mirror of
https://git.postgresql.org/git/postgresql.git
synced 2024-09-10 06:39:45 +02:00
92df220343
Get rid of section 8.5.6 (Date/Time Internals), which appears to confuse people more than it helps, and anyway discussion of Postgres' internal datetime calculation methods seems pretty out of place here. Instead, make datatype.sgml just say that we follow the Gregorian calendar (a bit of specification not previously present anywhere in that chapter :-() and link to the History of Units appendix for more info. Do some mild editorialization on that appendix, too, to make it clearer that we are following proleptic Gregorian calendar rules rather than anything more historically accurate. Per a question from Florence Cousin and subsequent discussion in pgsql-docs.
595 lines
18 KiB
Plaintext
595 lines
18 KiB
Plaintext
<!-- doc/src/sgml/datetime.sgml -->
|
|
|
|
<appendix id="datetime-appendix">
|
|
<title>Date/Time Support</title>
|
|
|
|
<para>
|
|
<productname>PostgreSQL</productname> uses an internal heuristic
|
|
parser for all date/time input support. Dates and times are input as
|
|
strings, and are broken up into distinct fields with a preliminary
|
|
determination of what kind of information can be in the
|
|
field. Each field is interpreted and either assigned a numeric
|
|
value, ignored, or rejected.
|
|
The parser contains internal lookup tables for all textual fields,
|
|
including months, days of the week, and time zones.
|
|
</para>
|
|
|
|
<para>
|
|
This appendix includes information on the content of these
|
|
lookup tables and describes the steps used by the parser to decode
|
|
dates and times.
|
|
</para>
|
|
|
|
<sect1 id="datetime-input-rules">
|
|
<title>Date/Time Input Interpretation</title>
|
|
|
|
<para>
|
|
The date/time type inputs are all decoded using the following procedure.
|
|
</para>
|
|
|
|
<procedure>
|
|
<step>
|
|
<para>
|
|
Break the input string into tokens and categorize each token as
|
|
a string, time, time zone, or number.
|
|
</para>
|
|
|
|
<substeps>
|
|
<step>
|
|
<para>
|
|
If the numeric token contains a colon (<literal>:</>), this is
|
|
a time string. Include all subsequent digits and colons.
|
|
</para>
|
|
</step>
|
|
|
|
<step>
|
|
<para>
|
|
If the numeric token contains a dash (<literal>-</>), slash
|
|
(<literal>/</>), or two or more dots (<literal>.</>), this is
|
|
a date string which might have a text month. If a date token has
|
|
already been seen, it is instead interpreted as a time zone
|
|
name (e.g., <literal>America/New_York</>).
|
|
</para>
|
|
</step>
|
|
|
|
<step>
|
|
<para>
|
|
If the token is numeric only, then it is either a single field
|
|
or an ISO 8601 concatenated date (e.g.,
|
|
<literal>19990113</literal> for January 13, 1999) or time
|
|
(e.g., <literal>141516</literal> for 14:15:16).
|
|
</para>
|
|
</step>
|
|
|
|
<step>
|
|
<para>
|
|
If the token starts with a plus (<literal>+</>) or minus
|
|
(<literal>-</>), then it is either a numeric time zone or a special
|
|
field.
|
|
</para>
|
|
</step>
|
|
</substeps>
|
|
</step>
|
|
|
|
<step>
|
|
<para>
|
|
If the token is a text string, match up with possible strings:
|
|
</para>
|
|
|
|
<substeps>
|
|
<step>
|
|
<para>
|
|
Do a binary-search table lookup for the token as a time zone
|
|
abbreviation.
|
|
</para>
|
|
</step>
|
|
|
|
<step>
|
|
<para>
|
|
If not found, do a similar binary-search table lookup to match
|
|
the token as either a special string (e.g., <literal>today</literal>),
|
|
day (e.g., <literal>Thursday</literal>),
|
|
month (e.g., <literal>January</literal>),
|
|
or noise word (e.g., <literal>at</literal>, <literal>on</literal>).
|
|
</para>
|
|
</step>
|
|
|
|
<step>
|
|
<para>
|
|
If still not found, throw an error.
|
|
</para>
|
|
</step>
|
|
</substeps>
|
|
</step>
|
|
|
|
<step>
|
|
<para>
|
|
When the token is a number or number field:
|
|
</para>
|
|
|
|
<substeps>
|
|
<step>
|
|
<para>
|
|
If there are eight or six digits,
|
|
and if no other date fields have been previously read, then interpret
|
|
as a <quote>concatenated date</quote> (e.g.,
|
|
<literal>19990118</literal> or <literal>990118</literal>).
|
|
The interpretation is <literal>YYYYMMDD</> or <literal>YYMMDD</>.
