timestamptz_part. I'm not too sure about what timestamp_trunc or > timestamptz_trunc should do, but they may be wrong as well. Sigh... as usual, what is not tested does not work:-( > Could we have a more complete patch? Please find a submission attached. I hope it really fixes all decade, century and millenium issues for extract and *_trunc functions on interval and other timestamp types. If someone could check that the results are reasonnable, it would be great. I indeed overlooked the fact that there were two functions. The patch fixes the code so that both variants agree. I added comments to interval extractions, because it relies on the C division to have a negative remainder: -7/10 = 0 and remains -7. As for *_trunc functions, I have chosen to put the first year of the century or millennium: -100, 1, 101... 1001 2001 etc. Indeed, I don't think it would make sense to put 2000 (last year of the 2nd millennium) for rounding all years of the third millenium. I also fixed the code so that all decades last 10 years and decade 199 means the 1990's. I have added some tests that are relevant to deal with tricky cases. The formula may be simplified, but all these cases must pass. Please keep them. Fabien Coelho
2004-08-20 05:45:14 +02:00
else
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
tm->tm_year = -((99 - (tm->tm_year - 1)) / 100) * 100 + 1;
Add -Wimplicit-fallthrough to CFLAGS and CXXFLAGS Use it at level 4, a bit more restrictive than the default level, and tweak our commanding comments to FALLTHROUGH. (However, leave zic.c alone, since it's external code; to avoid the warnings that would appear there, change CFLAGS for that file in the Makefile.) Author: Julien Rouhaud <rjuju123@gmail.com> Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Tom Lane <tgl@sss.pgh.pa.us> Discussion: https://postgr.es/m/20200412081825.qyo5vwwco3fv4gdo@nol Discussion: https://postgr.es/m/flat/E1fDenm-0000C8-IJ@gemulon.postgresql.org
2020-05-12 22:01:52 +02:00
/* FALLTHROUGH */
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 DTK_DECADE:
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
/*
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* truncating to the decade? first year of the decade. must
* not be applied if year was truncated before!
> After all that about numbering centuries and millenia correctly, > why does CVS tip still give me > > regression=# select extract(century from now()); > date_part > ----------- > 20 > (1 row) > [ ... looks in code ... ] > > Apparently it's because you fixed only timestamp_part, and not > timestamptz_part. I'm not too sure about what timestamp_trunc or > timestamptz_trunc should do, but they may be wrong as well. Sigh... as usual, what is not tested does not work:-( > Could we have a more complete patch? Please find a submission attached. I hope it really fixes all decade, century and millenium issues for extract and *_trunc functions on interval and other timestamp types. If someone could check that the results are reasonnable, it would be great. I indeed overlooked the fact that there were two functions. The patch fixes the code so that both variants agree. I added comments to interval extractions, because it relies on the C division to have a negative remainder: -7/10 = 0 and remains -7. As for *_trunc functions, I have chosen to put the first year of the century or millennium: -100, 1, 101... 1001 2001 etc. Indeed, I don't think it would make sense to put 2000 (last year of the 2nd millennium) for rounding all years of the third millenium. I also fixed the code so that all decades last 10 years and decade 199 means the 1990's. I have added some tests that are relevant to deal with tricky cases. The formula may be simplified, but all these cases must pass. Please keep them. Fabien Coelho
2004-08-20 05:45:14 +02:00
*/
if (val != DTK_MILLENNIUM && val != DTK_CENTURY)
{
if (tm->tm_year > 0)
tm->tm_year = (tm->tm_year / 10) * 10;
else
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
tm->tm_year = -((8 - (tm->tm_year - 1)) / 10) * 10;
> After all that about numbering centuries and millenia correctly, > why does CVS tip still give me > > regression=# select extract(century from now()); > date_part > ----------- > 20 > (1 row) > [ ... looks in code ... ] > > Apparently it's because you fixed only timestamp_part, and not > timestamptz_part. I'm not too sure about what timestamp_trunc or > timestamptz_trunc should do, but they may be wrong as well. Sigh... as usual, what is not tested does not work:-( > Could we have a more complete patch? Please find a submission attached. I hope it really fixes all decade, century and millenium issues for extract and *_trunc functions on interval and other timestamp types. If someone could check that the results are reasonnable, it would be great. I indeed overlooked the fact that there were two functions. The patch fixes the code so that both variants agree. I added comments to interval extractions, because it relies on the C division to have a negative remainder: -7/10 = 0 and remains -7. As for *_trunc functions, I have chosen to put the first year of the century or millennium: -100, 1, 101... 1001 2001 etc. Indeed, I don't think it would make sense to put 2000 (last year of the 2nd millennium) for rounding all years of the third millenium. I also fixed the code so that all decades last 10 years and decade 199 means the 1990's. I have added some tests that are relevant to deal with tricky cases. The formula may be simplified, but all these cases must pass. Please keep them. Fabien Coelho
2004-08-20 05:45:14 +02:00
}
Add -Wimplicit-fallthrough to CFLAGS and CXXFLAGS Use it at level 4, a bit more restrictive than the default level, and tweak our commanding comments to FALLTHROUGH. (However, leave zic.c alone, since it's external code; to avoid the warnings that would appear there, change CFLAGS for that file in the Makefile.) Author: Julien Rouhaud <rjuju123@gmail.com> Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Tom Lane <tgl@sss.pgh.pa.us> Discussion: https://postgr.es/m/20200412081825.qyo5vwwco3fv4gdo@nol Discussion: https://postgr.es/m/flat/E1fDenm-0000C8-IJ@gemulon.postgresql.org
2020-05-12 22:01:52 +02:00
/* FALLTHROUGH */
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 DTK_YEAR:
tm->tm_mon = 1;
Add -Wimplicit-fallthrough to CFLAGS and CXXFLAGS Use it at level 4, a bit more restrictive than the default level, and tweak our commanding comments to FALLTHROUGH. (However, leave zic.c alone, since it's external code; to avoid the warnings that would appear there, change CFLAGS for that file in the Makefile.) Author: Julien Rouhaud <rjuju123@gmail.com> Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Tom Lane <tgl@sss.pgh.pa.us> Discussion: https://postgr.es/m/20200412081825.qyo5vwwco3fv4gdo@nol Discussion: https://postgr.es/m/flat/E1fDenm-0000C8-IJ@gemulon.postgresql.org
2020-05-12 22:01:52 +02:00
/* FALLTHROUGH */
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 DTK_QUARTER:
I corecting date_trunc('quarter',...) and friends because orig version doing '2003-07-30' -> '2003-04-01', '2003-11-30' ->'2003-07-01' B?jthe Zolt?n
2003-07-26 17:17:36 +02:00
tm->tm_mon = (3 * ((tm->tm_mon - 1) / 3)) + 1;
Add -Wimplicit-fallthrough to CFLAGS and CXXFLAGS Use it at level 4, a bit more restrictive than the default level, and tweak our commanding comments to FALLTHROUGH. (However, leave zic.c alone, since it's external code; to avoid the warnings that would appear there, change CFLAGS for that file in the Makefile.) Author: Julien Rouhaud <rjuju123@gmail.com> Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Tom Lane <tgl@sss.pgh.pa.us> Discussion: https://postgr.es/m/20200412081825.qyo5vwwco3fv4gdo@nol Discussion: https://postgr.es/m/flat/E1fDenm-0000C8-IJ@gemulon.postgresql.org
2020-05-12 22:01:52 +02:00
/* FALLTHROUGH */
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 DTK_MONTH:
tm->tm_mday = 1;
Add -Wimplicit-fallthrough to CFLAGS and CXXFLAGS Use it at level 4, a bit more restrictive than the default level, and tweak our commanding comments to FALLTHROUGH. (However, leave zic.c alone, since it's external code; to avoid the warnings that would appear there, change CFLAGS for that file in the Makefile.) Author: Julien Rouhaud <rjuju123@gmail.com> Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Tom Lane <tgl@sss.pgh.pa.us> Discussion: https://postgr.es/m/20200412081825.qyo5vwwco3fv4gdo@nol Discussion: https://postgr.es/m/flat/E1fDenm-0000C8-IJ@gemulon.postgresql.org
2020-05-12 22:01:52 +02:00
/* FALLTHROUGH */
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 DTK_DAY:
tm->tm_hour = 0;
timestamptz_trunc() should only recalculate the timezone when truncating to DAY precision or coarser; leave the timezone alone when precision is HOUR or less. This avoids surprises for inputs near a DST transition time, as per example from Matthew Gabeler-Lee. (The only reason we recalculate at all is so that outputs that are supposed to represent days will come out as local midnight, and that's not relevant for sub-day precision.)
2004-11-01 23:00:30 +01:00
redotz = true; /* for all cases >= DAY */
Add -Wimplicit-fallthrough to CFLAGS and CXXFLAGS Use it at level 4, a bit more restrictive than the default level, and tweak our commanding comments to FALLTHROUGH. (However, leave zic.c alone, since it's external code; to avoid the warnings that would appear there, change CFLAGS for that file in the Makefile.) Author: Julien Rouhaud <rjuju123@gmail.com> Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Tom Lane <tgl@sss.pgh.pa.us> Discussion: https://postgr.es/m/20200412081825.qyo5vwwco3fv4gdo@nol Discussion: https://postgr.es/m/flat/E1fDenm-0000C8-IJ@gemulon.postgresql.org
2020-05-12 22:01:52 +02:00
/* FALLTHROUGH */
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 DTK_HOUR:
tm->tm_min = 0;
Add -Wimplicit-fallthrough to CFLAGS and CXXFLAGS Use it at level 4, a bit more restrictive than the default level, and tweak our commanding comments to FALLTHROUGH. (However, leave zic.c alone, since it's external code; to avoid the warnings that would appear there, change CFLAGS for that file in the Makefile.) Author: Julien Rouhaud <rjuju123@gmail.com> Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Tom Lane <tgl@sss.pgh.pa.us> Discussion: https://postgr.es/m/20200412081825.qyo5vwwco3fv4gdo@nol Discussion: https://postgr.es/m/flat/E1fDenm-0000C8-IJ@gemulon.postgresql.org
2020-05-12 22:01:52 +02:00
/* FALLTHROUGH */
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 DTK_MINUTE:
tm->tm_sec = 0;
Add -Wimplicit-fallthrough to CFLAGS and CXXFLAGS Use it at level 4, a bit more restrictive than the default level, and tweak our commanding comments to FALLTHROUGH. (However, leave zic.c alone, since it's external code; to avoid the warnings that would appear there, change CFLAGS for that file in the Makefile.) Author: Julien Rouhaud <rjuju123@gmail.com> Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Tom Lane <tgl@sss.pgh.pa.us> Discussion: https://postgr.es/m/20200412081825.qyo5vwwco3fv4gdo@nol Discussion: https://postgr.es/m/flat/E1fDenm-0000C8-IJ@gemulon.postgresql.org
2020-05-12 22:01:52 +02:00
/* FALLTHROUGH */
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 DTK_SECOND:
fsec = 0;
break;
case DTK_MILLISEC:
Fix AT TIME ZONE for timestamps without time zones: test=> select ('2005-07-20 00:00:00'::timestamp without time zone) at time zone 'Europe/Paris'; timezone ------------------------ 2005-07-19 22:00:00-04 Udpate documentation.
2005-07-22 23:16:15 +02:00
fsec = (fsec / 1000) * 1000;
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;
case DTK_MICROSEC:
break;
default:
Error message editing in utils/adt. Again thanks to Joe Conway for doing the bulk of the heavy lifting ...
2003-07-27 06:53:12 +02:00
ereport(ERROR,
(errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
errmsg("timestamp with time zone units \"%s\" not "
"supported", lowunits)));
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
result = 0;
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +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
timestamptz_trunc() should only recalculate the timezone when truncating to DAY precision or coarser; leave the timezone alone when precision is HOUR or less. This avoids surprises for inputs near a DST transition time, as per example from Matthew Gabeler-Lee. (The only reason we recalculate at all is so that outputs that are supposed to represent days will come out as local midnight, and that's not relevant for sub-day precision.)
2004-11-01 23:00:30 +01:00
if (redotz)
Add a timezone-specific variant of date_trunc(). date_trunc(field, timestamptz, zone_name) performs truncation using the named time zone as reference, rather than working in the session time zone as is the default behavior. It's equivalent to date_trunc(field, timestamptz at time zone zone_name) at time zone zone_name but it's faster, easier to type, and arguably easier to understand. Vik Fearing and Tom Lane Discussion: https://postgr.es/m/6249ffc4-2b22-4c1b-4e7d-7af84fedd7c6@2ndquadrant.com
2018-11-14 21:41:07 +01:00
tz = DetermineTimeZoneOffset(tm, tzp);
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 (tm2timestamp(tm, fsec, &tz, &result) != 0)
Error message editing in utils/adt. Again thanks to Joe Conway for doing the bulk of the heavy lifting ...
2003-07-27 06:53:12 +02:00
ereport(ERROR,
(errcode(ERRCODE_DATETIME_VALUE_OUT_OF_RANGE),
errmsg("timestamp out of range")));
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
{
Error message editing in utils/adt. Again thanks to Joe Conway for doing the bulk of the heavy lifting ...
2003-07-27 06:53:12 +02:00
ereport(ERROR,
(errcode(ERRCODE_INVALID_PARAMETER_VALUE),
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
errmsg("timestamp with time zone units \"%s\" not recognized",
lowunits)));
Error message editing in utils/adt. Again thanks to Joe Conway for doing the bulk of the heavy lifting ...
2003-07-27 06:53:12 +02:00
result = 0;
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
}
Add a timezone-specific variant of date_trunc(). date_trunc(field, timestamptz, zone_name) performs truncation using the named time zone as reference, rather than working in the session time zone as is the default behavior. It's equivalent to date_trunc(field, timestamptz at time zone zone_name) at time zone zone_name but it's faster, easier to type, and arguably easier to understand. Vik Fearing and Tom Lane Discussion: https://postgr.es/m/6249ffc4-2b22-4c1b-4e7d-7af84fedd7c6@2ndquadrant.com
2018-11-14 21:41:07 +01:00
return result;
}
/* timestamptz_trunc()
* Truncate timestamptz to specified units in session timezone.
*/
Datum
timestamptz_trunc(PG_FUNCTION_ARGS)
{
text *units = PG_GETARG_TEXT_PP(0);
TimestampTz timestamp = PG_GETARG_TIMESTAMPTZ(1);
TimestampTz result;
if (TIMESTAMP_NOT_FINITE(timestamp))
PG_RETURN_TIMESTAMPTZ(timestamp);
result = timestamptz_trunc_internal(units, timestamp, session_timezone);
PG_RETURN_TIMESTAMPTZ(result);
}
/* timestamptz_trunc_zone()
* Truncate timestamptz to specified units in specified timezone.
*/
Datum
timestamptz_trunc_zone(PG_FUNCTION_ARGS)
{
text *units = PG_GETARG_TEXT_PP(0);
TimestampTz timestamp = PG_GETARG_TIMESTAMPTZ(1);
text *zone = PG_GETARG_TEXT_PP(2);
TimestampTz result;
char tzname[TZ_STRLEN_MAX + 1];
char *lowzone;
int type,
val;
pg_tz *tzp;
/*
* timestamptz_zone() doesn't look up the zone for infinite inputs, so we
* don't do so here either.
*/
if (TIMESTAMP_NOT_FINITE(timestamp))
PG_RETURN_TIMESTAMP(timestamp);
/*
* Look up the requested timezone (see notes in timestamptz_zone()).
*/
text_to_cstring_buffer(zone, tzname, sizeof(tzname));
/* DecodeTimezoneAbbrev requires lowercase input */
lowzone = downcase_truncate_identifier(tzname,
strlen(tzname),
false);
type = DecodeTimezoneAbbrev(0, lowzone, &val, &tzp);
if (type == TZ || type == DTZ)
{
/* fixed-offset abbreviation, get a pg_tz descriptor for that */
tzp = pg_tzset_offset(-val);
}
else if (type == DYNTZ)
{
/* dynamic-offset abbreviation, use its referenced timezone */
}
else
{
/* try it as a full zone name */
tzp = pg_tzset(tzname);
if (!tzp)
ereport(ERROR,
(errcode(ERRCODE_INVALID_PARAMETER_VALUE),
errmsg("time zone \"%s\" not recognized", tzname)));
}
result = timestamptz_trunc_internal(units, timestamp, tzp);
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
PG_RETURN_TIMESTAMPTZ(result);
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
}
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
/* interval_trunc()
* Extract specified field from interval.
*/
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
Datum
interval_trunc(PG_FUNCTION_ARGS)
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
{
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
text *units = PG_GETARG_TEXT_PP(0);
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
Interval *interval = PG_GETARG_INTERVAL_P(1);
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
Interval *result;
int type,
val;
Solve the 'Turkish problem' with undesirable locale behavior for case conversion of basic ASCII letters. Remove all uses of strcasecmp and strncasecmp in favor of new functions pg_strcasecmp and pg_strncasecmp; remove most but not all direct uses of toupper and tolower in favor of pg_toupper and pg_tolower. These functions use the same notions of case folding already developed for identifier case conversion. I left the straight locale-based folding in place for situations where we are just manipulating user data and not trying to match it to built-in strings --- for example, the SQL upper() function is still locale dependent. Perhaps this will prove not to be what's wanted, but at the moment we can initdb and pass regression tests in Turkish locale.
2004-05-07 02:24:59 +02:00
char *lowunits;
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;
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
struct pg_tm tt,
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
*tm = &tt;
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
result = (Interval *) palloc(sizeof(Interval));
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
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
lowunits = downcase_truncate_identifier(VARDATA_ANY(units),
VARSIZE_ANY_EXHDR(units),
Solve the 'Turkish problem' with undesirable locale behavior for case conversion of basic ASCII letters. Remove all uses of strcasecmp and strncasecmp in favor of new functions pg_strcasecmp and pg_strncasecmp; remove most but not all direct uses of toupper and tolower in favor of pg_toupper and pg_tolower. These functions use the same notions of case folding already developed for identifier case conversion. I left the straight locale-based folding in place for situations where we are just manipulating user data and not trying to match it to built-in strings --- for example, the SQL upper() function is still locale dependent. Perhaps this will prove not to be what's wanted, but at the moment we can initdb and pass regression tests in Turkish locale.