|
|
</para>
|
|
</step>
|
|
|
|
<step>
|
|
<para>
|
|
If the token is three digits
|
|
and a year has already been read, then interpret as day of year.
|
|
</para>
|
|
</step>
|
|
|
|
<step>
|
|
<para>
|
|
If four or six digits and a year has already been read, then
|
|
interpret as a time (<literal>HHMM</> or <literal>HHMMSS</>).
|
|
</para>
|
|
</step>
|
|
|
|
<step>
|
|
<para>
|
|
If three or more digits and no date fields have yet been found,
|
|
interpret as a year (this forces yy-mm-dd ordering of the remaining
|
|
date fields).
|
|
</para>
|
|
</step>
|
|
|
|
<step>
|
|
<para>
|
|
Otherwise the date field ordering is assumed to follow the
|
|
<varname>DateStyle</> setting: mm-dd-yy, dd-mm-yy, or yy-mm-dd.
|
|
Throw an error if a month or day field is found to be out of range.
|
|
</para>
|
|
</step>
|
|
</substeps>
|
|
</step>
|
|
|
|
<step>
|
|
<para>
|
|
If BC has been specified, negate the year and add one for
|
|
internal storage. (There is no year zero in the Gregorian
|
|
calendar, so numerically 1 BC becomes year zero.)
|
|
</para>
|
|
</step>
|
|
|
|
<step>
|
|
<para>
|
|
If BC was not specified, and if the year field was two digits in length,
|
|
then adjust the year to four digits. If the field is less than 70, then
|
|
add 2000, otherwise add 1900.
|
|
|
|
<tip>
|
|
<para>
|
|
Gregorian years AD 1-99 can be entered by using 4 digits with leading
|
|
zeros (e.g., <literal>0099</> is AD 99).
|
|
</para>
|
|
</tip>
|
|
</para>
|
|
</step>
|
|
</procedure>
|
|
</sect1>
|
|
|
|
|
|
<sect1 id="datetime-keywords">
|
|
<title>Date/Time Key Words</title>
|
|
|
|
<para>
|
|
<xref linkend="datetime-month-table"> shows the tokens that are
|
|
recognized as names of months.
|
|
</para>
|
|
|
|
<table id="datetime-month-table">
|
|
<title>Month Names</title>
|
|
<tgroup cols="2">
|
|
<thead>
|
|
<row>
|
|
<entry>Month</entry>
|
|
<entry>Abbreviations</entry>
|
|
</row>
|
|
</thead>
|
|
<tbody>
|
|
<row>
|
|
<entry>January</entry>
|
|
<entry>Jan</entry>
|
|
</row>
|
|
<row>
|
|
<entry>February</entry>
|
|
<entry>Feb</entry>
|
|
</row>
|
|
<row>
|
|
<entry>March</entry>
|
|
<entry>Mar</entry>
|
|
</row>
|
|
<row>
|
|
<entry>April</entry>
|
|
<entry>Apr</entry>
|
|
</row>
|
|
<row>
|
|
<entry>May</entry>
|
|
<entry></entry>
|
|
</row>
|
|
<row>
|
|
<entry>June</entry>
|
|
<entry>Jun</entry>
|
|
</row>
|
|
<row>
|
|
<entry>July</entry>
|
|
<entry>Jul</entry>
|
|
</row>
|
|
<row>
|
|
<entry>August</entry>
|
|
<entry>Aug</entry>
|
|
</row>
|
|
<row>
|
|
<entry>September</entry>
|
|
<entry>Sep, Sept</entry>
|
|
</row>
|
|
<row>
|
|
<entry>October</entry>
|
|
<entry>Oct</entry>
|
|
</row>
|
|
<row>
|
|
<entry>November</entry>
|
|
<entry>Nov</entry>
|
|
</row>
|
|
<row>
|
|
<entry>December</entry>
|
|
<entry>Dec</entry>
|
|
</row>
|
|
</tbody>
|
|
</tgroup>
|
|
</table>
|
|
|
|
<para>
|
|
<xref linkend="datetime-dow-table"> shows the tokens that are
|
|
recognized as names of days of the week.