2004-05-07 02:24:59 +02:00
false);
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
type = DecodeUnits(0, lowunits, &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
if (type == UNITS)
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
{
if (interval2tm(*interval, tm, &fsec) == 0)
{
switch (val)
{
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:
> After all that about numbering centuries and millenia correctly, > why does CVS tip still give me > > regression=# select extract(century from now()); > date_part > ----------- > 20 > (1 row) > [ ... looks in code ... ] > > Apparently it's because you fixed only timestamp_part, and not > timestamptz_part. I'm not too sure about what timestamp_trunc or > timestamptz_trunc should do, but they may be wrong as well. Sigh... as usual, what is not tested does not work:-( > Could we have a more complete patch? Please find a submission attached. I hope it really fixes all decade, century and millenium issues for extract and *_trunc functions on interval and other timestamp types. If someone could check that the results are reasonnable, it would be great. I indeed overlooked the fact that there were two functions. The patch fixes the code so that both variants agree. I added comments to interval extractions, because it relies on the C division to have a negative remainder: -7/10 = 0 and remains -7. As for *_trunc functions, I have chosen to put the first year of the century or millennium: -100, 1, 101... 1001 2001 etc. Indeed, I don't think it would make sense to put 2000 (last year of the 2nd millennium) for rounding all years of the third millenium. I also fixed the code so that all decades last 10 years and decade 199 means the 1990's. I have added some tests that are relevant to deal with tricky cases. The formula may be simplified, but all these cases must pass. Please keep them. Fabien Coelho
2004-08-20 05:45:14 +02:00
/* caution: C division may have negative remainder */
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
tm->tm_year = (tm->tm_year / 1000) * 1000;
Add -Wimplicit-fallthrough to CFLAGS and CXXFLAGS Use it at level 4, a bit more restrictive than the default level, and tweak our commanding comments to FALLTHROUGH. (However, leave zic.c alone, since it's external code; to avoid the warnings that would appear there, change CFLAGS for that file in the Makefile.) Author: Julien Rouhaud <rjuju123@gmail.com> Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Tom Lane <tgl@sss.pgh.pa.us> Discussion: https://postgr.es/m/20200412081825.qyo5vwwco3fv4gdo@nol Discussion: https://postgr.es/m/flat/E1fDenm-0000C8-IJ@gemulon.postgresql.org
2020-05-12 22:01:52 +02:00
/* FALLTHROUGH */
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
case DTK_CENTURY:
> After all that about numbering centuries and millenia correctly, > why does CVS tip still give me > > regression=# select extract(century from now()); > date_part > ----------- > 20 > (1 row) > [ ... looks in code ... ] > > Apparently it's because you fixed only timestamp_part, and not > timestamptz_part. I'm not too sure about what timestamp_trunc or > timestamptz_trunc should do, but they may be wrong as well. Sigh... as usual, what is not tested does not work:-( > Could we have a more complete patch? Please find a submission attached. I hope it really fixes all decade, century and millenium issues for extract and *_trunc functions on interval and other timestamp types. If someone could check that the results are reasonnable, it would be great. I indeed overlooked the fact that there were two functions. The patch fixes the code so that both variants agree. I added comments to interval extractions, because it relies on the C division to have a negative remainder: -7/10 = 0 and remains -7. As for *_trunc functions, I have chosen to put the first year of the century or millennium: -100, 1, 101... 1001 2001 etc. Indeed, I don't think it would make sense to put 2000 (last year of the 2nd millennium) for rounding all years of the third millenium. I also fixed the code so that all decades last 10 years and decade 199 means the 1990's. I have added some tests that are relevant to deal with tricky cases. The formula may be simplified, but all these cases must pass. Please keep them. Fabien Coelho
2004-08-20 05:45:14 +02:00
/* caution: C division may have negative remainder */
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
tm->tm_year = (tm->tm_year / 100) * 100;
Add -Wimplicit-fallthrough to CFLAGS and CXXFLAGS Use it at level 4, a bit more restrictive than the default level, and tweak our commanding comments to FALLTHROUGH. (However, leave zic.c alone, since it's external code; to avoid the warnings that would appear there, change CFLAGS for that file in the Makefile.) Author: Julien Rouhaud <rjuju123@gmail.com> Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Tom Lane <tgl@sss.pgh.pa.us> Discussion: https://postgr.es/m/20200412081825.qyo5vwwco3fv4gdo@nol Discussion: https://postgr.es/m/flat/E1fDenm-0000C8-IJ@gemulon.postgresql.org
2020-05-12 22:01:52 +02:00
/* FALLTHROUGH */
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
case DTK_DECADE:
> After all that about numbering centuries and millenia correctly, > why does CVS tip still give me > > regression=# select extract(century from now()); > date_part > ----------- > 20 > (1 row) > [ ... looks in code ... ] > > Apparently it's because you fixed only timestamp_part, and not > timestamptz_part. I'm not too sure about what timestamp_trunc or > timestamptz_trunc should do, but they may be wrong as well. Sigh... as usual, what is not tested does not work:-( > Could we have a more complete patch? Please find a submission attached. I hope it really fixes all decade, century and millenium issues for extract and *_trunc functions on interval and other timestamp types. If someone could check that the results are reasonnable, it would be great. I indeed overlooked the fact that there were two functions. The patch fixes the code so that both variants agree. I added comments to interval extractions, because it relies on the C division to have a negative remainder: -7/10 = 0 and remains -7. As for *_trunc functions, I have chosen to put the first year of the century or millennium: -100, 1, 101... 1001 2001 etc. Indeed, I don't think it would make sense to put 2000 (last year of the 2nd millennium) for rounding all years of the third millenium. I also fixed the code so that all decades last 10 years and decade 199 means the 1990's. I have added some tests that are relevant to deal with tricky cases. The formula may be simplified, but all these cases must pass. Please keep them. Fabien Coelho
2004-08-20 05:45:14 +02:00
/* caution: C division may have negative remainder */
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
tm->tm_year = (tm->tm_year / 10) * 10;
Add -Wimplicit-fallthrough to CFLAGS and CXXFLAGS Use it at level 4, a bit more restrictive than the default level, and tweak our commanding comments to FALLTHROUGH. (However, leave zic.c alone, since it's external code; to avoid the warnings that would appear there, change CFLAGS for that file in the Makefile.) Author: Julien Rouhaud <rjuju123@gmail.com> Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Tom Lane <tgl@sss.pgh.pa.us> Discussion: https://postgr.es/m/20200412081825.qyo5vwwco3fv4gdo@nol Discussion: https://postgr.es/m/flat/E1fDenm-0000C8-IJ@gemulon.postgresql.org
2020-05-12 22:01:52 +02:00
/* FALLTHROUGH */
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
case DTK_YEAR:
tm->tm_mon = 0;
Add -Wimplicit-fallthrough to CFLAGS and CXXFLAGS Use it at level 4, a bit more restrictive than the default level, and tweak our commanding comments to FALLTHROUGH. (However, leave zic.c alone, since it's external code; to avoid the warnings that would appear there, change CFLAGS for that file in the Makefile.) Author: Julien Rouhaud <rjuju123@gmail.com> Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Tom Lane <tgl@sss.pgh.pa.us> Discussion: https://postgr.es/m/20200412081825.qyo5vwwco3fv4gdo@nol Discussion: https://postgr.es/m/flat/E1fDenm-0000C8-IJ@gemulon.postgresql.org
2020-05-12 22:01:52 +02:00
/* FALLTHROUGH */
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
case DTK_QUARTER:
Remove more extraneous parentheses in date/time functions.
2005-05-24 04:09:45 +02:00
tm->tm_mon = 3 * (tm->tm_mon / 3);
Add -Wimplicit-fallthrough to CFLAGS and CXXFLAGS Use it at level 4, a bit more restrictive than the default level, and tweak our commanding comments to FALLTHROUGH. (However, leave zic.c alone, since it's external code; to avoid the warnings that would appear there, change CFLAGS for that file in the Makefile.) Author: Julien Rouhaud <rjuju123@gmail.com> Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Tom Lane <tgl@sss.pgh.pa.us> Discussion: https://postgr.es/m/20200412081825.qyo5vwwco3fv4gdo@nol Discussion: https://postgr.es/m/flat/E1fDenm-0000C8-IJ@gemulon.postgresql.org
2020-05-12 22:01:52 +02:00
/* FALLTHROUGH */
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
case DTK_MONTH:
tm->tm_mday = 0;
Add -Wimplicit-fallthrough to CFLAGS and CXXFLAGS Use it at level 4, a bit more restrictive than the default level, and tweak our commanding comments to FALLTHROUGH. (However, leave zic.c alone, since it's external code; to avoid the warnings that would appear there, change CFLAGS for that file in the Makefile.) Author: Julien Rouhaud <rjuju123@gmail.com> Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Tom Lane <tgl@sss.pgh.pa.us> Discussion: https://postgr.es/m/20200412081825.qyo5vwwco3fv4gdo@nol Discussion: https://postgr.es/m/flat/E1fDenm-0000C8-IJ@gemulon.postgresql.org
2020-05-12 22:01:52 +02:00
/* FALLTHROUGH */
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
case DTK_DAY:
tm->tm_hour = 0;
Add -Wimplicit-fallthrough to CFLAGS and CXXFLAGS Use it at level 4, a bit more restrictive than the default level, and tweak our commanding comments to FALLTHROUGH. (However, leave zic.c alone, since it's external code; to avoid the warnings that would appear there, change CFLAGS for that file in the Makefile.) Author: Julien Rouhaud <rjuju123@gmail.com> Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Tom Lane <tgl@sss.pgh.pa.us> Discussion: https://postgr.es/m/20200412081825.qyo5vwwco3fv4gdo@nol Discussion: https://postgr.es/m/flat/E1fDenm-0000C8-IJ@gemulon.postgresql.org
2020-05-12 22:01:52 +02:00
/* FALLTHROUGH */
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
case DTK_HOUR:
tm->tm_min = 0;
Add -Wimplicit-fallthrough to CFLAGS and CXXFLAGS Use it at level 4, a bit more restrictive than the default level, and tweak our commanding comments to FALLTHROUGH. (However, leave zic.c alone, since it's external code; to avoid the warnings that would appear there, change CFLAGS for that file in the Makefile.) Author: Julien Rouhaud <rjuju123@gmail.com> Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Tom Lane <tgl@sss.pgh.pa.us> Discussion: https://postgr.es/m/20200412081825.qyo5vwwco3fv4gdo@nol Discussion: https://postgr.es/m/flat/E1fDenm-0000C8-IJ@gemulon.postgresql.org
2020-05-12 22:01:52 +02:00
/* FALLTHROUGH */
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
case DTK_MINUTE:
tm->tm_sec = 0;
Add -Wimplicit-fallthrough to CFLAGS and CXXFLAGS Use it at level 4, a bit more restrictive than the default level, and tweak our commanding comments to FALLTHROUGH. (However, leave zic.c alone, since it's external code; to avoid the warnings that would appear there, change CFLAGS for that file in the Makefile.) Author: Julien Rouhaud <rjuju123@gmail.com> Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Tom Lane <tgl@sss.pgh.pa.us> Discussion: https://postgr.es/m/20200412081825.qyo5vwwco3fv4gdo@nol Discussion: https://postgr.es/m/flat/E1fDenm-0000C8-IJ@gemulon.postgresql.org
2020-05-12 22:01:52 +02:00
/* FALLTHROUGH */
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
case DTK_SECOND:
fsec = 0;
break;
case DTK_MILLISEC:
Fix AT TIME ZONE for timestamps without time zones: test=> select ('2005-07-20 00:00:00'::timestamp without time zone) at time zone 'Europe/Paris'; timezone ------------------------ 2005-07-19 22:00:00-04 Udpate documentation.
2005-07-22 23:16:15 +02:00
fsec = (fsec / 1000) * 1000;
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
break;
case DTK_MICROSEC:
break;
default:
On second thought, explain why date_trunc("week") on interval values is not supported in the error message, rather than the docs.
2012-08-15 22:48:05 +02:00
if (val == DTK_WEEK)
ereport(ERROR,
(errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
errmsg("interval units \"%s\" not supported "
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
"because months usually have fractional weeks",
On second thought, explain why date_trunc("week") on interval values is not supported in the error message, rather than the docs.
2012-08-15 22:48:05 +02:00
lowunits)));
else
ereport(ERROR,
(errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
errmsg("interval units \"%s\" not supported",
lowunits)));
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
}
if (tm2interval(tm, fsec, result) != 0)
Error message editing in utils/adt. Again thanks to Joe Conway for doing the bulk of the heavy lifting ...
2003-07-27 06:53:12 +02:00
ereport(ERROR,
(errcode(ERRCODE_DATETIME_VALUE_OUT_OF_RANGE),
errmsg("interval out of range")));
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
}
else
Error message editing in utils/adt. Again thanks to Joe Conway for doing the bulk of the heavy lifting ...
2003-07-27 06:53:12 +02:00
elog(ERROR, "could not convert interval to tm");
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
}
else
{
Error message editing in utils/adt. Again thanks to Joe Conway for doing the bulk of the heavy lifting ...
2003-07-27 06:53:12 +02:00
ereport(ERROR,
(errcode(ERRCODE_INVALID_PARAMETER_VALUE),
errmsg("interval units \"%s\" not recognized",
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
lowunits)));
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
}
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
PG_RETURN_INTERVAL_P(result);
}
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
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
/* isoweek2j()
*
pgindent run for 8.3.
2007-11-15 22:14:46 +01:00
* Return the Julian day which corresponds to the first day (Monday) of the given ISO 8601 year and week.
* Julian days are used to convert between ISO week dates and Gregorian dates.
* to_char: - full support for IW (ISO week) and vice versa conversion for IW too (the to_char 'week' support is now complete and I hope correct). Thomas, I use for IW code from timestamp.c, for this I create separate function date2isoweek() from original 'case DTK_WEEK:' code in the timestamp_part(). I mean will better use one code for same feature in date_part() and in to_char(). The isoweek2date() is added to timestamp.c too. Right? IMHO in 7.1 will all to_char's features complete. It is cca 41 templates for date/time and cca 21 for numbers. * to_ascii: - gcc, is it correct now? :-) In the patch is documentation for to_char's IW and for to_ascii(). Karel
2000-08-29 06:41:48 +02:00
*/
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
int
isoweek2j(int year, int week)
* to_char: - full support for IW (ISO week) and vice versa conversion for IW too (the to_char 'week' support is now complete and I hope correct). Thomas, I use for IW code from timestamp.c, for this I create separate function date2isoweek() from original 'case DTK_WEEK:' code in the timestamp_part(). I mean will better use one code for same feature in date_part() and in to_char(). The isoweek2date() is added to timestamp.c too. Right? IMHO in 7.1 will all to_char's features complete. It is cca 41 templates for date/time and cca 21 for numbers. * to_ascii: - gcc, is it correct now? :-) In the patch is documentation for to_char's IW and for to_ascii(). Karel
2000-08-29 06:41:48 +02:00
{
pgindent run. Make it all clean.
2001-03-22 05:01:46 +01:00
int day0,
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
day4;
pgindent run. Make it all clean.
2001-03-22 05:01:46 +01:00
* to_char: - full support for IW (ISO week) and vice versa conversion for IW too (the to_char 'week' support is now complete and I hope correct). Thomas, I use for IW code from timestamp.c, for this I create separate function date2isoweek() from original 'case DTK_WEEK:' code in the timestamp_part(). I mean will better use one code for same feature in date_part() and in to_char(). The isoweek2date() is added to timestamp.c too. Right? IMHO in 7.1 will all to_char's features complete. It is cca 41 templates for date/time and cca 21 for numbers. * to_ascii: - gcc, is it correct now? :-) In the patch is documentation for to_char's IW and for to_ascii(). Karel
2000-08-29 06:41:48 +02:00
/* fourth day of current year */
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
day4 = date2j(year, 1, 4);
pgindent run. Make it all clean.
2001-03-22 05:01:46 +01:00
* to_char: - full support for IW (ISO week) and vice versa conversion for IW too (the to_char 'week' support is now complete and I hope correct). Thomas, I use for IW code from timestamp.c, for this I create separate function date2isoweek() from original 'case DTK_WEEK:' code in the timestamp_part(). I mean will better use one code for same feature in date_part() and in to_char(). The isoweek2date() is added to timestamp.c too. Right? IMHO in 7.1 will all to_char's features complete. It is cca 41 templates for date/time and cca 21 for numbers. * to_ascii: - gcc, is it correct now? :-) In the patch is documentation for to_char's IW and for to_ascii(). Karel
2000-08-29 06:41:48 +02:00
/* day0 == offset to first day of week (Monday) */
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
day0 = j2day(day4 - 1);
* to_char: - full support for IW (ISO week) and vice versa conversion for IW too (the to_char 'week' support is now complete and I hope correct). Thomas, I use for IW code from timestamp.c, for this I create separate function date2isoweek() from original 'case DTK_WEEK:' code in the timestamp_part(). I mean will better use one code for same feature in date_part() and in to_char(). The isoweek2date() is added to timestamp.c too. Right? IMHO in 7.1 will all to_char's features complete. It is cca 41 templates for date/time and cca 21 for numbers. * to_ascii: - gcc, is it correct now? :-) In the patch is documentation for to_char's IW and for to_ascii(). Karel
2000-08-29 06:41:48 +02:00
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
return ((week - 1) * 7) + (day4 - day0);
}
/* isoweek2date()
* Convert ISO week of year number to date.
* The year field must be specified with the ISO year!
* karel 2000/08/07
*/
void
isoweek2date(int woy, int *year, int *mon, int *mday)
{
j2date(isoweek2j(*year, woy), year, mon, mday);
}
/* isoweekdate2date()
*
Fix to_date() and to_timestamp() to allow specification of the day of the week via ISO or Gregorian designations. The fix is to store the day-of-week consistently as 1-7, Sunday = 1. Fixes bug reported by Marc Munro
2012-09-04 04:52:34 +02:00
* Convert an ISO 8601 week date (ISO year, ISO week) into a Gregorian date.
* Gregorian day of week sent so weekday strings can be supplied.
pgindent run for 8.3.
2007-11-15 22:14:46 +01:00
* Populates year, mon, and mday with the correct Gregorian values.
* year must be passed in as the ISO year.
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
*/
void
Fix to_date() and to_timestamp() to allow specification of the day of the week via ISO or Gregorian designations. The fix is to store the day-of-week consistently as 1-7, Sunday = 1. Fixes bug reported by Marc Munro
2012-09-04 04:52:34 +02:00
isoweekdate2date(int isoweek, int wday, int *year, int *mon, int *mday)
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
{
pgindent run for 8.3.
2007-11-15 22:14:46 +01:00
int jday;
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
jday = isoweek2j(*year, isoweek);
Fix to_date() and to_timestamp() to allow specification of the day of the week via ISO or Gregorian designations. The fix is to store the day-of-week consistently as 1-7, Sunday = 1. Fixes bug reported by Marc Munro
2012-09-04 04:52:34 +02:00
/* convert Gregorian week start (Sunday=1) to ISO week start (Monday=1) */
if (wday > 1)
jday += wday - 2;
else
jday += 6;
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
j2date(jday, year, mon, mday);
* to_char: - full support for IW (ISO week) and vice versa conversion for IW too (the to_char 'week' support is now complete and I hope correct). Thomas, I use for IW code from timestamp.c, for this I create separate function date2isoweek() from original 'case DTK_WEEK:' code in the timestamp_part(). I mean will better use one code for same feature in date_part() and in to_char(). The isoweek2date() is added to timestamp.c too. Right? IMHO in 7.1 will all to_char's features complete. It is cca 41 templates for date/time and cca 21 for numbers. * to_ascii: - gcc, is it correct now? :-) In the patch is documentation for to_char's IW and for to_ascii(). Karel
2000-08-29 06:41:48 +02:00
}
/* date2isoweek()
pgindent run. Make it all clean.
2001-03-22 05:01:46 +01:00
*
* to_char: - full support for IW (ISO week) and vice versa conversion for IW too (the to_char 'week' support is now complete and I hope correct). Thomas, I use for IW code from timestamp.c, for this I create separate function date2isoweek() from original 'case DTK_WEEK:' code in the timestamp_part(). I mean will better use one code for same feature in date_part() and in to_char(). The isoweek2date() is added to timestamp.c too. Right? IMHO in 7.1 will all to_char's features complete. It is cca 41 templates for date/time and cca 21 for numbers. * to_ascii: - gcc, is it correct now? :-) In the patch is documentation for to_char's IW and for to_ascii(). Karel
2000-08-29 06:41:48 +02:00
* Returns ISO week number of year.
*/
int
pgindent run. Make it all clean.
2001-03-22 05:01:46 +01:00
date2isoweek(int year, int mon, int mday)
* to_char: - full support for IW (ISO week) and vice versa conversion for IW too (the to_char 'week' support is now complete and I hope correct). Thomas, I use for IW code from timestamp.c, for this I create separate function date2isoweek() from original 'case DTK_WEEK:' code in the timestamp_part(). I mean will better use one code for same feature in date_part() and in to_char(). The isoweek2date() is added to timestamp.c too. Right? IMHO in 7.1 will all to_char's features complete. It is cca 41 templates for date/time and cca 21 for numbers. * to_ascii: - gcc, is it correct now? :-) In the patch is documentation for to_char's IW and for to_ascii(). Karel
2000-08-29 06:41:48 +02:00
{
pgindent run. Make it all clean.
2001-03-22 05:01:46 +01:00
float8 result;
int day0,
day4,
dayn;
/* current day */
* to_char: - full support for IW (ISO week) and vice versa conversion for IW too (the to_char 'week' support is now complete and I hope correct). Thomas, I use for IW code from timestamp.c, for this I create separate function date2isoweek() from original 'case DTK_WEEK:' code in the timestamp_part(). I mean will better use one code for same feature in date_part() and in to_char(). The isoweek2date() is added to timestamp.c too. Right? IMHO in 7.1 will all to_char's features complete. It is cca 41 templates for date/time and cca 21 for numbers. * to_ascii: - gcc, is it correct now? :-) In the patch is documentation for to_char's IW and for to_ascii(). Karel
2000-08-29 06:41:48 +02:00
dayn = date2j(year, mon, mday);
pgindent run. Make it all clean.
2001-03-22 05:01:46 +01:00
* to_char: - full support for IW (ISO week) and vice versa conversion for IW too (the to_char 'week' support is now complete and I hope correct). Thomas, I use for IW code from timestamp.c, for this I create separate function date2isoweek() from original 'case DTK_WEEK:' code in the timestamp_part(). I mean will better use one code for same feature in date_part() and in to_char(). The isoweek2date() is added to timestamp.c too. Right? IMHO in 7.1 will all to_char's features complete. It is cca 41 templates for date/time and cca 21 for numbers. * to_ascii: - gcc, is it correct now? :-) In the patch is documentation for to_char's IW and for to_ascii(). Karel
2000-08-29 06:41:48 +02:00
/* fourth day of current year */
day4 = date2j(year, 1, 4);
pgindent run. Make it all clean.
2001-03-22 05:01:46 +01:00
* to_char: - full support for IW (ISO week) and vice versa conversion for IW too (the to_char 'week' support is now complete and I hope correct). Thomas, I use for IW code from timestamp.c, for this I create separate function date2isoweek() from original 'case DTK_WEEK:' code in the timestamp_part(). I mean will better use one code for same feature in date_part() and in to_char(). The isoweek2date() is added to timestamp.c too. Right? IMHO in 7.1 will all to_char's features complete. It is cca 41 templates for date/time and cca 21 for numbers. * to_ascii: - gcc, is it correct now? :-) In the patch is documentation for to_char's IW and for to_ascii(). Karel
2000-08-29 06:41:48 +02:00
/* day0 == offset to first day of week (Monday) */
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
day0 = j2day(day4 - 1);
pgindent run. Make it all clean.
2001-03-22 05:01:46 +01:00
/*
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* We need the first week containing a Thursday, otherwise this day falls
* into the previous year for purposes of counting weeks
* to_char: - full support for IW (ISO week) and vice versa conversion for IW too (the to_char 'week' support is now complete and I hope correct). Thomas, I use for IW code from timestamp.c, for this I create separate function date2isoweek() from original 'case DTK_WEEK:' code in the timestamp_part(). I mean will better use one code for same feature in date_part() and in to_char(). The isoweek2date() is added to timestamp.c too. Right? IMHO in 7.1 will all to_char's features complete. It is cca 41 templates for date/time and cca 21 for numbers. * to_ascii: - gcc, is it correct now? :-) In the patch is documentation for to_char's IW and for to_ascii(). Karel
2000-08-29 06:41:48 +02:00
*/
Remove unnecessary parentheses in datetime/timestamp code.
2005-05-23 19:13:14 +02:00
if (dayn < day4 - day0)
* to_char: - full support for IW (ISO week) and vice versa conversion for IW too (the to_char 'week' support is now complete and I hope correct). Thomas, I use for IW code from timestamp.c, for this I create separate function date2isoweek() from original 'case DTK_WEEK:' code in the timestamp_part(). I mean will better use one code for same feature in date_part() and in to_char(). The isoweek2date() is added to timestamp.c too. Right? IMHO in 7.1 will all to_char's features complete. It is cca 41 templates for date/time and cca 21 for numbers. * to_ascii: - gcc, is it correct now? :-) In the patch is documentation for to_char's IW and for to_ascii(). Karel
2000-08-29 06:41:48 +02:00
{
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
day4 = date2j(year - 1, 1, 4);
pgindent run. Make it all clean.
2001-03-22 05:01:46 +01:00
* to_char: - full support for IW (ISO week) and vice versa conversion for IW too (the to_char 'week' support is now complete and I hope correct). Thomas, I use for IW code from timestamp.c, for this I create separate function date2isoweek() from original 'case DTK_WEEK:' code in the timestamp_part(). I mean will better use one code for same feature in date_part() and in to_char(). The isoweek2date() is added to timestamp.c too. Right? IMHO in 7.1 will all to_char's features complete. It is cca 41 templates for date/time and cca 21 for numbers. * to_ascii: - gcc, is it correct now? :-) In the patch is documentation for to_char's IW and for to_ascii(). Karel
2000-08-29 06:41:48 +02:00
/* day0 == offset to first day of week (Monday) */
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
day0 = j2day(day4 - 1);
* to_char: - full support for IW (ISO week) and vice versa conversion for IW too (the to_char 'week' support is now complete and I hope correct). Thomas, I use for IW code from timestamp.c, for this I create separate function date2isoweek() from original 'case DTK_WEEK:' code in the timestamp_part(). I mean will better use one code for same feature in date_part() and in to_char(). The isoweek2date() is added to timestamp.c too. Right? IMHO in 7.1 will all to_char's features complete. It is cca 41 templates for date/time and cca 21 for numbers. * to_ascii: - gcc, is it correct now? :-) In the patch is documentation for to_char's IW and for to_ascii(). Karel
2000-08-29 06:41:48 +02:00
}
pgindent run. Make it all clean.
2001-03-22 05:01:46 +01:00
Remove unnecessary parentheses in datetime/timestamp code.
2005-05-23 19:13:14 +02:00
result = (dayn - (day4 - day0)) / 7 + 1;
pgindent run. Make it all clean.
2001-03-22 05:01:46 +01:00
/*
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* Sometimes the last few days in a year will fall into the first week of
* the next year, so check for this.
* to_char: - full support for IW (ISO week) and vice versa conversion for IW too (the to_char 'week' support is now complete and I hope correct). Thomas, I use for IW code from timestamp.c, for this I create separate function date2isoweek() from original 'case DTK_WEEK:' code in the timestamp_part(). I mean will better use one code for same feature in date_part() and in to_char(). The isoweek2date() is added to timestamp.c too. Right? IMHO in 7.1 will all to_char's features complete. It is cca 41 templates for date/time and cca 21 for numbers. * to_ascii: - gcc, is it correct now? :-) In the patch is documentation for to_char's IW and for to_ascii(). Karel
2000-08-29 06:41:48 +02:00
*/
Fix wrong week returnded by date_trunc('week') for early dates in January --- would return wrong year for 2005-01-01 and 2006-01-01. per report from Robert Creager. Backpatch to 8.0.X.
2005-04-01 16:25:23 +02:00
if (result >= 52)
* to_char: - full support for IW (ISO week) and vice versa conversion for IW too (the to_char 'week' support is now complete and I hope correct). Thomas, I use for IW code from timestamp.c, for this I create separate function date2isoweek() from original 'case DTK_WEEK:' code in the timestamp_part(). I mean will better use one code for same feature in date_part() and in to_char(). The isoweek2date() is added to timestamp.c too. Right? IMHO in 7.1 will all to_char's features complete. It is cca 41 templates for date/time and cca 21 for numbers. * to_ascii: - gcc, is it correct now? :-) In the patch is documentation for to_char's IW and for to_ascii(). Karel
2000-08-29 06:41:48 +02:00
{
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
day4 = date2j(year + 1, 1, 4);
pgindent run. Make it all clean.
2001-03-22 05:01:46 +01:00
* to_char: - full support for IW (ISO week) and vice versa conversion for IW too (the to_char 'week' support is now complete and I hope correct). Thomas, I use for IW code from timestamp.c, for this I create separate function date2isoweek() from original 'case DTK_WEEK:' code in the timestamp_part(). I mean will better use one code for same feature in date_part() and in to_char(). The isoweek2date() is added to timestamp.c too. Right? IMHO in 7.1 will all to_char's features complete. It is cca 41 templates for date/time and cca 21 for numbers. * to_ascii: - gcc, is it correct now? :-) In the patch is documentation for to_char's IW and for to_ascii(). Karel
2000-08-29 06:41:48 +02:00
/* day0 == offset to first day of week (Monday) */
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
day0 = j2day(day4 - 1);
pgindent run. Make it all clean.
2001-03-22 05:01:46 +01:00
Remove unnecessary parentheses in datetime/timestamp code.
2005-05-23 19:13:14 +02:00
if (dayn >= day4 - day0)
result = (dayn - (day4 - day0)) / 7 + 1;
* to_char: - full support for IW (ISO week) and vice versa conversion for IW too (the to_char 'week' support is now complete and I hope correct). Thomas, I use for IW code from timestamp.c, for this I create separate function date2isoweek() from original 'case DTK_WEEK:' code in the timestamp_part(). I mean will better use one code for same feature in date_part() and in to_char(). The isoweek2date() is added to timestamp.c too. Right? IMHO in 7.1 will all to_char's features complete. It is cca 41 templates for date/time and cca 21 for numbers. * to_ascii: - gcc, is it correct now? :-) In the patch is documentation for to_char's IW and for to_ascii(). Karel
2000-08-29 06:41:48 +02:00
}
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
* to_char: - full support for IW (ISO week) and vice versa conversion for IW too (the to_char 'week' support is now complete and I hope correct). Thomas, I use for IW code from timestamp.c, for this I create separate function date2isoweek() from original 'case DTK_WEEK:' code in the timestamp_part(). I mean will better use one code for same feature in date_part() and in to_char(). The isoweek2date() is added to timestamp.c too. Right? IMHO in 7.1 will all to_char's features complete. It is cca 41 templates for date/time and cca 21 for numbers. * to_ascii: - gcc, is it correct now? :-) In the patch is documentation for to_char's IW and for to_ascii(). Karel
2000-08-29 06:41:48 +02:00
return (int) result;
pgindent run. Make it all clean.
2001-03-22 05:01:46 +01:00
}
> > I have no idea if this in Oracle or not. But it's something I > > needed, and other people in the past asked about it too. > > It is in Oracle, but you aren't exactly on the spot. It should be > > IYYY - 4 digits ('2003') > IYY - 3 digits ('003') > IY - 2 digits ('03') > I - 1 digit ('3') Here is an updated patch that does that. Kurt Roeckx
2003-12-25 04:36:24 +01:00
/* date2isoyear()
*
* Returns ISO 8601 year number.
Fix EXTRACT(ISOYEAR FROM timestamp) for years BC. The test cases added by commit 26ae3aa80 exposed an old oversight in timestamp[tz]_part: they didn't correct the result of date2isoyear() for BC years, so that we produced an off-by-one answer for such years. Fix that, and back-patch to all supported branches. Discussion: https://postgr.es/m/SG2PR06MB37762CAE45DB0F6CA7001EA9B6550@SG2PR06MB3776.apcprd06.prod.outlook.com
2019-12-12 18:30:43 +01:00
* Note: zero or negative results follow the year-zero-exists convention.
> > I have no idea if this in Oracle or not. But it's something I > > needed, and other people in the past asked about it too. > > It is in Oracle, but you aren't exactly on the spot. It should be > > IYYY - 4 digits ('2003') > IYY - 3 digits ('003') > IY - 2 digits ('03') > I - 1 digit ('3') Here is an updated patch that does that. Kurt Roeckx
2003-12-25 04:36:24 +01:00
*/
int
date2isoyear(int year, int mon, int mday)
{
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
float8 result;
int day0,
day4,
dayn;
> > I have no idea if this in Oracle or not. But it's something I > > needed, and other people in the past asked about it too. > > It is in Oracle, but you aren't exactly on the spot. It should be > > IYYY - 4 digits ('2003') > IYY - 3 digits ('003') > IY - 2 digits ('03') > I - 1 digit ('3') Here is an updated patch that does that. Kurt Roeckx
2003-12-25 04:36:24 +01:00
/* current day */
dayn = date2j(year, mon, mday);
/* fourth day of current year */
day4 = date2j(year, 1, 4);
/* day0 == offset to first day of week (Monday) */
day0 = j2day(day4 - 1);
/*
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* We need the first week containing a Thursday, otherwise this day falls
* into the previous year for purposes of counting weeks
> > I have no idea if this in Oracle or not. But it's something I > > needed, and other people in the past asked about it too. > > It is in Oracle, but you aren't exactly on the spot. It should be > > IYYY - 4 digits ('2003') > IYY - 3 digits ('003') > IY - 2 digits ('03') > I - 1 digit ('3') Here is an updated patch that does that. Kurt Roeckx
2003-12-25 04:36:24 +01:00
*/
Remove unnecessary parentheses in datetime/timestamp code.
2005-05-23 19:13:14 +02:00
if (dayn < day4 - day0)
> > I have no idea if this in Oracle or not. But it's something I > > needed, and other people in the past asked about it too. > > It is in Oracle, but you aren't exactly on the spot. It should be > > IYYY - 4 digits ('2003') > IYY - 3 digits ('003') > IY - 2 digits ('03') > I - 1 digit ('3') Here is an updated patch that does that. Kurt Roeckx
2003-12-25 04:36:24 +01:00
{
day4 = date2j(year - 1, 1, 4);
/* day0 == offset to first day of week (Monday) */
day0 = j2day(day4 - 1);
year--;
}
Remove unnecessary parentheses in datetime/timestamp code.
2005-05-23 19:13:14 +02:00
result = (dayn - (day4 - day0)) / 7 + 1;
> > I have no idea if this in Oracle or not. But it's something I > > needed, and other people in the past asked about it too. > > It is in Oracle, but you aren't exactly on the spot. It should be > > IYYY - 4 digits ('2003') > IYY - 3 digits ('003') > IY - 2 digits ('03') > I - 1 digit ('3') Here is an updated patch that does that. Kurt Roeckx
2003-12-25 04:36:24 +01:00
/*
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* Sometimes the last few days in a year will fall into the first week of
* the next year, so check for this.
> > I have no idea if this in Oracle or not. But it's something I > > needed, and other people in the past asked about it too. > > It is in Oracle, but you aren't exactly on the spot. It should be > > IYYY - 4 digits ('2003') > IYY - 3 digits ('003') > IY - 2 digits ('03') > I - 1 digit ('3') Here is an updated patch that does that. Kurt Roeckx
2003-12-25 04:36:24 +01:00
*/
Fix wrong week returnded by date_trunc('week') for early dates in January --- would return wrong year for 2005-01-01 and 2006-01-01. per report from Robert Creager. Backpatch to 8.0.X.
2005-04-01 16:25:23 +02:00
if (result >= 52)
> > I have no idea if this in Oracle or not. But it's something I > > needed, and other people in the past asked about it too. > > It is in Oracle, but you aren't exactly on the spot. It should be > > IYYY - 4 digits ('2003') > IYY - 3 digits ('003') > IY - 2 digits ('03') > I - 1 digit ('3') Here is an updated patch that does that. Kurt Roeckx
2003-12-25 04:36:24 +01:00
{
day4 = date2j(year + 1, 1, 4);
/* day0 == offset to first day of week (Monday) */
day0 = j2day(day4 - 1);
Remove unnecessary parentheses in datetime/timestamp code.
2005-05-23 19:13:14 +02:00
if (dayn >= day4 - day0)
> > I have no idea if this in Oracle or not. But it's something I > > needed, and other people in the past asked about it too. > > It is in Oracle, but you aren't exactly on the spot. It should be > > IYYY - 4 digits ('2003') > IYY - 3 digits ('003') > IY - 2 digits ('03') > I - 1 digit ('3') Here is an updated patch that does that. Kurt Roeckx
2003-12-25 04:36:24 +01:00
year++;
}
return year;
}
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
/* date2isoyearday()
*
pgindent run for 8.3.
2007-11-15 22:14:46 +01:00
* Returns the ISO 8601 day-of-year, given a Gregorian year, month and day.
* Possible return values are 1 through 371 (364 in non-leap years).
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
*/
int
date2isoyearday(int year, int mon, int mday)
{
return date2j(year, mon, mday) - isoweek2j(date2isoyear(year, mon, mday), 1) + 1;
}
Make extract() do something more reasonable with infinite datetimes. Historically, extract() just returned zero for any case involving an infinite timestamp[tz] input; even cases in which the unit name was invalid. This is not very sensible. Instead, return infinity or -infinity as appropriate when the requested field is one that is monotonically increasing (e.g, year, epoch), or NULL when it is not (e.g., day, hour). Also, throw the expected errors for bad unit names. BACKWARDS INCOMPATIBLE CHANGE Vitaly Burovoy, reviewed by Vik Fearing
2016-01-22 04:26:20 +01:00
/*
* NonFiniteTimestampTzPart
*
* Used by timestamp_part and timestamptz_part when extracting from infinite
* timestamp[tz]. Returns +/-Infinity if that is the appropriate result,
* otherwise returns zero (which should be taken as meaning to return NULL).
*
* Errors thrown here for invalid units should exactly match those that
* would be thrown in the calling functions, else there will be unexpected
* discrepancies between finite- and infinite-input cases.
*/
static float8
NonFiniteTimestampTzPart(int type, int unit, char *lowunits,
bool isNegative, bool isTz)
{
if ((type != UNITS) && (type != RESERV))
{
if (isTz)
ereport(ERROR,
(errcode(ERRCODE_INVALID_PARAMETER_VALUE),
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
errmsg("timestamp with time zone units \"%s\" not recognized",
lowunits)));
Make extract() do something more reasonable with infinite datetimes. Historically, extract() just returned zero for any case involving an infinite timestamp[tz] input; even cases in which the unit name was invalid. This is not very sensible. Instead, return infinity or -infinity as appropriate when the requested field is one that is monotonically increasing (e.g, year, epoch), or NULL when it is not (e.g., day, hour). Also, throw the expected errors for bad unit names. BACKWARDS INCOMPATIBLE CHANGE Vitaly Burovoy, reviewed by Vik Fearing
2016-01-22 04:26:20 +01:00
else
ereport(ERROR,
(errcode(ERRCODE_INVALID_PARAMETER_VALUE),
errmsg("timestamp units \"%s\" not recognized",
lowunits)));
}
switch (unit)
{
/* Oscillating units */
case DTK_MICROSEC:
case DTK_MILLISEC:
case DTK_SECOND:
case DTK_MINUTE:
case DTK_HOUR:
case DTK_DAY:
case DTK_MONTH:
case DTK_QUARTER:
case DTK_WEEK:
case DTK_DOW:
case DTK_ISODOW:
case DTK_DOY:
case DTK_TZ:
case DTK_TZ_MINUTE:
case DTK_TZ_HOUR:
return 0.0;
/* Monotonically-increasing units */
case DTK_YEAR:
case DTK_DECADE:
case DTK_CENTURY:
case DTK_MILLENNIUM:
case DTK_JULIAN:
case DTK_ISOYEAR:
case DTK_EPOCH:
if (isNegative)
return -get_float8_infinity();
else
return get_float8_infinity();
default:
if (isTz)
ereport(ERROR,
(errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
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
errmsg("timestamp with time zone units \"%s\" not supported",
lowunits)));
Make extract() do something more reasonable with infinite datetimes. Historically, extract() just returned zero for any case involving an infinite timestamp[tz] input; even cases in which the unit name was invalid. This is not very sensible. Instead, return infinity or -infinity as appropriate when the requested field is one that is monotonically increasing (e.g, year, epoch), or NULL when it is not (e.g., day, hour). Also, throw the expected errors for bad unit names. BACKWARDS INCOMPATIBLE CHANGE Vitaly Burovoy, reviewed by Vik Fearing
2016-01-22 04:26:20 +01:00
else
ereport(ERROR,
(errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
errmsg("timestamp units \"%s\" not supported",
lowunits)));
return 0.0; /* keep compiler quiet */
}
}
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
/* timestamp_part()
* Extract specified field from timestamp.
*/
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
Datum
timestamp_part(PG_FUNCTION_ARGS)
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
{
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
text *units = PG_GETARG_TEXT_PP(0);
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
Timestamp timestamp = PG_GETARG_TIMESTAMP(1);
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
float8 result;
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
Timestamp epoch;
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
int type,
val;
Solve the 'Turkish problem' with undesirable locale behavior for case conversion of basic ASCII letters. Remove all uses of strcasecmp and strncasecmp in favor of new functions pg_strcasecmp and pg_strncasecmp; remove most but not all direct uses of toupper and tolower in favor of pg_toupper and pg_tolower. These functions use the same notions of case folding already developed for identifier case conversion. I left the straight locale-based folding in place for situations where we are just manipulating user data and not trying to match it to built-in strings --- for example, the SQL upper() function is still locale dependent. Perhaps this will prove not to be what's wanted, but at the moment we can initdb and pass regression tests in Turkish locale.
2004-05-07 02:24:59 +02:00
char *lowunits;
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;
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
struct pg_tm tt,
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
*tm = &tt;
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
lowunits = downcase_truncate_identifier(VARDATA_ANY(units),
VARSIZE_ANY_EXHDR(units),
Solve the 'Turkish problem' with undesirable locale behavior for case conversion of basic ASCII letters. Remove all uses of strcasecmp and strncasecmp in favor of new functions pg_strcasecmp and pg_strncasecmp; remove most but not all direct uses of toupper and tolower in favor of pg_toupper and pg_tolower. These functions use the same notions of case folding already developed for identifier case conversion. I left the straight locale-based folding in place for situations where we are just manipulating user data and not trying to match it to built-in strings --- for example, the SQL upper() function is still locale dependent. Perhaps this will prove not to be what's wanted, but at the moment we can initdb and pass regression tests in Turkish locale.
2004-05-07 02:24:59 +02:00
false);
type = DecodeUnits(0, lowunits, &val);
if (type == UNKNOWN_FIELD)
type = DecodeSpecial(0, lowunits, &val);
Make extract() do something more reasonable with infinite datetimes. Historically, extract() just returned zero for any case involving an infinite timestamp[tz] input; even cases in which the unit name was invalid. This is not very sensible. Instead, return infinity or -infinity as appropriate when the requested field is one that is monotonically increasing (e.g, year, epoch), or NULL when it is not (e.g., day, hour). Also, throw the expected errors for bad unit names. BACKWARDS INCOMPATIBLE CHANGE Vitaly Burovoy, reviewed by Vik Fearing
2016-01-22 04:26:20 +01:00
if (TIMESTAMP_NOT_FINITE(timestamp))
{
result = NonFiniteTimestampTzPart(type, val, lowunits,
TIMESTAMP_IS_NOBEGIN(timestamp),
false);
if (result)
PG_RETURN_FLOAT8(result);
else
PG_RETURN_NULL();
}
Error message editing in utils/adt. Again thanks to Joe Conway for doing the bulk of the heavy lifting ...
2003-07-27 06:53:12 +02:00
if (type == UNITS)
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 unnecessary parentheses in assignments. Add spaces where needed. Reference time interval variables as tinterval.
2005-07-21 06:41:43 +02:00
if (timestamp2tm(timestamp, NULL, tm, &fsec, NULL, NULL) != 0)
Error message editing in utils/adt. Again thanks to Joe Conway for doing the bulk of the heavy lifting ...
2003-07-27 06:53:12 +02:00
ereport(ERROR,
(errcode(ERRCODE_DATETIME_VALUE_OUT_OF_RANGE),
errmsg("timestamp out of range")));
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
switch (val)
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +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
case DTK_MICROSEC:
Change 5e0 to 5.0, for consistency.
2005-07-12 17:17:44 +02:00
result = tm->tm_sec * 1000000.0 + fsec;
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;
case DTK_MILLISEC:
Change 5e0 to 5.0, for consistency.
2005-07-12 17:17:44 +02:00
result = tm->tm_sec * 1000.0 + fsec / 1000.0;
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;
case DTK_SECOND:
Change 5e0 to 5.0, for consistency.
2005-07-12 17:17:44 +02:00
result = tm->tm_sec + fsec / 1000000.0;
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;
case DTK_MINUTE:
result = tm->tm_min;
break;
case DTK_HOUR:
result = tm->tm_hour;
break;
case DTK_DAY:
result = tm->tm_mday;
break;
case DTK_MONTH:
result = tm->tm_mon;
break;
case DTK_QUARTER:
Remove unnecessary parentheses in datetime/timestamp code.
2005-05-23 19:13:14 +02:00
result = (tm->tm_mon - 1) / 3 + 1;
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;
case DTK_WEEK:
result = (float8) date2isoweek(tm->tm_year, tm->tm_mon, tm->tm_mday);
break;
case DTK_YEAR:
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
if (tm->tm_year > 0)
result = tm->tm_year;
else
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
/* there is no year 0, just 1 BC and 1 AD */
result = tm->tm_year - 1;
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;
case DTK_DECADE:
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
/*
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* what is a decade wrt dates? let us assume that decade 199
* is 1990 thru 1999... decade 0 starts on year 1 BC, and -1
* is 11 BC thru 2 BC...
> After all that about numbering centuries and millenia correctly, > why does CVS tip still give me > > regression=# select extract(century from now()); > date_part > ----------- > 20 > (1 row) > [ ... looks in code ... ] > > Apparently it's because you fixed only timestamp_part, and not > timestamptz_part. I'm not too sure about what timestamp_trunc or > timestamptz_trunc should do, but they may be wrong as well. Sigh... as usual, what is not tested does not work:-( > Could we have a more complete patch? Please find a submission attached. I hope it really fixes all decade, century and millenium issues for extract and *_trunc functions on interval and other timestamp types. If someone could check that the results are reasonnable, it would be great. I indeed overlooked the fact that there were two functions. The patch fixes the code so that both variants agree. I added comments to interval extractions, because it relies on the C division to have a negative remainder: -7/10 = 0 and remains -7. As for *_trunc functions, I have chosen to put the first year of the century or millennium: -100, 1, 101... 1001 2001 etc. Indeed, I don't think it would make sense to put 2000 (last year of the 2nd millennium) for rounding all years of the third millenium. I also fixed the code so that all decades last 10 years and decade 199 means the 1990's. I have added some tests that are relevant to deal with tricky cases. The formula may be simplified, but all these cases must pass. Please keep them. Fabien Coelho
2004-08-20 05:45:14 +02:00
*/
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
if (tm->tm_year >= 0)
Remove unnecessary parentheses in assignments. Add spaces where needed. Reference time interval variables as tinterval.
2005-07-21 06:41:43 +02:00
result = tm->tm_year / 10;
> After all that about numbering centuries and millenia correctly, > why does CVS tip still give me > > regression=# select extract(century from now()); > date_part > ----------- > 20 > (1 row) > [ ... looks in code ... ] > > Apparently it's because you fixed only timestamp_part, and not > timestamptz_part. I'm not too sure about what timestamp_trunc or > timestamptz_trunc should do, but they may be wrong as well. Sigh... as usual, what is not tested does not work:-( > Could we have a more complete patch? Please find a submission attached. I hope it really fixes all decade, century and millenium issues for extract and *_trunc functions on interval and other timestamp types. If someone could check that the results are reasonnable, it would be great. I indeed overlooked the fact that there were two functions. The patch fixes the code so that both variants agree. I added comments to interval extractions, because it relies on the C division to have a negative remainder: -7/10 = 0 and remains -7. As for *_trunc functions, I have chosen to put the first year of the century or millennium: -100, 1, 101... 1001 2001 etc. Indeed, I don't think it would make sense to put 2000 (last year of the 2nd millennium) for rounding all years of the third millenium. I also fixed the code so that all decades last 10 years and decade 199 means the 1990's. I have added some tests that are relevant to deal with tricky cases. The formula may be simplified, but all these cases must pass. Please keep them. Fabien Coelho
2004-08-20 05:45:14 +02:00
else
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
result = -((8 - (tm->tm_year - 1)) / 10);
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;
case DTK_CENTURY:
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
Remove unnecessary parentheses in datetime/timestamp code.
2005-05-23 19:13:14 +02:00
/* ----
* centuries AD, c>0: year in [ (c-1)* 100 + 1 : c*100 ]
* centuries BC, c<0: year in [ c*100 : (c+1) * 100 - 1]
* there is no number 0 century.
* ----
Please find a small patch to fix the brain damage "century" and "millennium" date part implementation in postgresql, both in the code and the documentation, so that it conforms to the official definition. If you do not agree with the official definition, please send your complaint to "pope@vatican.org". I'm not responsible for them;-) With the previous version, the centuries and millenniums had a wrong number and started the wrong year. Moreover century number 0, which does not exist in reality, lasted 200 years. Also, millennium number 0 lasted 2000 years. If you want postgresql to have it's own definition of "century" and "millennium" that does not conform to the one of the society, just give them another name. I would suggest "pgCENTURY" and "pgMILLENNIUM";-) IMO, if someone may use the options, it means that postgresql is used for historical data, so it make sense to have an historical definition. Also, I just want to divide the year by 100 or 1000, I can do that quite easily. BACKWARD INCOMPATIBLE CHANGE Fabien Coelho - coelho@cri.ensmp.fr
2004-04-10 20:02:59 +02:00
*/
if (tm->tm_year > 0)
Remove more extraneous parentheses in date/time functions.
2005-05-24 04:09:45 +02:00
result = (tm->tm_year + 99) / 100;
Please find a small patch to fix the brain damage "century" and "millennium" date part implementation in postgresql, both in the code and the documentation, so that it conforms to the official definition. If you do not agree with the official definition, please send your complaint to "pope@vatican.org". I'm not responsible for them;-) With the previous version, the centuries and millenniums had a wrong number and started the wrong year. Moreover century number 0, which does not exist in reality, lasted 200 years. Also, millennium number 0 lasted 2000 years. If you want postgresql to have it's own definition of "century" and "millennium" that does not conform to the one of the society, just give them another name. I would suggest "pgCENTURY" and "pgMILLENNIUM";-) IMO, if someone may use the options, it means that postgresql is used for historical data, so it make sense to have an historical definition. Also, I just want to divide the year by 100 or 1000, I can do that quite easily. BACKWARD INCOMPATIBLE CHANGE Fabien Coelho - coelho@cri.ensmp.fr
2004-04-10 20:02:59 +02:00
else
> After all that about numbering centuries and millenia correctly, > why does CVS tip still give me > > regression=# select extract(century from now()); > date_part > ----------- > 20 > (1 row) > [ ... looks in code ... ] > > Apparently it's because you fixed only timestamp_part, and not > timestamptz_part. I'm not too sure about what timestamp_trunc or > timestamptz_trunc should do, but they may be wrong as well. Sigh... as usual, what is not tested does not work:-( > Could we have a more complete patch? Please find a submission attached. I hope it really fixes all decade, century and millenium issues for extract and *_trunc functions on interval and other timestamp types. If someone could check that the results are reasonnable, it would be great. I indeed overlooked the fact that there were two functions. The patch fixes the code so that both variants agree. I added comments to interval extractions, because it relies on the C division to have a negative remainder: -7/10 = 0 and remains -7. As for *_trunc functions, I have chosen to put the first year of the century or millennium: -100, 1, 101... 1001 2001 etc. Indeed, I don't think it would make sense to put 2000 (last year of the 2nd millennium) for rounding all years of the third millenium. I also fixed the code so that all decades last 10 years and decade 199 means the 1990's. I have added some tests that are relevant to deal with tricky cases. The formula may be simplified, but all these cases must pass. Please keep them. Fabien Coelho
2004-08-20 05:45:14 +02:00
/* caution: C division may have negative remainder */
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
result = -((99 - (tm->tm_year - 1)) / 100);
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;
case DTK_MILLENNIUM:
Please find a small patch to fix the brain damage "century" and "millennium" date part implementation in postgresql, both in the code and the documentation, so that it conforms to the official definition. If you do not agree with the official definition, please send your complaint to "pope@vatican.org". I'm not responsible for them;-) With the previous version, the centuries and millenniums had a wrong number and started the wrong year. Moreover century number 0, which does not exist in reality, lasted 200 years. Also, millennium number 0 lasted 2000 years. If you want postgresql to have it's own definition of "century" and "millennium" that does not conform to the one of the society, just give them another name. I would suggest "pgCENTURY" and "pgMILLENNIUM";-) IMO, if someone may use the options, it means that postgresql is used for historical data, so it make sense to have an historical definition. Also, I just want to divide the year by 100 or 1000, I can do that quite easily. BACKWARD INCOMPATIBLE CHANGE Fabien Coelho - coelho@cri.ensmp.fr
2004-04-10 20:02:59 +02:00
/* see comments above. */
if (tm->tm_year > 0)
Remove unnecessary parentheses in assignments. Add spaces where needed. Reference time interval variables as tinterval.
2005-07-21 06:41:43 +02:00
result = (tm->tm_year + 999) / 1000;
Please find a small patch to fix the brain damage "century" and "millennium" date part implementation in postgresql, both in the code and the documentation, so that it conforms to the official definition. If you do not agree with the official definition, please send your complaint to "pope@vatican.org". I'm not responsible for them;-) With the previous version, the centuries and millenniums had a wrong number and started the wrong year. Moreover century number 0, which does not exist in reality, lasted 200 years. Also, millennium number 0 lasted 2000 years. If you want postgresql to have it's own definition of "century" and "millennium" that does not conform to the one of the society, just give them another name. I would suggest "pgCENTURY" and "pgMILLENNIUM";-) IMO, if someone may use the options, it means that postgresql is used for historical data, so it make sense to have an historical definition. Also, I just want to divide the year by 100 or 1000, I can do that quite easily. BACKWARD INCOMPATIBLE CHANGE Fabien Coelho - coelho@cri.ensmp.fr
2004-04-10 20:02:59 +02:00
else
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
result = -((999 - (tm->tm_year - 1)) / 1000);
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_JULIAN:
result = date2j(tm->tm_year, tm->tm_mon, tm->tm_mday);
Update DAYS_PER_MONTH comment. Add SECS_PER_YEAR and MINS_PER_HOUR macros.
2005-07-21 20:06:13 +02:00
result += ((((tm->tm_hour * MINS_PER_HOUR) + tm->tm_min) * SECS_PER_MINUTE) +
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_sec + (fsec / 1000000.0)) / (double) SECS_PER_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
break;
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
case DTK_ISOYEAR:
result = date2isoyear(tm->tm_year, tm->tm_mon, tm->tm_mday);
Fix EXTRACT(ISOYEAR FROM timestamp) for years BC. The test cases added by commit 26ae3aa80 exposed an old oversight in timestamp[tz]_part: they didn't correct the result of date2isoyear() for BC years, so that we produced an off-by-one answer for such years. Fix that, and back-patch to all supported branches. Discussion: https://postgr.es/m/SG2PR06MB37762CAE45DB0F6CA7001EA9B6550@SG2PR06MB3776.apcprd06.prod.outlook.com
2019-12-12 18:30:43 +01:00
/* Adjust BC years */
if (result <= 0)
result -= 1;
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
break;
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
case DTK_DOW:
case DTK_ISODOW:
result = j2day(date2j(tm->tm_year, tm->tm_mon, tm->tm_mday));
if (val == DTK_ISODOW && result == 0)
result = 7;
break;
case DTK_DOY:
result = (date2j(tm->tm_year, tm->tm_mon, tm->tm_mday)
- date2j(tm->tm_year, 1, 1) + 1);
break;
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
case DTK_TZ:
case DTK_TZ_MINUTE:
case DTK_TZ_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
default:
Error message editing in utils/adt. Again thanks to Joe Conway for doing the bulk of the heavy lifting ...
2003-07-27 06:53:12 +02:00
ereport(ERROR,
(errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
errmsg("timestamp units \"%s\" not supported",
pgindent run.
2003-08-04 02:43:34 +02:00
lowunits)));
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
result = 0;
}
}
else if (type == RESERV)
{
switch (val)
{
case DTK_EPOCH:
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
epoch = SetEpochTimestamp();
/* try to avoid precision loss in subtraction */
if (timestamp < (PG_INT64_MAX + epoch))
result = (timestamp - epoch) / 1000000.0;
else
result = ((float8) timestamp - epoch) / 1000000.0;
Run pgindent on 9.2 source tree in preparation for first 9.3 commit-fest.
2012-06-10 21:20:04 +02:00
break;
Measure epoch of timestamp-without-time-zone from local not UTC midnight. This patch reverts commit 191ef2b407f065544ceed5700e42400857d9270f and thereby restores the pre-7.3 behavior of EXTRACT(EPOCH FROM timestamp-without-tz). Per discussion, the more recent behavior was misguided on a couple of grounds: it makes it hard to get a non-timezone-aware epoch value for a timestamp, and it makes this one case dependent on the value of the timezone GUC, which is incompatible with having timestamp_part() labeled as immutable. The other behavior is still available (in all releases) by explicitly casting the timestamp to timestamp with time zone before applying EXTRACT. This will need to be called out as an incompatible change in the 9.2 release notes. Although having mutable behavior in a function marked immutable is clearly a bug, we're not going to back-patch such a change.
2012-04-10 18:04:42 +02:00
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
default:
Error message editing in utils/adt. Again thanks to Joe Conway for doing the bulk of the heavy lifting ...
2003-07-27 06:53:12 +02:00
ereport(ERROR,
(errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
errmsg("timestamp units \"%s\" not supported",
pgindent run.
2003-08-04 02:43:34 +02:00
lowunits)));
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
result = 0;
}
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +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
{
Error message editing in utils/adt. Again thanks to Joe Conway for doing the bulk of the heavy lifting ...
2003-07-27 06:53:12 +02:00
ereport(ERROR,
(errcode(ERRCODE_INVALID_PARAMETER_VALUE),
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
errmsg("timestamp units \"%s\" not recognized", lowunits)));
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
result = 0;
}
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
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
PG_RETURN_FLOAT8(result);
}
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +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
/* timestamptz_part()
* Extract specified field from timestamp with time zone.
*/
Datum
timestamptz_part(PG_FUNCTION_ARGS)
{
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
text *units = PG_GETARG_TEXT_PP(0);
Add binary I/O routines for a bunch more datatypes. Still a few to go, but that was enough tedium for one day. Along the way, move the few support routines for types xid and cid into a more logical place.
2003-05-13 01:08:52 +02:00
TimestampTz timestamp = PG_GETARG_TIMESTAMPTZ(1);
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
float8 result;
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
Timestamp epoch;
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
int tz;
int type,
val;
Solve the 'Turkish problem' with undesirable locale behavior for case conversion of basic ASCII letters. Remove all uses of strcasecmp and strncasecmp in favor of new functions pg_strcasecmp and pg_strncasecmp; remove most but not all direct uses of toupper and tolower in favor of pg_toupper and pg_tolower. These functions use the same notions of case folding already developed for identifier case conversion. I left the straight locale-based folding in place for situations where we are just manipulating user data and not trying to match it to built-in strings --- for example, the SQL upper() function is still locale dependent. Perhaps this will prove not to be what's wanted, but at the moment we can initdb and pass regression tests in Turkish locale.
2004-05-07 02:24:59 +02:00
char *lowunits;
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
double dummy;
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;
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
struct pg_tm tt,
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 = &tt;
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
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
lowunits = downcase_truncate_identifier(VARDATA_ANY(units),
VARSIZE_ANY_EXHDR(units),
Solve the 'Turkish problem' with undesirable locale behavior for case conversion of basic ASCII letters. Remove all uses of strcasecmp and strncasecmp in favor of new functions pg_strcasecmp and pg_strncasecmp; remove most but not all direct uses of toupper and tolower in favor of pg_toupper and pg_tolower. These functions use the same notions of case folding already developed for identifier case conversion. I left the straight locale-based folding in place for situations where we are just manipulating user data and not trying to match it to built-in strings --- for example, the SQL upper() function is still locale dependent. Perhaps this will prove not to be what's wanted, but at the moment we can initdb and pass regression tests in Turkish locale.
2004-05-07 02:24:59 +02:00
false);
type = DecodeUnits(0, lowunits, &val);
if (type == UNKNOWN_FIELD)
type = DecodeSpecial(0, lowunits, &val);
Make extract() do something more reasonable with infinite datetimes. Historically, extract() just returned zero for any case involving an infinite timestamp[tz] input; even cases in which the unit name was invalid. This is not very sensible. Instead, return infinity or -infinity as appropriate when the requested field is one that is monotonically increasing (e.g, year, epoch), or NULL when it is not (e.g., day, hour). Also, throw the expected errors for bad unit names. BACKWARDS INCOMPATIBLE CHANGE Vitaly Burovoy, reviewed by Vik Fearing
2016-01-22 04:26:20 +01:00
if (TIMESTAMP_NOT_FINITE(timestamp))
{
result = NonFiniteTimestampTzPart(type, val, lowunits,
TIMESTAMP_IS_NOBEGIN(timestamp),
true);
if (result)
PG_RETURN_FLOAT8(result);
else
PG_RETURN_NULL();
}
Error message editing in utils/adt. Again thanks to Joe Conway for doing the bulk of the heavy lifting ...
2003-07-27 06:53:12 +02:00
if (type == UNITS)
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 unused tzn arguments for timestamp2tm()
2012-03-15 20:13:35 +01:00
if (timestamp2tm(timestamp, &tz, tm, &fsec, NULL, NULL) != 0)
Error message editing in utils/adt. Again thanks to Joe Conway for doing the bulk of the heavy lifting ...
2003-07-27 06:53:12 +02:00
ereport(ERROR,
(errcode(ERRCODE_DATETIME_VALUE_OUT_OF_RANGE),
errmsg("timestamp out of range")));
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
switch (val)
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +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
case DTK_TZ:
Make EXTRACT(TIMEZONE) and SET/SHOW TIMEZONE follow the SQL convention for the sign of timezone offsets, ie, positive is east from UTC. These were previously out of step with other operations that accept or show timezones, such as I/O of timestamptz values.
2003-07-17 02:55:37 +02:00
result = -tz;
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;
case DTK_TZ_MINUTE:
Make EXTRACT(TIMEZONE) and SET/SHOW TIMEZONE follow the SQL convention for the sign of timezone offsets, ie, positive is east from UTC. These were previously out of step with other operations that accept or show timezones, such as I/O of timestamptz values.
2003-07-17 02:55:37 +02:00
result = -tz;
Update DAYS_PER_MONTH comment. Add SECS_PER_YEAR and MINS_PER_HOUR macros.
2005-07-21 20:06:13 +02:00
result /= MINS_PER_HOUR;
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
FMODULO(result, dummy, (double) MINS_PER_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;
case DTK_TZ_HOUR:
Make EXTRACT(TIMEZONE) and SET/SHOW TIMEZONE follow the SQL convention for the sign of timezone offsets, ie, positive is east from UTC. These were previously out of step with other operations that accept or show timezones, such as I/O of timestamptz values.
2003-07-17 02:55:37 +02:00
dummy = -tz;
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
FMODULO(dummy, result, (double) SECS_PER_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;
case DTK_MICROSEC:
Change 5e0 to 5.0, for consistency.
2005-07-12 17:17:44 +02:00
result = tm->tm_sec * 1000000.0 + fsec;
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;
case DTK_MILLISEC:
Change 5e0 to 5.0, for consistency.
2005-07-12 17:17:44 +02:00
result = tm->tm_sec * 1000.0 + fsec / 1000.0;
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;
case DTK_SECOND:
Change 5e0 to 5.0, for consistency.
2005-07-12 17:17:44 +02:00
result = tm->tm_sec + fsec / 1000000.0;
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;
case DTK_MINUTE:
result = tm->tm_min;
break;
case DTK_HOUR:
result = tm->tm_hour;
break;
case DTK_DAY:
result = tm->tm_mday;
break;
case DTK_MONTH:
result = tm->tm_mon;
break;
case DTK_QUARTER:
Remove unnecessary parentheses in datetime/timestamp code.
2005-05-23 19:13:14 +02:00
result = (tm->tm_mon - 1) / 3 + 1;
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;
case DTK_WEEK:
result = (float8) date2isoweek(tm->tm_year, tm->tm_mon, tm->tm_mday);
break;
case DTK_YEAR:
Patch of 2004-03-30 corrected date_part(timestamp) for extracting the year from a BC date, but failed to make the same fix in date_part(timestamptz).
2004-11-20 23:12:44 +01:00
if (tm->tm_year > 0)
result = tm->tm_year;
else
/* there is no year 0, just 1 BC and 1 AD */
result = tm->tm_year - 1;
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;
case DTK_DECADE:
> After all that about numbering centuries and millenia correctly, > why does CVS tip still give me > > regression=# select extract(century from now()); > date_part > ----------- > 20 > (1 row) > [ ... looks in code ... ] > > Apparently it's because you fixed only timestamp_part, and not > timestamptz_part. I'm not too sure about what timestamp_trunc or > timestamptz_trunc should do, but they may be wrong as well. Sigh... as usual, what is not tested does not work:-( > Could we have a more complete patch? Please find a submission attached. I hope it really fixes all decade, century and millenium issues for extract and *_trunc functions on interval and other timestamp types. If someone could check that the results are reasonnable, it would be great. I indeed overlooked the fact that there were two functions. The patch fixes the code so that both variants agree. I added comments to interval extractions, because it relies on the C division to have a negative remainder: -7/10 = 0 and remains -7. As for *_trunc functions, I have chosen to put the first year of the century or millennium: -100, 1, 101... 1001 2001 etc. Indeed, I don't think it would make sense to put 2000 (last year of the 2nd millennium) for rounding all years of the third millenium. I also fixed the code so that all decades last 10 years and decade 199 means the 1990's. I have added some tests that are relevant to deal with tricky cases. The formula may be simplified, but all these cases must pass. Please keep them. Fabien Coelho
2004-08-20 05:45:14 +02:00
/* see comments in timestamp_part */
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
if (tm->tm_year > 0)
Remove unnecessary parentheses in assignments. Add spaces where needed. Reference time interval variables as tinterval.
2005-07-21 06:41:43 +02:00
result = tm->tm_year / 10;
> After all that about numbering centuries and millenia correctly, > why does CVS tip still give me > > regression=# select extract(century from now()); > date_part > ----------- > 20 > (1 row) > [ ... looks in code ... ] > > Apparently it's because you fixed only timestamp_part, and not > timestamptz_part. I'm not too sure about what timestamp_trunc or > timestamptz_trunc should do, but they may be wrong as well. Sigh... as usual, what is not tested does not work:-( > Could we have a more complete patch? Please find a submission attached. I hope it really fixes all decade, century and millenium issues for extract and *_trunc functions on interval and other timestamp types. If someone could check that the results are reasonnable, it would be great. I indeed overlooked the fact that there were two functions. The patch fixes the code so that both variants agree. I added comments to interval extractions, because it relies on the C division to have a negative remainder: -7/10 = 0 and remains -7. As for *_trunc functions, I have chosen to put the first year of the century or millennium: -100, 1, 101... 1001 2001 etc. Indeed, I don't think it would make sense to put 2000 (last year of the 2nd millennium) for rounding all years of the third millenium. I also fixed the code so that all decades last 10 years and decade 199 means the 1990's. I have added some tests that are relevant to deal with tricky cases. The formula may be simplified, but all these cases must pass. Please keep them. Fabien Coelho
2004-08-20 05:45:14 +02:00
else
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
result = -((8 - (tm->tm_year - 1)) / 10);
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;
case DTK_CENTURY:
> After all that about numbering centuries and millenia correctly, > why does CVS tip still give me > > regression=# select extract(century from now()); > date_part > ----------- > 20 > (1 row) > [ ... looks in code ... ] > > Apparently it's because you fixed only timestamp_part, and not > timestamptz_part. I'm not too sure about what timestamp_trunc or > timestamptz_trunc should do, but they may be wrong as well. Sigh... as usual, what is not tested does not work:-( > Could we have a more complete patch? Please find a submission attached. I hope it really fixes all decade, century and millenium issues for extract and *_trunc functions on interval and other timestamp types. If someone could check that the results are reasonnable, it would be great. I indeed overlooked the fact that there were two functions. The patch fixes the code so that both variants agree. I added comments to interval extractions, because it relies on the C division to have a negative remainder: -7/10 = 0 and remains -7. As for *_trunc functions, I have chosen to put the first year of the century or millennium: -100, 1, 101... 1001 2001 etc. Indeed, I don't think it would make sense to put 2000 (last year of the 2nd millennium) for rounding all years of the third millenium. I also fixed the code so that all decades last 10 years and decade 199 means the 1990's. I have added some tests that are relevant to deal with tricky cases. The formula may be simplified, but all these cases must pass. Please keep them. Fabien Coelho
2004-08-20 05:45:14 +02:00
/* see comments in timestamp_part */
if (tm->tm_year > 0)
Remove unnecessary parentheses in datetime/timestamp code.
2005-05-23 19:13:14 +02:00
result = (tm->tm_year + 99) / 100;
> After all that about numbering centuries and millenia correctly, > why does CVS tip still give me > > regression=# select extract(century from now()); > date_part > ----------- > 20 > (1 row) > [ ... looks in code ... ] > > Apparently it's because you fixed only timestamp_part, and not > timestamptz_part. I'm not too sure about what timestamp_trunc or > timestamptz_trunc should do, but they may be wrong as well. Sigh... as usual, what is not tested does not work:-( > Could we have a more complete patch? Please find a submission attached. I hope it really fixes all decade, century and millenium issues for extract and *_trunc functions on interval and other timestamp types. If someone could check that the results are reasonnable, it would be great. I indeed overlooked the fact that there were two functions. The patch fixes the code so that both variants agree. I added comments to interval extractions, because it relies on the C division to have a negative remainder: -7/10 = 0 and remains -7. As for *_trunc functions, I have chosen to put the first year of the century or millennium: -100, 1, 101... 1001 2001 etc. Indeed, I don't think it would make sense to put 2000 (last year of the 2nd millennium) for rounding all years of the third millenium. I also fixed the code so that all decades last 10 years and decade 199 means the 1990's. I have added some tests that are relevant to deal with tricky cases. The formula may be simplified, but all these cases must pass. Please keep them. Fabien Coelho
2004-08-20 05:45:14 +02:00
else
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
result = -((99 - (tm->tm_year - 1)) / 100);
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;
case DTK_MILLENNIUM:
> After all that about numbering centuries and millenia correctly, > why does CVS tip still give me > > regression=# select extract(century from now()); > date_part > ----------- > 20 > (1 row) > [ ... looks in code ... ] > > Apparently it's because you fixed only timestamp_part, and not > timestamptz_part. I'm not too sure about what timestamp_trunc or > timestamptz_trunc should do, but they may be wrong as well. Sigh... as usual, what is not tested does not work:-( > Could we have a more complete patch? Please find a submission attached. I hope it really fixes all decade, century and millenium issues for extract and *_trunc functions on interval and other timestamp types. If someone could check that the results are reasonnable, it would be great. I indeed overlooked the fact that there were two functions. The patch fixes the code so that both variants agree. I added comments to interval extractions, because it relies on the C division to have a negative remainder: -7/10 = 0 and remains -7. As for *_trunc functions, I have chosen to put the first year of the century or millennium: -100, 1, 101... 1001 2001 etc. Indeed, I don't think it would make sense to put 2000 (last year of the 2nd millennium) for rounding all years of the third millenium. I also fixed the code so that all decades last 10 years and decade 199 means the 1990's. I have added some tests that are relevant to deal with tricky cases. The formula may be simplified, but all these cases must pass. Please keep them. Fabien Coelho
2004-08-20 05:45:14 +02:00
/* see comments in timestamp_part */
if (tm->tm_year > 0)
Remove unnecessary parentheses in datetime/timestamp code.
2005-05-23 19:13:14 +02:00
result = (tm->tm_year + 999) / 1000;
> After all that about numbering centuries and millenia correctly, > why does CVS tip still give me > > regression=# select extract(century from now()); > date_part > ----------- > 20 > (1 row) > [ ... looks in code ... ] > > Apparently it's because you fixed only timestamp_part, and not > timestamptz_part. I'm not too sure about what timestamp_trunc or > timestamptz_trunc should do, but they may be wrong as well. Sigh... as usual, what is not tested does not work:-( > Could we have a more complete patch? Please find a submission attached. I hope it really fixes all decade, century and millenium issues for extract and *_trunc functions on interval and other timestamp types. If someone could check that the results are reasonnable, it would be great. I indeed overlooked the fact that there were two functions. The patch fixes the code so that both variants agree. I added comments to interval extractions, because it relies on the C division to have a negative remainder: -7/10 = 0 and remains -7. As for *_trunc functions, I have chosen to put the first year of the century or millennium: -100, 1, 101... 1001 2001 etc. Indeed, I don't think it would make sense to put 2000 (last year of the 2nd millennium) for rounding all years of the third millenium. I also fixed the code so that all decades last 10 years and decade 199 means the 1990's. I have added some tests that are relevant to deal with tricky cases. The formula may be simplified, but all these cases must pass. Please keep them. Fabien Coelho
2004-08-20 05:45:14 +02:00
else
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
result = -((999 - (tm->tm_year - 1)) / 1000);
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_JULIAN:
result = date2j(tm->tm_year, tm->tm_mon, tm->tm_mday);
Update DAYS_PER_MONTH comment. Add SECS_PER_YEAR and MINS_PER_HOUR macros.
2005-07-21 20:06:13 +02:00
result += ((((tm->tm_hour * MINS_PER_HOUR) + tm->tm_min) * SECS_PER_MINUTE) +
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_sec + (fsec / 1000000.0)) / (double) SECS_PER_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
break;
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
case DTK_ISOYEAR:
result = date2isoyear(tm->tm_year, tm->tm_mon, tm->tm_mday);
Fix EXTRACT(ISOYEAR FROM timestamp) for years BC. The test cases added by commit 26ae3aa80 exposed an old oversight in timestamp[tz]_part: they didn't correct the result of date2isoyear() for BC years, so that we produced an off-by-one answer for such years. Fix that, and back-patch to all supported branches. Discussion: https://postgr.es/m/SG2PR06MB37762CAE45DB0F6CA7001EA9B6550@SG2PR06MB3776.apcprd06.prod.outlook.com
2019-12-12 18:30:43 +01:00
/* Adjust BC years */
if (result <= 0)
result -= 1;
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
break;
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
case DTK_DOW:
case DTK_ISODOW:
result = j2day(date2j(tm->tm_year, tm->tm_mon, tm->tm_mday));
if (val == DTK_ISODOW && result == 0)
result = 7;
break;
case DTK_DOY:
result = (date2j(tm->tm_year, tm->tm_mon, tm->tm_mday)
- date2j(tm->tm_year, 1, 1) + 1);
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
default:
Error message editing in utils/adt. Again thanks to Joe Conway for doing the bulk of the heavy lifting ...
2003-07-27 06:53:12 +02:00
ereport(ERROR,
(errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
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
errmsg("timestamp with time zone units \"%s\" not supported",
lowunits)));
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
result = 0;
}
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +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 if (type == RESERV)
{
switch (val)
{
case DTK_EPOCH:
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
epoch = SetEpochTimestamp();
/* try to avoid precision loss in subtraction */
if (timestamp < (PG_INT64_MAX + epoch))
result = (timestamp - epoch) / 1000000.0;
else
result = ((float8) timestamp - epoch) / 1000000.0;
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;
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +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
default:
Error message editing in utils/adt. Again thanks to Joe Conway for doing the bulk of the heavy lifting ...
2003-07-27 06:53:12 +02:00
ereport(ERROR,
(errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
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
errmsg("timestamp with time zone units \"%s\" not supported",
lowunits)));
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
result = 0;
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +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
{
Error message editing in utils/adt. Again thanks to Joe Conway for doing the bulk of the heavy lifting ...
2003-07-27 06:53:12 +02:00
ereport(ERROR,
(errcode(ERRCODE_INVALID_PARAMETER_VALUE),
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
errmsg("timestamp with time zone units \"%s\" not recognized",
lowunits)));
Error message editing in utils/adt. Again thanks to Joe Conway for doing the bulk of the heavy lifting ...
2003-07-27 06:53:12 +02:00
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
result = 0;
}
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
PG_RETURN_FLOAT8(result);
}
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
/* interval_part()
* Extract specified field from interval.
*/
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
Datum
interval_part(PG_FUNCTION_ARGS)
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
{
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
text *units = PG_GETARG_TEXT_PP(0);
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
Interval *interval = PG_GETARG_INTERVAL_P(1);
float8 result;
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
int type,
val;
Solve the 'Turkish problem' with undesirable locale behavior for case conversion of basic ASCII letters. Remove all uses of strcasecmp and strncasecmp in favor of new functions pg_strcasecmp and pg_strncasecmp; remove most but not all direct uses of toupper and tolower in favor of pg_toupper and pg_tolower. These functions use the same notions of case folding already developed for identifier case conversion. I left the straight locale-based folding in place for situations where we are just manipulating user data and not trying to match it to built-in strings --- for example, the SQL upper() function is still locale dependent. Perhaps this will prove not to be what's wanted, but at the moment we can initdb and pass regression tests in Turkish locale.
2004-05-07 02:24:59 +02:00
char *lowunits;
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;
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
struct pg_tm tt,
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
*tm = &tt;
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
lowunits = downcase_truncate_identifier(VARDATA_ANY(units),
VARSIZE_ANY_EXHDR(units),
Solve the 'Turkish problem' with undesirable locale behavior for case conversion of basic ASCII letters. Remove all uses of strcasecmp and strncasecmp in favor of new functions pg_strcasecmp and pg_strncasecmp; remove most but not all direct uses of toupper and tolower in favor of pg_toupper and pg_tolower. These functions use the same notions of case folding already developed for identifier case conversion. I left the straight locale-based folding in place for situations where we are just manipulating user data and not trying to match it to built-in strings --- for example, the SQL upper() function is still locale dependent. Perhaps this will prove not to be what's wanted, but at the moment we can initdb and pass regression tests in Turkish locale.
2004-05-07 02:24:59 +02:00
false);
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
type = DecodeUnits(0, lowunits, &val);
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
if (type == UNKNOWN_FIELD)
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
type = DecodeSpecial(0, lowunits, &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
if (type == UNITS)
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
{
if (interval2tm(*interval, tm, &fsec) == 0)
{
switch (val)
{
pgindent run.
2002-09-04 22:31:48 +02:00
case DTK_MICROSEC:
Change 5e0 to 5.0, for consistency.
2005-07-12 17:17:44 +02:00
result = tm->tm_sec * 1000000.0 + fsec;
pgindent run.
2002-09-04 22:31:48 +02:00
break;
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
pgindent run.
2002-09-04 22:31:48 +02:00
case DTK_MILLISEC:
Change 5e0 to 5.0, for consistency.
2005-07-12 17:17:44 +02:00
result = tm->tm_sec * 1000.0 + fsec / 1000.0;
pgindent run.
2002-09-04 22:31:48 +02:00
break;
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
pgindent run.
2002-09-04 22:31:48 +02:00
case DTK_SECOND:
Change 5e0 to 5.0, for consistency.
2005-07-12 17:17:44 +02:00
result = tm->tm_sec + fsec / 1000000.0;
pgindent run.
2002-09-04 22:31:48 +02:00
break;
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
case DTK_MINUTE:
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
result = tm->tm_min;
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
break;
case DTK_HOUR:
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
result = tm->tm_hour;
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
break;
case DTK_DAY:
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
result = tm->tm_mday;
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
break;
case DTK_MONTH:
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
result = tm->tm_mon;
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
break;
case DTK_QUARTER:
I corecting date_trunc('quarter',...) and friends because orig version doing '2003-07-30' -> '2003-04-01', '2003-11-30' ->'2003-07-01' B?jthe Zolt?n
2003-07-26 17:17:36 +02:00
result = (tm->tm_mon / 3) + 1;
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
break;
case DTK_YEAR:
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
result = tm->tm_year;
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
break;
case DTK_DECADE:
> After all that about numbering centuries and millenia correctly, > why does CVS tip still give me > > regression=# select extract(century from now()); > date_part > ----------- > 20 > (1 row) > [ ... looks in code ... ] > > Apparently it's because you fixed only timestamp_part, and not > timestamptz_part. I'm not too sure about what timestamp_trunc or > timestamptz_trunc should do, but they may be wrong as well. Sigh... as usual, what is not tested does not work:-( > Could we have a more complete patch? Please find a submission attached. I hope it really fixes all decade, century and millenium issues for extract and *_trunc functions on interval and other timestamp types. If someone could check that the results are reasonnable, it would be great. I indeed overlooked the fact that there were two functions. The patch fixes the code so that both variants agree. I added comments to interval extractions, because it relies on the C division to have a negative remainder: -7/10 = 0 and remains -7. As for *_trunc functions, I have chosen to put the first year of the century or millennium: -100, 1, 101... 1001 2001 etc. Indeed, I don't think it would make sense to put 2000 (last year of the 2nd millennium) for rounding all years of the third millenium. I also fixed the code so that all decades last 10 years and decade 199 means the 1990's. I have added some tests that are relevant to deal with tricky cases. The formula may be simplified, but all these cases must pass. Please keep them. Fabien Coelho
2004-08-20 05:45:14 +02:00
/* caution: C division may have negative remainder */
Remove unnecessary parentheses in assignments. Add spaces where needed. Reference time interval variables as tinterval.
2005-07-21 06:41:43 +02:00
result = tm->tm_year / 10;
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
break;
case DTK_CENTURY:
> After all that about numbering centuries and millenia correctly, > why does CVS tip still give me > > regression=# select extract(century from now()); > date_part > ----------- > 20 > (1 row) > [ ... looks in code ... ] > > Apparently it's because you fixed only timestamp_part, and not > timestamptz_part. I'm not too sure about what timestamp_trunc or > timestamptz_trunc should do, but they may be wrong as well. Sigh... as usual, what is not tested does not work:-( > Could we have a more complete patch? Please find a submission attached. I hope it really fixes all decade, century and millenium issues for extract and *_trunc functions on interval and other timestamp types. If someone could check that the results are reasonnable, it would be great. I indeed overlooked the fact that there were two functions. The patch fixes the code so that both variants agree. I added comments to interval extractions, because it relies on the C division to have a negative remainder: -7/10 = 0 and remains -7. As for *_trunc functions, I have chosen to put the first year of the century or millennium: -100, 1, 101... 1001 2001 etc. Indeed, I don't think it would make sense to put 2000 (last year of the 2nd millennium) for rounding all years of the third millenium. I also fixed the code so that all decades last 10 years and decade 199 means the 1990's. I have added some tests that are relevant to deal with tricky cases. The formula may be simplified, but all these cases must pass. Please keep them. Fabien Coelho
2004-08-20 05:45:14 +02:00
/* caution: C division may have negative remainder */
Remove unnecessary parentheses in assignments. Add spaces where needed. Reference time interval variables as tinterval.
2005-07-21 06:41:43 +02:00
result = tm->tm_year / 100;
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +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:
> After all that about numbering centuries and millenia correctly, > why does CVS tip still give me > > regression=# select extract(century from now()); > date_part > ----------- > 20 > (1 row) > [ ... looks in code ... ] > > Apparently it's because you fixed only timestamp_part, and not > timestamptz_part. I'm not too sure about what timestamp_trunc or > timestamptz_trunc should do, but they may be wrong as well. Sigh... as usual, what is not tested does not work:-( > Could we have a more complete patch? Please find a submission attached. I hope it really fixes all decade, century and millenium issues for extract and *_trunc functions on interval and other timestamp types. If someone could check that the results are reasonnable, it would be great. I indeed overlooked the fact that there were two functions. The patch fixes the code so that both variants agree. I added comments to interval extractions, because it relies on the C division to have a negative remainder: -7/10 = 0 and remains -7. As for *_trunc functions, I have chosen to put the first year of the century or millennium: -100, 1, 101... 1001 2001 etc. Indeed, I don't think it would make sense to put 2000 (last year of the 2nd millennium) for rounding all years of the third millenium. I also fixed the code so that all decades last 10 years and decade 199 means the 1990's. I have added some tests that are relevant to deal with tricky cases. The formula may be simplified, but all these cases must pass. Please keep them. Fabien Coelho
2004-08-20 05:45:14 +02:00
/* caution: C division may have negative remainder */
Remove unnecessary parentheses in assignments. Add spaces where needed. Reference time interval variables as tinterval.
2005-07-21 06:41:43 +02:00
result = tm->tm_year / 1000;
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
break;
default:
Error message editing in utils/adt. Again thanks to Joe Conway for doing the bulk of the heavy lifting ...
2003-07-27 06:53:12 +02:00
ereport(ERROR,
(errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
errmsg("interval units \"%s\" not supported",
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
lowunits)));
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
result = 0;
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
}
}
else
{
Error message editing in utils/adt. Again thanks to Joe Conway for doing the bulk of the heavy lifting ...
2003-07-27 06:53:12 +02:00
elog(ERROR, "could not convert interval to tm");
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
result = 0;
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
}
}
Remove unnecessary parentheses in datetime/timestamp code.
2005-05-23 19:13:14 +02:00
else if (type == RESERV && val == DTK_EPOCH)
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
{
Change 5e0 to 5.0, for consistency.
2005-07-12 17:17:44 +02:00
result = interval->time / 1000000.0;
Fix overflow in extract(epoch from interval) for intervals exceeding 68 years. Seems to have been introduced in 8.1 by careless SECS_PER_DAY search-and-replace.
2007-09-16 17:56:20 +02:00
result += ((double) DAYS_PER_YEAR * SECS_PER_DAY) * (interval->month / MONTHS_PER_YEAR);
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
result += ((double) DAYS_PER_MONTH * SECS_PER_DAY) * (interval->month % MONTHS_PER_YEAR);
Fix overflow in extract(epoch from interval) for intervals exceeding 68 years. Seems to have been introduced in 8.1 by careless SECS_PER_DAY search-and-replace.
2007-09-16 17:56:20 +02:00
result += ((double) SECS_PER_DAY) * interval->day;
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
}
else
{
Error message editing in utils/adt. Again thanks to Joe Conway for doing the bulk of the heavy lifting ...
2003-07-27 06:53:12 +02:00
ereport(ERROR,
(errcode(ERRCODE_INVALID_PARAMETER_VALUE),
errmsg("interval units \"%s\" not recognized",
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
lowunits)));
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
result = 0;
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
}
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
PG_RETURN_FLOAT8(result);
}
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
/* timestamp_zone()
* Encode timestamp type with specified time zone.
* This function is just timestamp2timestamptz() except instead of
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
* shifting to the global timezone, we shift to the specified timezone.
* This is different from the other AT TIME ZONE cases because instead
Copyedit comments and documentation.
2016-04-02 03:53:10 +02:00
* of shifting _to_ a new time zone, it sets the time to _be_ the
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
* specified timezone.
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
*/
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
Datum
timestamp_zone(PG_FUNCTION_ARGS)
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
{
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
text *zone = PG_GETARG_TEXT_PP(0);
Fix the various forms of AT TIME ZONE to accept either timezones found in the zic database or zone names found in the date token table. This preserves the old ability to do AT TIME ZONE 'PST' along with the new ability to do AT TIME ZONE 'PST8PDT'. Per gripe from Bricklen Anderson. Also, fix some inconsistencies in usage of TZ_STRLEN_MAX --- the old code had the potential for one-byte buffer overruns, though given alignment considerations it's unlikely there was any real risk.
2005-09-09 04:31:50 +02:00
Timestamp timestamp = PG_GETARG_TIMESTAMP(1);
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
TimestampTz result;
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
int tz;
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
char tzname[TZ_STRLEN_MAX + 1];
Fix AT TIME ZONE (in all three variants) so that we first try to interpret the timezone argument as a timezone abbreviation, and only try it as a full timezone name if that fails. The zic database has four zones (CET, EET, MET, WET) that are full daylight-savings zones and yet have names that are the same as their abbreviations for standard time, resulting in ambiguity. In the timestamp input functions we resolve the ambiguity by preferring the abbreviation, and AT TIME ZONE should work the same way. (No functionality is lost because the zic database also has other names for these zones, eg Europe/Zurich.) Per gripe from Jaromir Talir. Backpatch to 8.1. Older releases did not have the issue because AT TIME ZONE only accepted abbreviations not zone names. (Thus, this patch also arguably fixes a compatibility botch introduced at 8.1: in ambiguous cases we now behave the same as 8.0 did.)
2008-07-07 20:09:46 +02:00
char *lowzone;
int type,
val;
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
struct pg_tm tm;
fsec_t fsec;
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
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
if (TIMESTAMP_NOT_FINITE(timestamp))
PG_RETURN_TIMESTAMPTZ(timestamp);
Fix the various forms of AT TIME ZONE to accept either timezones found in the zic database or zone names found in the date token table. This preserves the old ability to do AT TIME ZONE 'PST' along with the new ability to do AT TIME ZONE 'PST8PDT'. Per gripe from Bricklen Anderson. Also, fix some inconsistencies in usage of TZ_STRLEN_MAX --- the old code had the potential for one-byte buffer overruns, though given alignment considerations it's unlikely there was any real risk.
2005-09-09 04:31:50 +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
* Look up the requested timezone. First we look in the timezone
* abbreviation table (to handle cases like "EST"), and if that fails, we
* look in the timezone database (to handle cases like
* "America/New_York"). (This matches the order in which timestamp input
* checks the cases; it's important because the timezone database unwisely
* uses a few zone names that are identical to offset abbreviations.)
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
*/
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
text_to_cstring_buffer(zone, tzname, sizeof(tzname));
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 requires lowercase input */
Fix AT TIME ZONE (in all three variants) so that we first try to interpret the timezone argument as a timezone abbreviation, and only try it as a full timezone name if that fails. The zic database has four zones (CET, EET, MET, WET) that are full daylight-savings zones and yet have names that are the same as their abbreviations for standard time, resulting in ambiguity. In the timestamp input functions we resolve the ambiguity by preferring the abbreviation, and AT TIME ZONE should work the same way. (No functionality is lost because the zic database also has other names for these zones, eg Europe/Zurich.) Per gripe from Jaromir Talir. Backpatch to 8.1. Older releases did not have the issue because AT TIME ZONE only accepted abbreviations not zone names. (Thus, this patch also arguably fixes a compatibility botch introduced at 8.1: in ambiguous cases we now behave the same as 8.0 did.)
2008-07-07 20:09:46 +02:00
lowzone = downcase_truncate_identifier(tzname,
strlen(tzname),
false);
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
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
type = DecodeTimezoneAbbrev(0, lowzone, &val, &tzp);
Fix AT TIME ZONE (in all three variants) so that we first try to interpret the timezone argument as a timezone abbreviation, and only try it as a full timezone name if that fails. The zic database has four zones (CET, EET, MET, WET) that are full daylight-savings zones and yet have names that are the same as their abbreviations for standard time, resulting in ambiguity. In the timestamp input functions we resolve the ambiguity by preferring the abbreviation, and AT TIME ZONE should work the same way. (No functionality is lost because the zic database also has other names for these zones, eg Europe/Zurich.) Per gripe from Jaromir Talir. Backpatch to 8.1. Older releases did not have the issue because AT TIME ZONE only accepted abbreviations not zone names. (Thus, this patch also arguably fixes a compatibility botch introduced at 8.1: in ambiguous cases we now behave the same as 8.0 did.)
2008-07-07 20:09:46 +02:00
if (type == TZ || type == DTZ)
{
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
/* fixed-offset abbreviation */
tz = val;
result = dt2local(timestamp, tz);
}
else if (type == DYNTZ)
{
/* dynamic-offset abbreviation, resolve using specified time */
if (timestamp2tm(timestamp, NULL, &tm, &fsec, NULL, tzp) != 0)
ereport(ERROR,
(errcode(ERRCODE_DATETIME_VALUE_OUT_OF_RANGE),
errmsg("timestamp out of range")));
tz = -DetermineTimeZoneAbbrevOffset(&tm, tzname, tzp);
Fix AT TIME ZONE (in all three variants) so that we first try to interpret the timezone argument as a timezone abbreviation, and only try it as a full timezone name if that fails. The zic database has four zones (CET, EET, MET, WET) that are full daylight-savings zones and yet have names that are the same as their abbreviations for standard time, resulting in ambiguity. In the timestamp input functions we resolve the ambiguity by preferring the abbreviation, and AT TIME ZONE should work the same way. (No functionality is lost because the zic database also has other names for these zones, eg Europe/Zurich.) Per gripe from Jaromir Talir. Backpatch to 8.1. Older releases did not have the issue because AT TIME ZONE only accepted abbreviations not zone names. (Thus, this patch also arguably fixes a compatibility botch introduced at 8.1: in ambiguous cases we now behave the same as 8.0 did.)
2008-07-07 20:09:46 +02:00
result = dt2local(timestamp, tz);
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
}
Fix the various forms of AT TIME ZONE to accept either timezones found in the zic database or zone names found in the date token table. This preserves the old ability to do AT TIME ZONE 'PST' along with the new ability to do AT TIME ZONE 'PST8PDT'. Per gripe from Bricklen Anderson. Also, fix some inconsistencies in usage of TZ_STRLEN_MAX --- the old code had the potential for one-byte buffer overruns, though given alignment considerations it's unlikely there was any real risk.
2005-09-09 04:31:50 +02:00
else
More spacing adjustments
2005-07-22 17:15:38 +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
/* try it as a full zone name */
Fix AT TIME ZONE (in all three variants) so that we first try to interpret the timezone argument as a timezone abbreviation, and only try it as a full timezone name if that fails. The zic database has four zones (CET, EET, MET, WET) that are full daylight-savings zones and yet have names that are the same as their abbreviations for standard time, resulting in ambiguity. In the timestamp input functions we resolve the ambiguity by preferring the abbreviation, and AT TIME ZONE should work the same way. (No functionality is lost because the zic database also has other names for these zones, eg Europe/Zurich.) Per gripe from Jaromir Talir. Backpatch to 8.1. Older releases did not have the issue because AT TIME ZONE only accepted abbreviations not zone names. (Thus, this patch also arguably fixes a compatibility botch introduced at 8.1: in ambiguous cases we now behave the same as 8.0 did.)
2008-07-07 20:09:46 +02:00
tzp = pg_tzset(tzname);
if (tzp)
{
/* Apply the timezone change */
if (timestamp2tm(timestamp, NULL, &tm, &fsec, NULL, tzp) != 0)
ereport(ERROR,
(errcode(ERRCODE_DATETIME_VALUE_OUT_OF_RANGE),
errmsg("timestamp out of range")));
tz = DetermineTimeZoneOffset(&tm, tzp);
if (tm2timestamp(&tm, fsec, &tz, &result) != 0)
ereport(ERROR,
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
(errcode(ERRCODE_DATETIME_VALUE_OUT_OF_RANGE),
errmsg("timestamp out of range")));
Fix AT TIME ZONE (in all three variants) so that we first try to interpret the timezone argument as a timezone abbreviation, and only try it as a full timezone name if that fails. The zic database has four zones (CET, EET, MET, WET) that are full daylight-savings zones and yet have names that are the same as their abbreviations for standard time, resulting in ambiguity. In the timestamp input functions we resolve the ambiguity by preferring the abbreviation, and AT TIME ZONE should work the same way. (No functionality is lost because the zic database also has other names for these zones, eg Europe/Zurich.) Per gripe from Jaromir Talir. Backpatch to 8.1. Older releases did not have the issue because AT TIME ZONE only accepted abbreviations not zone names. (Thus, this patch also arguably fixes a compatibility botch introduced at 8.1: in ambiguous cases we now behave the same as 8.0 did.)
2008-07-07 20:09:46 +02:00
}
Fix the various forms of AT TIME ZONE to accept either timezones found in the zic database or zone names found in the date token table. This preserves the old ability to do AT TIME ZONE 'PST' along with the new ability to do AT TIME ZONE 'PST8PDT'. Per gripe from Bricklen Anderson. Also, fix some inconsistencies in usage of TZ_STRLEN_MAX --- the old code had the potential for one-byte buffer overruns, though given alignment considerations it's unlikely there was any real risk.
2005-09-09 04:31:50 +02:00
else
{
ereport(ERROR,
(errcode(ERRCODE_INVALID_PARAMETER_VALUE),
errmsg("time zone \"%s\" not recognized", tzname)));
8.4 pgindent run, with new combined Linux/FreeBSD/MinGW typedef list provided by Andrew.
2009-06-11 16:49:15 +02:00
result = 0; /* keep compiler quiet */
Fix the various forms of AT TIME ZONE to accept either timezones found in the zic database or zone names found in the date token table. This preserves the old ability to do AT TIME ZONE 'PST' along with the new ability to do AT TIME ZONE 'PST8PDT'. Per gripe from Bricklen Anderson. Also, fix some inconsistencies in usage of TZ_STRLEN_MAX --- the old code had the potential for one-byte buffer overruns, though given alignment considerations it's unlikely there was any real risk.
2005-09-09 04:31:50 +02:00
}
This patch makes it possible to use the full set of timezones when doing "AT TIME ZONE", and not just the shorlist previously available. For example: SELECT CURRENT_TIMESTAMP AT TIME ZONE 'Europe/London'; works fine now. It will also obey whatever DST rules were in effect at just that date, which the previous implementation did not. It also supports the AT TIME ZONE on the timetz datatype. The whole handling of DST is a bit bogus there, so I chose to make it use whatever DST rules are in effect at the time of executig the query. not sure if anybody is actuallyi *using* timetz though, it seems pretty unpredictable just because of this... Magnus Hagander
2005-06-15 02:34:11 +02:00
}
Fix AT TIME ZONE for timestamps without time zones: test=> select (CURRENT_DATE + '05:00'::time)::timestamp at time zone 'Canada/Pacific'; timezone ------------------------ 2005-07-22 08:00:00-04 (1 row)
2005-07-23 04:02:27 +02:00
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
if (!IS_VALID_TIMESTAMP(result))
ereport(ERROR,
(errcode(ERRCODE_DATETIME_VALUE_OUT_OF_RANGE),
errmsg("timestamp out of range")));
Fix AT TIME ZONE for timestamps without time zones: test=> select ('2005-07-20 00:00:00'::timestamp without time zone) at time zone 'Europe/Paris'; timezone ------------------------ 2005-07-19 22:00:00-04 Udpate documentation.
2005-07-22 23:16:15 +02:00
PG_RETURN_TIMESTAMPTZ(result);
More spacing adjustments
2005-07-22 17:15:38 +02:00
}
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
/* timestamp_izone()
* Encode timestamp type with specified time interval as time zone.
*/
Datum
timestamp_izone(PG_FUNCTION_ARGS)
{
Interval *zone = PG_GETARG_INTERVAL_P(0);
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
Timestamp timestamp = PG_GETARG_TIMESTAMP(1);
pgindent run on all C files. Java run to follow. initdb/regression tests pass.
2001-10-25 07:50:21 +02:00
TimestampTz result;
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
int tz;
if (TIMESTAMP_NOT_FINITE(timestamp))
PG_RETURN_TIMESTAMPTZ(timestamp);
Reject nonzero day fields in AT TIME ZONE INTERVAL functions. It's not sensible for an interval that's used as a time zone value to be larger than a day. When we changed the interval type to contain a separate day field, check_timezone() was adjusted to reject nonzero day values, but timetz_izone(), timestamp_izone(), and timestamptz_izone() evidently were overlooked. While at it, make the error messages for these three cases consistent.
2013-01-31 18:12:23 +01:00
if (zone->month != 0 || zone->day != 0)
Error message editing in utils/adt. Again thanks to Joe Conway for doing the bulk of the heavy lifting ...
2003-07-27 06:53:12 +02:00
ereport(ERROR,
(errcode(ERRCODE_INVALID_PARAMETER_VALUE),
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
errmsg("interval time zone \"%s\" must not include months or days",
DatumGetCString(DirectFunctionCall1(interval_out,
PointerGetDatum(zone))))));
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
Add datetime macros for constants, for clarity: #define SECS_PER_DAY 86400 #define USECS_PER_DAY INT64CONST(86400000000) #define USECS_PER_HOUR INT64CONST(3600000000) #define USECS_PER_MINUTE INT64CONST(60000000) #define USECS_PER_SEC INT64CONST(1000000)
2005-05-23 20:56:55 +02:00
tz = zone->time / USECS_PER_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
result = dt2local(timestamp, tz);
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
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
if (!IS_VALID_TIMESTAMP(result))
ereport(ERROR,
(errcode(ERRCODE_DATETIME_VALUE_OUT_OF_RANGE),
errmsg("timestamp out of range")));
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
PG_RETURN_TIMESTAMPTZ(result);
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
} /* timestamp_izone() */
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
Avoid some table rewrites for ALTER TABLE .. SET DATA TYPE timestamp. When the timezone is UTC, timestamptz and timestamp are binary coercible in both directions. See b8a18ad4850ea5ad7884aa6ab731fd392e73b4ad and c22ecc6562aac895f0f0529707d7bdb460fd2a49 for the previous attempt in this problem space. Skip the table rewrite; for now, continue to needlessly rewrite any index on an affected column. Reviewed by Simon Riggs and Tom Lane. Discussion: https://postgr.es/m/20190226061450.GA1665944@rfd.leadboat.com
2019-03-09 05:16:27 +01:00
/* TimestampTimestampTzRequiresRewrite()
*
* Returns false if the TimeZone GUC setting causes timestamp_timestamptz and
* timestamptz_timestamp to be no-ops, where the return value has the same
* bits as the argument. Since project convention is to assume a GUC changes
* no more often than STABLE functions change, the answer is valid that long.
*/
bool
TimestampTimestampTzRequiresRewrite(void)
{
long offset;
if (pg_get_timezone_offset(session_timezone, &offset) && offset == 0)
Don't call PG_RETURN_BOOL() in a function not returning Datum. This code is new in v12, and the defect probably was not user-visible.
2019-06-23 21:02:19 +02:00
return false;
return true;
Avoid some table rewrites for ALTER TABLE .. SET DATA TYPE timestamp. When the timezone is UTC, timestamptz and timestamp are binary coercible in both directions. See b8a18ad4850ea5ad7884aa6ab731fd392e73b4ad and c22ecc6562aac895f0f0529707d7bdb460fd2a49 for the previous attempt in this problem space. Skip the table rewrite; for now, continue to needlessly rewrite any index on an affected column. Reviewed by Simon Riggs and Tom Lane. Discussion: https://postgr.es/m/20190226061450.GA1665944@rfd.leadboat.com
2019-03-09 05:16:27 +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
/* timestamp_timestamptz()
* Convert local timestamp to timestamp at GMT
*/
Datum
timestamp_timestamptz(PG_FUNCTION_ARGS)
{
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
Timestamp timestamp = PG_GETARG_TIMESTAMP(0);
Add timestamp-versus-timestamptz cross-type comparison functions, flesh out the index operator classes to include these. In passing, fix erroneous volatility marking of ACL functions.
2004-03-22 02:38:18 +01:00
PG_RETURN_TIMESTAMPTZ(timestamp2timestamptz(timestamp));
}
Error suppression support for upcoming jsonpath .datetime() method Add support of error suppression in some date and time manipulation functions as it's required for jsonpath .datetime() method support. This commit doesn't use PG_TRY()/PG_CATCH() in order to implement that. Instead, it provides internal versions of date and time functions used, which support error suppression. Discussion: https://postgr.es/m/CAPpHfdsZgYEra_PeCLGNoXOWYx6iU-S3wF8aX0ObQUcZU%2B4XTw%40mail.gmail.com Author: Alexander Korotkov, Nikita Glukhov Reviewed-by: Anastasia Lubennikova, Peter Eisentraut
2019-09-25 20:51:47 +02:00
/*
* Convert timestamp to timestamp with time zone.
*
Refactor jsonpath's compareDatetime() This commit refactors come ridiculous coding in compareDatetime(). Also, it provides correct cross-datatype comparison even when one of values overflows during cast. That eliminates dilemma on whether we should suppress overflow errors during cast. Reported-by: Tom Lane Discussion: https://postgr.es/m/32308.1569455803%40sss.pgh.pa.us Discussion: https://postgr.es/m/a5629d0c-8162-7559-16aa-0c8390d6ba5f%40postgrespro.ru Author: Nikita Glukhov, Alexander Korotkov
2019-10-21 22:04:14 +02:00
* On overflow error is thrown if 'overflow' is NULL. Otherwise, '*overflow'
* is set to -1 (+1) when result value exceed lower (upper) boundary and zero
* returned.
Error suppression support for upcoming jsonpath .datetime() method Add support of error suppression in some date and time manipulation functions as it's required for jsonpath .datetime() method support. This commit doesn't use PG_TRY()/PG_CATCH() in order to implement that. Instead, it provides internal versions of date and time functions used, which support error suppression. Discussion: https://postgr.es/m/CAPpHfdsZgYEra_PeCLGNoXOWYx6iU-S3wF8aX0ObQUcZU%2B4XTw%40mail.gmail.com Author: Alexander Korotkov, Nikita Glukhov Reviewed-by: Anastasia Lubennikova, Peter Eisentraut
2019-09-25 20:51:47 +02:00
*/
TimestampTz
Refactor jsonpath's compareDatetime() This commit refactors come ridiculous coding in compareDatetime(). Also, it provides correct cross-datatype comparison even when one of values overflows during cast. That eliminates dilemma on whether we should suppress overflow errors during cast. Reported-by: Tom Lane Discussion: https://postgr.es/m/32308.1569455803%40sss.pgh.pa.us Discussion: https://postgr.es/m/a5629d0c-8162-7559-16aa-0c8390d6ba5f%40postgrespro.ru Author: Nikita Glukhov, Alexander Korotkov
2019-10-21 22:04:14 +02:00
timestamp2timestamptz_opt_overflow(Timestamp timestamp, int *overflow)
Add timestamp-versus-timestamptz cross-type comparison functions, flesh out the index operator classes to include these. In passing, fix erroneous volatility marking of ACL functions.
2004-03-22 02:38:18 +01:00
{
pgindent run on all C files. Java run to follow. initdb/regression tests pass.
2001-10-25 07:50:21 +02:00
TimestampTz result;
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
struct pg_tm tt,
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 = &tt;
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;
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
int tz;
if (TIMESTAMP_NOT_FINITE(timestamp))
Error suppression support for upcoming jsonpath .datetime() method Add support of error suppression in some date and time manipulation functions as it's required for jsonpath .datetime() method support. This commit doesn't use PG_TRY()/PG_CATCH() in order to implement that. Instead, it provides internal versions of date and time functions used, which support error suppression. Discussion: https://postgr.es/m/CAPpHfdsZgYEra_PeCLGNoXOWYx6iU-S3wF8aX0ObQUcZU%2B4XTw%40mail.gmail.com Author: Alexander Korotkov, Nikita Glukhov Reviewed-by: Anastasia Lubennikova, Peter Eisentraut
2019-09-25 20:51:47 +02:00
return timestamp;
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
Error suppression support for upcoming jsonpath .datetime() method Add support of error suppression in some date and time manipulation functions as it's required for jsonpath .datetime() method support. This commit doesn't use PG_TRY()/PG_CATCH() in order to implement that. Instead, it provides internal versions of date and time functions used, which support error suppression. Discussion: https://postgr.es/m/CAPpHfdsZgYEra_PeCLGNoXOWYx6iU-S3wF8aX0ObQUcZU%2B4XTw%40mail.gmail.com Author: Alexander Korotkov, Nikita Glukhov Reviewed-by: Anastasia Lubennikova, Peter Eisentraut
2019-09-25 20:51:47 +02:00
if (!timestamp2tm(timestamp, NULL, tm, &fsec, NULL, NULL))
{
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
tz = DetermineTimeZoneOffset(tm, session_timezone);
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
Refactor timestamp2timestamptz_opt_error() While casting from timestamp to timestamptz we do timestamp2tm() then tm2timestamp(). This commit eliminates call to tm2timestamp(). Instead, it directly applies timezone offset to the original timestamp value. That makes upcoming datetime overflow handling in jsonpath easier. That should also save us some CPU cycles. Discussion: https://postgr.es/m/CAPpHfdvRPRh_mTGar5WmDeRZ%3DU5dOXHdxspYYD%3D76m3knNGjXA%40mail.gmail.com Author: Alexander Korotkov Reviewed-by: Tom Lane
2019-10-21 22:03:55 +02:00
result = dt2local(timestamp, -tz);
if (IS_VALID_TIMESTAMP(result))
{
Error suppression support for upcoming jsonpath .datetime() method Add support of error suppression in some date and time manipulation functions as it's required for jsonpath .datetime() method support. This commit doesn't use PG_TRY()/PG_CATCH() in order to implement that. Instead, it provides internal versions of date and time functions used, which support error suppression. Discussion: https://postgr.es/m/CAPpHfdsZgYEra_PeCLGNoXOWYx6iU-S3wF8aX0ObQUcZU%2B4XTw%40mail.gmail.com Author: Alexander Korotkov, Nikita Glukhov Reviewed-by: Anastasia Lubennikova, Peter Eisentraut
2019-09-25 20:51:47 +02:00
return result;
Refactor timestamp2timestamptz_opt_error() While casting from timestamp to timestamptz we do timestamp2tm() then tm2timestamp(). This commit eliminates call to tm2timestamp(). Instead, it directly applies timezone offset to the original timestamp value. That makes upcoming datetime overflow handling in jsonpath easier. That should also save us some CPU cycles. Discussion: https://postgr.es/m/CAPpHfdvRPRh_mTGar5WmDeRZ%3DU5dOXHdxspYYD%3D76m3knNGjXA%40mail.gmail.com Author: Alexander Korotkov Reviewed-by: Tom Lane
2019-10-21 22:03:55 +02:00
}
Refactor jsonpath's compareDatetime() This commit refactors come ridiculous coding in compareDatetime(). Also, it provides correct cross-datatype comparison even when one of values overflows during cast. That eliminates dilemma on whether we should suppress overflow errors during cast. Reported-by: Tom Lane Discussion: https://postgr.es/m/32308.1569455803%40sss.pgh.pa.us Discussion: https://postgr.es/m/a5629d0c-8162-7559-16aa-0c8390d6ba5f%40postgrespro.ru Author: Nikita Glukhov, Alexander Korotkov
2019-10-21 22:04:14 +02:00
else if (overflow)
Refactor timestamp2timestamptz_opt_error() While casting from timestamp to timestamptz we do timestamp2tm() then tm2timestamp(). This commit eliminates call to tm2timestamp(). Instead, it directly applies timezone offset to the original timestamp value. That makes upcoming datetime overflow handling in jsonpath easier. That should also save us some CPU cycles. Discussion: https://postgr.es/m/CAPpHfdvRPRh_mTGar5WmDeRZ%3DU5dOXHdxspYYD%3D76m3knNGjXA%40mail.gmail.com Author: Alexander Korotkov Reviewed-by: Tom Lane
2019-10-21 22:03:55 +02:00
{
Refactor jsonpath's compareDatetime() This commit refactors come ridiculous coding in compareDatetime(). Also, it provides correct cross-datatype comparison even when one of values overflows during cast. That eliminates dilemma on whether we should suppress overflow errors during cast. Reported-by: Tom Lane Discussion: https://postgr.es/m/32308.1569455803%40sss.pgh.pa.us Discussion: https://postgr.es/m/a5629d0c-8162-7559-16aa-0c8390d6ba5f%40postgrespro.ru Author: Nikita Glukhov, Alexander Korotkov
2019-10-21 22:04:14 +02:00
if (result < MIN_TIMESTAMP)
*overflow = -1;
else
{
Assert(result >= END_TIMESTAMP);
*overflow = 1;
}
Refactor timestamp2timestamptz_opt_error() While casting from timestamp to timestamptz we do timestamp2tm() then tm2timestamp(). This commit eliminates call to tm2timestamp(). Instead, it directly applies timezone offset to the original timestamp value. That makes upcoming datetime overflow handling in jsonpath easier. That should also save us some CPU cycles. Discussion: https://postgr.es/m/CAPpHfdvRPRh_mTGar5WmDeRZ%3DU5dOXHdxspYYD%3D76m3knNGjXA%40mail.gmail.com Author: Alexander Korotkov Reviewed-by: Tom Lane
2019-10-21 22:03:55 +02:00
return (TimestampTz) 0;
}
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
}
Refactor jsonpath's compareDatetime() This commit refactors come ridiculous coding in compareDatetime(). Also, it provides correct cross-datatype comparison even when one of values overflows during cast. That eliminates dilemma on whether we should suppress overflow errors during cast. Reported-by: Tom Lane Discussion: https://postgr.es/m/32308.1569455803%40sss.pgh.pa.us Discussion: https://postgr.es/m/a5629d0c-8162-7559-16aa-0c8390d6ba5f%40postgrespro.ru Author: Nikita Glukhov, Alexander Korotkov
2019-10-21 22:04:14 +02:00
ereport(ERROR,
(errcode(ERRCODE_DATETIME_VALUE_OUT_OF_RANGE),
errmsg("timestamp out of range")));
Error suppression support for upcoming jsonpath .datetime() method Add support of error suppression in some date and time manipulation functions as it's required for jsonpath .datetime() method support. This commit doesn't use PG_TRY()/PG_CATCH() in order to implement that. Instead, it provides internal versions of date and time functions used, which support error suppression. Discussion: https://postgr.es/m/CAPpHfdsZgYEra_PeCLGNoXOWYx6iU-S3wF8aX0ObQUcZU%2B4XTw%40mail.gmail.com Author: Alexander Korotkov, Nikita Glukhov Reviewed-by: Anastasia Lubennikova, Peter Eisentraut
2019-09-25 20:51:47 +02:00
return 0;
}
/*
Refactor jsonpath's compareDatetime() This commit refactors come ridiculous coding in compareDatetime(). Also, it provides correct cross-datatype comparison even when one of values overflows during cast. That eliminates dilemma on whether we should suppress overflow errors during cast. Reported-by: Tom Lane Discussion: https://postgr.es/m/32308.1569455803%40sss.pgh.pa.us Discussion: https://postgr.es/m/a5629d0c-8162-7559-16aa-0c8390d6ba5f%40postgrespro.ru Author: Nikita Glukhov, Alexander Korotkov
2019-10-21 22:04:14 +02:00
* Single-argument version of timestamp2timestamptz_opt_overflow().
Error suppression support for upcoming jsonpath .datetime() method Add support of error suppression in some date and time manipulation functions as it's required for jsonpath .datetime() method support. This commit doesn't use PG_TRY()/PG_CATCH() in order to implement that. Instead, it provides internal versions of date and time functions used, which support error suppression. Discussion: https://postgr.es/m/CAPpHfdsZgYEra_PeCLGNoXOWYx6iU-S3wF8aX0ObQUcZU%2B4XTw%40mail.gmail.com Author: Alexander Korotkov, Nikita Glukhov Reviewed-by: Anastasia Lubennikova, Peter Eisentraut
2019-09-25 20:51:47 +02:00
*/
static TimestampTz
timestamp2timestamptz(Timestamp timestamp)
{
Refactor jsonpath's compareDatetime() This commit refactors come ridiculous coding in compareDatetime(). Also, it provides correct cross-datatype comparison even when one of values overflows during cast. That eliminates dilemma on whether we should suppress overflow errors during cast. Reported-by: Tom Lane Discussion: https://postgr.es/m/32308.1569455803%40sss.pgh.pa.us Discussion: https://postgr.es/m/a5629d0c-8162-7559-16aa-0c8390d6ba5f%40postgrespro.ru Author: Nikita Glukhov, Alexander Korotkov
2019-10-21 22:04:14 +02:00
return timestamp2timestamptz_opt_overflow(timestamp, 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
}
/* timestamptz_timestamp()
* Convert timestamp at GMT to local timestamp
*/
Datum
timestamptz_timestamp(PG_FUNCTION_ARGS)
{
Add binary I/O routines for a bunch more datatypes. Still a few to go, but that was enough tedium for one day. Along the way, move the few support routines for types xid and cid into a more logical place.
2003-05-13 01:08:52 +02:00
TimestampTz timestamp = PG_GETARG_TIMESTAMPTZ(0);
Improve parsetree representation of special functions such as CURRENT_DATE. We implement a dozen or so parameterless functions that the SQL standard defines special syntax for. Up to now, that was done by converting them into more or less ad-hoc constructs such as "'now'::text::date". That's messy for multiple reasons: it exposes what should be implementation details to users, and performance is worse than it needs to be in several cases. To improve matters, invent a new expression node type SQLValueFunction that can represent any of these parameterless functions. Bump catversion because this changes stored parsetrees for rules. Discussion: <30058.1463091294@sss.pgh.pa.us>
2016-08-17 02:33:01 +02:00
PG_RETURN_TIMESTAMP(timestamptz2timestamp(timestamp));
}
static Timestamp
timestamptz2timestamp(TimestampTz timestamp)
{
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
Timestamp result;
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
struct pg_tm tt,
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 = &tt;
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;
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
int tz;
if (TIMESTAMP_NOT_FINITE(timestamp))
result = timestamp;
else
{
Remove unused tzn arguments for timestamp2tm()
2012-03-15 20:13:35 +01:00
if (timestamp2tm(timestamp, &tz, tm, &fsec, NULL, NULL) != 0)
Error message editing in utils/adt. Again thanks to Joe Conway for doing the bulk of the heavy lifting ...
2003-07-27 06:53:12 +02:00
ereport(ERROR,
(errcode(ERRCODE_DATETIME_VALUE_OUT_OF_RANGE),
errmsg("timestamp out of range")));
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 (tm2timestamp(tm, fsec, NULL, &result) != 0)
Error message editing in utils/adt. Again thanks to Joe Conway for doing the bulk of the heavy lifting ...
2003-07-27 06:53:12 +02:00
ereport(ERROR,
(errcode(ERRCODE_DATETIME_VALUE_OUT_OF_RANGE),
errmsg("timestamp out of range")));
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
}
Improve parsetree representation of special functions such as CURRENT_DATE. We implement a dozen or so parameterless functions that the SQL standard defines special syntax for. Up to now, that was done by converting them into more or less ad-hoc constructs such as "'now'::text::date". That's messy for multiple reasons: it exposes what should be implementation details to users, and performance is worse than it needs to be in several cases. To improve matters, invent a new expression node type SQLValueFunction that can represent any of these parameterless functions. Bump catversion because this changes stored parsetrees for rules. Discussion: <30058.1463091294@sss.pgh.pa.us>
2016-08-17 02:33:01 +02:00
return result;
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
}
/* timestamptz_zone()
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
* Evaluate timestamp with time zone type at the specified time zone.
* Returns a timestamp without time zone.
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
*/
Datum
timestamptz_zone(PG_FUNCTION_ARGS)
{
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
text *zone = PG_GETARG_TEXT_PP(0);
Add binary I/O routines for a bunch more datatypes. Still a few to go, but that was enough tedium for one day. Along the way, move the few support routines for types xid and cid into a more logical place.
2003-05-13 01:08:52 +02:00
TimestampTz timestamp = PG_GETARG_TIMESTAMPTZ(1);
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
Timestamp result;
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
int tz;
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
char tzname[TZ_STRLEN_MAX + 1];
Fix AT TIME ZONE (in all three variants) so that we first try to interpret the timezone argument as a timezone abbreviation, and only try it as a full timezone name if that fails. The zic database has four zones (CET, EET, MET, WET) that are full daylight-savings zones and yet have names that are the same as their abbreviations for standard time, resulting in ambiguity. In the timestamp input functions we resolve the ambiguity by preferring the abbreviation, and AT TIME ZONE should work the same way. (No functionality is lost because the zic database also has other names for these zones, eg Europe/Zurich.) Per gripe from Jaromir Talir. Backpatch to 8.1. Older releases did not have the issue because AT TIME ZONE only accepted abbreviations not zone names. (Thus, this patch also arguably fixes a compatibility botch introduced at 8.1: in ambiguous cases we now behave the same as 8.0 did.)
2008-07-07 20:09:46 +02:00
char *lowzone;
int type,
val;
pg_tz *tzp;
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
Another round of updates for new fmgr, mostly in the datetime code.
2000-06-09 03:11:16 +02:00
if (TIMESTAMP_NOT_FINITE(timestamp))
Fix the various forms of AT TIME ZONE to accept either timezones found in the zic database or zone names found in the date token table. This preserves the old ability to do AT TIME ZONE 'PST' along with the new ability to do AT TIME ZONE 'PST8PDT'. Per gripe from Bricklen Anderson. Also, fix some inconsistencies in usage of TZ_STRLEN_MAX --- the old code had the potential for one-byte buffer overruns, though given alignment considerations it's unlikely there was any real risk.
2005-09-09 04:31:50 +02:00
PG_RETURN_TIMESTAMP(timestamp);
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
Fix the various forms of AT TIME ZONE to accept either timezones found in the zic database or zone names found in the date token table. This preserves the old ability to do AT TIME ZONE 'PST' along with the new ability to do AT TIME ZONE 'PST8PDT'. Per gripe from Bricklen Anderson. Also, fix some inconsistencies in usage of TZ_STRLEN_MAX --- the old code had the potential for one-byte buffer overruns, though given alignment considerations it's unlikely there was any real risk.
2005-09-09 04:31:50 +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
* Look up the requested timezone. First we look in the timezone
* abbreviation table (to handle cases like "EST"), and if that fails, we
* look in the timezone database (to handle cases like
* "America/New_York"). (This matches the order in which timestamp input
* checks the cases; it's important because the timezone database unwisely
* uses a few zone names that are identical to offset abbreviations.)
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
*/
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
text_to_cstring_buffer(zone, tzname, sizeof(tzname));
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 requires lowercase input */
Fix AT TIME ZONE (in all three variants) so that we first try to interpret the timezone argument as a timezone abbreviation, and only try it as a full timezone name if that fails. The zic database has four zones (CET, EET, MET, WET) that are full daylight-savings zones and yet have names that are the same as their abbreviations for standard time, resulting in ambiguity. In the timestamp input functions we resolve the ambiguity by preferring the abbreviation, and AT TIME ZONE should work the same way. (No functionality is lost because the zic database also has other names for these zones, eg Europe/Zurich.) Per gripe from Jaromir Talir. Backpatch to 8.1. Older releases did not have the issue because AT TIME ZONE only accepted abbreviations not zone names. (Thus, this patch also arguably fixes a compatibility botch introduced at 8.1: in ambiguous cases we now behave the same as 8.0 did.)
2008-07-07 20:09:46 +02:00
lowzone = downcase_truncate_identifier(tzname,
strlen(tzname),
false);
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +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
type = DecodeTimezoneAbbrev(0, lowzone, &val, &tzp);
Fix AT TIME ZONE (in all three variants) so that we first try to interpret the timezone argument as a timezone abbreviation, and only try it as a full timezone name if that fails. The zic database has four zones (CET, EET, MET, WET) that are full daylight-savings zones and yet have names that are the same as their abbreviations for standard time, resulting in ambiguity. In the timestamp input functions we resolve the ambiguity by preferring the abbreviation, and AT TIME ZONE should work the same way. (No functionality is lost because the zic database also has other names for these zones, eg Europe/Zurich.) Per gripe from Jaromir Talir. Backpatch to 8.1. Older releases did not have the issue because AT TIME ZONE only accepted abbreviations not zone names. (Thus, this patch also arguably fixes a compatibility botch introduced at 8.1: in ambiguous cases we now behave the same as 8.0 did.)
2008-07-07 20:09:46 +02:00
if (type == TZ || type == DTZ)
{
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
/* fixed-offset abbreviation */
tz = -val;
result = dt2local(timestamp, tz);
}
else if (type == DYNTZ)
{
/* dynamic-offset abbreviation, resolve using specified time */
int isdst;
tz = DetermineTimeZoneAbbrevOffsetTS(timestamp, tzname, tzp, &isdst);
Fix AT TIME ZONE (in all three variants) so that we first try to interpret the timezone argument as a timezone abbreviation, and only try it as a full timezone name if that fails. The zic database has four zones (CET, EET, MET, WET) that are full daylight-savings zones and yet have names that are the same as their abbreviations for standard time, resulting in ambiguity. In the timestamp input functions we resolve the ambiguity by preferring the abbreviation, and AT TIME ZONE should work the same way. (No functionality is lost because the zic database also has other names for these zones, eg Europe/Zurich.) Per gripe from Jaromir Talir. Backpatch to 8.1. Older releases did not have the issue because AT TIME ZONE only accepted abbreviations not zone names. (Thus, this patch also arguably fixes a compatibility botch introduced at 8.1: in ambiguous cases we now behave the same as 8.0 did.)
2008-07-07 20:09:46 +02:00
result = dt2local(timestamp, tz);
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
}
Fix the various forms of AT TIME ZONE to accept either timezones found in the zic database or zone names found in the date token table. This preserves the old ability to do AT TIME ZONE 'PST' along with the new ability to do AT TIME ZONE 'PST8PDT'. Per gripe from Bricklen Anderson. Also, fix some inconsistencies in usage of TZ_STRLEN_MAX --- the old code had the potential for one-byte buffer overruns, though given alignment considerations it's unlikely there was any real risk.
2005-09-09 04:31:50 +02:00
else
{
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
/* try it as a full zone name */
Fix AT TIME ZONE (in all three variants) so that we first try to interpret the timezone argument as a timezone abbreviation, and only try it as a full timezone name if that fails. The zic database has four zones (CET, EET, MET, WET) that are full daylight-savings zones and yet have names that are the same as their abbreviations for standard time, resulting in ambiguity. In the timestamp input functions we resolve the ambiguity by preferring the abbreviation, and AT TIME ZONE should work the same way. (No functionality is lost because the zic database also has other names for these zones, eg Europe/Zurich.) Per gripe from Jaromir Talir. Backpatch to 8.1. Older releases did not have the issue because AT TIME ZONE only accepted abbreviations not zone names. (Thus, this patch also arguably fixes a compatibility botch introduced at 8.1: in ambiguous cases we now behave the same as 8.0 did.)
2008-07-07 20:09:46 +02:00
tzp = pg_tzset(tzname);
if (tzp)
{
/* Apply the timezone change */
struct pg_tm tm;
fsec_t fsec;
Fix the various forms of AT TIME ZONE to accept either timezones found in the zic database or zone names found in the date token table. This preserves the old ability to do AT TIME ZONE 'PST' along with the new ability to do AT TIME ZONE 'PST8PDT'. Per gripe from Bricklen Anderson. Also, fix some inconsistencies in usage of TZ_STRLEN_MAX --- the old code had the potential for one-byte buffer overruns, though given alignment considerations it's unlikely there was any real risk.
2005-09-09 04:31:50 +02:00
Fix AT TIME ZONE (in all three variants) so that we first try to interpret the timezone argument as a timezone abbreviation, and only try it as a full timezone name if that fails. The zic database has four zones (CET, EET, MET, WET) that are full daylight-savings zones and yet have names that are the same as their abbreviations for standard time, resulting in ambiguity. In the timestamp input functions we resolve the ambiguity by preferring the abbreviation, and AT TIME ZONE should work the same way. (No functionality is lost because the zic database also has other names for these zones, eg Europe/Zurich.) Per gripe from Jaromir Talir. Backpatch to 8.1. Older releases did not have the issue because AT TIME ZONE only accepted abbreviations not zone names. (Thus, this patch also arguably fixes a compatibility botch introduced at 8.1: in ambiguous cases we now behave the same as 8.0 did.)
2008-07-07 20:09:46 +02:00
if (timestamp2tm(timestamp, &tz, &tm, &fsec, NULL, tzp) != 0)
ereport(ERROR,
(errcode(ERRCODE_DATETIME_VALUE_OUT_OF_RANGE),
errmsg("timestamp out of range")));
if (tm2timestamp(&tm, fsec, NULL, &result) != 0)
ereport(ERROR,
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
(errcode(ERRCODE_DATETIME_VALUE_OUT_OF_RANGE),
errmsg("timestamp out of range")));
Fix AT TIME ZONE (in all three variants) so that we first try to interpret the timezone argument as a timezone abbreviation, and only try it as a full timezone name if that fails. The zic database has four zones (CET, EET, MET, WET) that are full daylight-savings zones and yet have names that are the same as their abbreviations for standard time, resulting in ambiguity. In the timestamp input functions we resolve the ambiguity by preferring the abbreviation, and AT TIME ZONE should work the same way. (No functionality is lost because the zic database also has other names for these zones, eg Europe/Zurich.) Per gripe from Jaromir Talir. Backpatch to 8.1. Older releases did not have the issue because AT TIME ZONE only accepted abbreviations not zone names. (Thus, this patch also arguably fixes a compatibility botch introduced at 8.1: in ambiguous cases we now behave the same as 8.0 did.)
2008-07-07 20:09:46 +02:00
}
Fix the various forms of AT TIME ZONE to accept either timezones found in the zic database or zone names found in the date token table. This preserves the old ability to do AT TIME ZONE 'PST' along with the new ability to do AT TIME ZONE 'PST8PDT'. Per gripe from Bricklen Anderson. Also, fix some inconsistencies in usage of TZ_STRLEN_MAX --- the old code had the potential for one-byte buffer overruns, though given alignment considerations it's unlikely there was any real risk.
2005-09-09 04:31:50 +02:00
else
{
ereport(ERROR,
(errcode(ERRCODE_INVALID_PARAMETER_VALUE),
errmsg("time zone \"%s\" not recognized", tzname)));
8.4 pgindent run, with new combined Linux/FreeBSD/MinGW typedef list provided by Andrew.
2009-06-11 16:49:15 +02:00
result = 0; /* keep compiler quiet */
Fix the various forms of AT TIME ZONE to accept either timezones found in the zic database or zone names found in the date token table. This preserves the old ability to do AT TIME ZONE 'PST' along with the new ability to do AT TIME ZONE 'PST8PDT'. Per gripe from Bricklen Anderson. Also, fix some inconsistencies in usage of TZ_STRLEN_MAX --- the old code had the potential for one-byte buffer overruns, though given alignment considerations it's unlikely there was any real risk.
2005-09-09 04:31:50 +02:00
}
Implement "date/time grand unification". Transform datetime and timespan into timestamp and interval. Deprecate datetime and timespan, though translate to new types in gram.y. Transform all datetime and timespan catalog entries into new types. Make "INTERVAL" reserved word allowed as a column identifier in gram.y. Remove dt.h, dt.c files, and retarget datetime.h, datetime.c as utility routines for all date/time types. date.{h,c} now deals with date, time types. timestamp.{h,c} now deals with timestamp, interval types. nabstime.{h,c} now deals with abstime, reltime, tinterval types. Make NUMERIC a known native type for purposes of type coersion. Not tested.
2000-02-16 18:26:26 +01:00
}
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
if (!IS_VALID_TIMESTAMP(result))
ereport(ERROR,
(errcode(ERRCODE_DATETIME_VALUE_OUT_OF_RANGE),
errmsg("timestamp out of range")));
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
PG_RETURN_TIMESTAMP(result);
More spacing adjustments
2005-07-22 17:15:38 +02:00
}
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
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
/* timestamptz_izone()
* Encode timestamp with time zone type with specified time interval as time zone.
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
* Returns a timestamp without time zone.
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
*/
Datum
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
timestamptz_izone(PG_FUNCTION_ARGS)
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
{
Interval *zone = PG_GETARG_INTERVAL_P(0);
Add binary I/O routines for a bunch more datatypes. Still a few to go, but that was enough tedium for one day. Along the way, move the few support routines for types xid and cid into a more logical place.
2003-05-13 01:08:52 +02:00
TimestampTz timestamp = PG_GETARG_TIMESTAMPTZ(1);
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
Timestamp result;
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 tz;
if (TIMESTAMP_NOT_FINITE(timestamp))
timestamptz_izone should return the input, not NULL, when the input is a non-finite timestamp, for consistency with related functions. In other words: +infinity rotated to a different timezone is still +infinity.
2005-09-09 08:46:14 +02:00
PG_RETURN_TIMESTAMP(timestamp);
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
Reject nonzero day fields in AT TIME ZONE INTERVAL functions. It's not sensible for an interval that's used as a time zone value to be larger than a day. When we changed the interval type to contain a separate day field, check_timezone() was adjusted to reject nonzero day values, but timetz_izone(), timestamp_izone(), and timestamptz_izone() evidently were overlooked. While at it, make the error messages for these three cases consistent.
2013-01-31 18:12:23 +01:00
if (zone->month != 0 || zone->day != 0)
Error message editing in utils/adt. Again thanks to Joe Conway for doing the bulk of the heavy lifting ...
2003-07-27 06:53:12 +02:00
ereport(ERROR,
(errcode(ERRCODE_INVALID_PARAMETER_VALUE),
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
errmsg("interval time zone \"%s\" must not include months or days",
DatumGetCString(DirectFunctionCall1(interval_out,
PointerGetDatum(zone))))));
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
Add datetime macros for constants, for clarity: #define SECS_PER_DAY 86400 #define USECS_PER_DAY INT64CONST(86400000000) #define USECS_PER_HOUR INT64CONST(3600000000) #define USECS_PER_MINUTE INT64CONST(60000000) #define USECS_PER_SEC INT64CONST(1000000)
2005-05-23 20:56:55 +02:00
tz = -(zone->time / USECS_PER_SEC);
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
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
result = dt2local(timestamp, tz);
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
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
if (!IS_VALID_TIMESTAMP(result))
ereport(ERROR,
(errcode(ERRCODE_DATETIME_VALUE_OUT_OF_RANGE),
errmsg("timestamp out of range")));
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
PG_RETURN_TIMESTAMP(result);
More spacing adjustments
2005-07-22 17:15:38 +02:00
}
Add timestamp and timestamptz versions of generate_series(). Hitoshi Harada
2008-05-05 01:19:24 +02:00
/* generate_series_timestamp()
* Generate the set of timestamps from start to finish by step
*/
Datum
generate_series_timestamp(PG_FUNCTION_ARGS)
{
FuncCallContext *funcctx;
generate_series_timestamp_fctx *fctx;
8.4 pgindent run, with new combined Linux/FreeBSD/MinGW typedef list provided by Andrew.
2009-06-11 16:49:15 +02:00
Timestamp result;
Add timestamp and timestamptz versions of generate_series(). Hitoshi Harada
2008-05-05 01:19:24 +02:00
/* stuff done only on the first call of the function */
if (SRF_IS_FIRSTCALL())
{
8.4 pgindent run, with new combined Linux/FreeBSD/MinGW typedef list provided by Andrew.
2009-06-11 16:49:15 +02:00
Timestamp start = PG_GETARG_TIMESTAMP(0);
Timestamp finish = PG_GETARG_TIMESTAMP(1);
Interval *step = PG_GETARG_INTERVAL_P(2);
Add timestamp and timestamptz versions of generate_series(). Hitoshi Harada
2008-05-05 01:19:24 +02:00
MemoryContext oldcontext;
8.4 pgindent run, with new combined Linux/FreeBSD/MinGW typedef list provided by Andrew.
2009-06-11 16:49:15 +02:00
Interval interval_zero;
Add timestamp and timestamptz versions of generate_series(). Hitoshi Harada
2008-05-05 01:19:24 +02:00
/* create a function context for cross-call persistence */
funcctx = SRF_FIRSTCALL_INIT();
/*
* switch to memory context appropriate for multiple function calls
*/
oldcontext = MemoryContextSwitchTo(funcctx->multi_call_memory_ctx);
/* allocate memory for user context */
fctx = (generate_series_timestamp_fctx *)
palloc(sizeof(generate_series_timestamp_fctx));
/*
* Use fctx to keep state from call to call. Seed current with the
* original start value
*/
fctx->current = start;
fctx->finish = finish;
fctx->step = *step;
/* Determine sign of the interval */
MemSet(&interval_zero, 0, sizeof(Interval));
fctx->step_sign = interval_cmp_internal(&fctx->step, &interval_zero);
if (fctx->step_sign == 0)
ereport(ERROR,
(errcode(ERRCODE_INVALID_PARAMETER_VALUE),
errmsg("step size cannot equal zero")));
funcctx->user_fctx = fctx;
MemoryContextSwitchTo(oldcontext);
}
/* stuff done on every call of the function */
funcctx = SRF_PERCALL_SETUP();
/*
* get the saved state and use current as the result for this iteration
*/
fctx = funcctx->user_fctx;
result = fctx->current;
if (fctx->step_sign > 0 ?
timestamp_cmp_internal(result, fctx->finish) <= 0 :
timestamp_cmp_internal(result, fctx->finish) >= 0)
{
/* increment current in preparation for next iteration */
Clean up newlines following left parentheses We used to strategically place newlines after some function call left parentheses to make pgindent move the argument list a few chars to the left, so that the whole line would fit under 80 chars. However, pgindent no longer does that, so the newlines just made the code vertically longer for no reason. Remove those newlines, and reflow some of those lines for some extra naturality. Reviewed-by: Michael Paquier, Tom Lane Discussion: https://postgr.es/m/20200129200401.GA6303@alvherre.pgsql
2020-01-30 17:42:14 +01:00
fctx->current = DatumGetTimestamp(DirectFunctionCall2(timestamp_pl_interval,
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
TimestampGetDatum(fctx->current),
PointerGetDatum(&fctx->step)));
Add timestamp and timestamptz versions of generate_series(). Hitoshi Harada
2008-05-05 01:19:24 +02:00
/* do when there is more left to send */
SRF_RETURN_NEXT(funcctx, TimestampGetDatum(result));
}
else
{
/* do when there is no more left */
SRF_RETURN_DONE(funcctx);
}
}
/* generate_series_timestamptz()
* Generate the set of timestamps from start to finish by step
*/
Datum
generate_series_timestamptz(PG_FUNCTION_ARGS)
{
FuncCallContext *funcctx;
generate_series_timestamptz_fctx *fctx;
TimestampTz result;
/* stuff done only on the first call of the function */
if (SRF_IS_FIRSTCALL())
{
TimestampTz start = PG_GETARG_TIMESTAMPTZ(0);
TimestampTz finish = PG_GETARG_TIMESTAMPTZ(1);
8.4 pgindent run, with new combined Linux/FreeBSD/MinGW typedef list provided by Andrew.
2009-06-11 16:49:15 +02:00
Interval *step = PG_GETARG_INTERVAL_P(2);
Add timestamp and timestamptz versions of generate_series(). Hitoshi Harada
2008-05-05 01:19:24 +02:00
MemoryContext oldcontext;
8.4 pgindent run, with new combined Linux/FreeBSD/MinGW typedef list provided by Andrew.
2009-06-11 16:49:15 +02:00
Interval interval_zero;
Add timestamp and timestamptz versions of generate_series(). Hitoshi Harada
2008-05-05 01:19:24 +02:00
/* create a function context for cross-call persistence */
funcctx = SRF_FIRSTCALL_INIT();
/*
* switch to memory context appropriate for multiple function calls
*/
oldcontext = MemoryContextSwitchTo(funcctx->multi_call_memory_ctx);
/* allocate memory for user context */
fctx = (generate_series_timestamptz_fctx *)
palloc(sizeof(generate_series_timestamptz_fctx));
/*
* Use fctx to keep state from call to call. Seed current with the
* original start value
*/
fctx->current = start;
fctx->finish = finish;
fctx->step = *step;
/* Determine sign of the interval */
MemSet(&interval_zero, 0, sizeof(Interval));
fctx->step_sign = interval_cmp_internal(&fctx->step, &interval_zero);
if (fctx->step_sign == 0)
ereport(ERROR,
(errcode(ERRCODE_INVALID_PARAMETER_VALUE),
errmsg("step size cannot equal zero")));
funcctx->user_fctx = fctx;
MemoryContextSwitchTo(oldcontext);
}
/* stuff done on every call of the function */
funcctx = SRF_PERCALL_SETUP();
/*
* get the saved state and use current as the result for this iteration
*/
fctx = funcctx->user_fctx;
result = fctx->current;
if (fctx->step_sign > 0 ?
timestamp_cmp_internal(result, fctx->finish) <= 0 :
timestamp_cmp_internal(result, fctx->finish) >= 0)
{
/* increment current in preparation for next iteration */
Clean up newlines following left parentheses We used to strategically place newlines after some function call left parentheses to make pgindent move the argument list a few chars to the left, so that the whole line would fit under 80 chars. However, pgindent no longer does that, so the newlines just made the code vertically longer for no reason. Remove those newlines, and reflow some of those lines for some extra naturality. Reviewed-by: Michael Paquier, Tom Lane Discussion: https://postgr.es/m/20200129200401.GA6303@alvherre.pgsql
2020-01-30 17:42:14 +01:00
fctx->current = DatumGetTimestampTz(DirectFunctionCall2(timestamptz_pl_interval,
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
TimestampTzGetDatum(fctx->current),
PointerGetDatum(&fctx->step)));
Add timestamp and timestamptz versions of generate_series(). Hitoshi Harada
2008-05-05 01:19:24 +02:00
/* do when there is more left to send */
SRF_RETURN_NEXT(funcctx, TimestampTzGetDatum(result));
}
else
{
/* do when there is no more left */
SRF_RETURN_DONE(funcctx);
}
}