|
|
</para>
|
|
|
|
<table id="datetime-dow-table">
|
|
<title>Day of the Week Names</title>
|
|
<tgroup cols="2">
|
|
<thead>
|
|
<row>
|
|
<entry>Day</entry>
|
|
<entry>Abbreviations</entry>
|
|
</row>
|
|
</thead>
|
|
<tbody>
|
|
<row>
|
|
<entry>Sunday</entry>
|
|
<entry>Sun</entry>
|
|
</row>
|
|
<row>
|
|
<entry>Monday</entry>
|
|
<entry>Mon</entry>
|
|
</row>
|
|
<row>
|
|
<entry>Tuesday</entry>
|
|
<entry>Tue, Tues</entry>
|
|
</row>
|
|
<row>
|
|
<entry>Wednesday</entry>
|
|
<entry>Wed, Weds</entry>
|
|
</row>
|
|
<row>
|
|
<entry>Thursday</entry>
|
|
<entry>Thu, Thur, Thurs</entry>
|
|
</row>
|
|
<row>
|
|
<entry>Friday</entry>
|
|
<entry>Fri</entry>
|
|
</row>
|
|
<row>
|
|
<entry>Saturday</entry>
|
|
<entry>Sat</entry>
|
|
</row>
|
|
</tbody>
|
|
</tgroup>
|
|
</table>
|
|
|
|
<para>
|
|
<xref linkend="datetime-mod-table"> shows the tokens that serve
|
|
various modifier purposes.
|
|
</para>
|
|
|
|
<table id="datetime-mod-table">
|
|
<title>Date/Time Field Modifiers</title>
|
|
<tgroup cols="2">
|
|
<thead>
|
|
<row>
|
|
<entry>Identifier</entry>
|
|
<entry>Description</entry>
|
|
</row>
|
|
</thead>
|
|
<tbody>
|
|
<row>
|
|
<entry><literal>AM</literal></entry>
|
|
<entry>Time is before 12:00</entry>
|
|
</row>
|
|
<row>
|
|
<entry><literal>AT</literal></entry>
|
|
<entry>Ignored</entry>
|
|
</row>
|
|
<row>
|
|
<entry><literal>JULIAN</>, <literal>JD</>, <literal>J</></entry>
|
|
<entry>Next field is Julian Date</entry>
|
|
</row>
|
|
<row>
|
|
<entry><literal>ON</literal></entry>
|
|
<entry>Ignored</entry>
|
|
</row>
|
|
<row>
|
|
<entry><literal>PM</literal></entry>
|
|
<entry>Time is on or after 12:00</entry>
|
|
</row>
|
|
<row>
|
|
<entry><literal>T</literal></entry>
|
|
<entry>Next field is time</entry>
|
|
</row>
|
|
</tbody>
|
|
</tgroup>
|
|
</table>
|
|
</sect1>
|
|
|
|
<sect1 id="datetime-config-files">
|
|
<title>Date/Time Configuration Files</title>
|
|
|
|
<indexterm>
|
|
<primary>time zone</primary>
|
|
<secondary>input abbreviations</secondary>
|
|
</indexterm>
|
|
|
|
<para>
|
|
Since timezone abbreviations are not well standardized,
|
|
<productname>PostgreSQL</productname> provides a means to customize
|
|
the set of abbreviations accepted by the server. The
|
|
<xref linkend="guc-timezone-abbreviations"> run-time parameter
|
|
determines the active set of abbreviations. While this parameter
|
|
can be altered by any database user, the possible values for it
|
|
are under the control of the database administrator — they
|
|
are in fact names of configuration files stored in
|
|
<filename>.../share/timezonesets/</> of the installation directory.
|
|
By adding or altering files in that directory, the administrator
|
|
can set local policy for timezone abbreviations.
|
|
</para>
|
|
|
|
<para>
|
|
<varname>timezone_abbreviations</> can be set to any file name
|
|
found in <filename>.../share/timezonesets/</>, if the file's name
|
|
is entirely alphabetic. (The prohibition against non-alphabetic
|
|
characters in <varname>timezone_abbreviations</> prevents reading
|
|
files outside the intended directory, as well as reading editor
|
|
backup files and other extraneous files.)
|
|
</para>
|
|
|
|
<para>
|
|
A timezone abbreviation file can contain blank lines and comments
|
|
beginning with <literal>#</>. Non-comment lines must have one of
|
|
these formats:
|
|
|
|
<synopsis>
|
|
<replaceable>time_zone_name</replaceable> <replaceable>offset</replaceable>
|
|
<replaceable>time_zone_name</replaceable> <replaceable>offset</replaceable> D
|
|
@INCLUDE <replaceable>file_name</replaceable>
|
|
@OVERRIDE
|
|
</synopsis>
|
|
</para>
|
|
|
|
<para>
|
|
A <replaceable>time_zone_name</replaceable> is just the abbreviation
|
|
being defined. The <replaceable>offset</replaceable> is the zone's
|
|
offset in seconds from UTC, positive being east from Greenwich and
|
|
negative being west. For example, -18000 would be five hours west
|
|
of Greenwich, or North American east coast standard time. <literal>D</>
|
|
indicates that the zone name represents local daylight-savings time
|
|
rather than standard time. Since all known time zone offsets are on
|
|
15 minute boundaries, the number of seconds has to be a multiple of 900.
|
|
</para>
|
|
|
|
<para>
|
|
The <literal>@INCLUDE</> syntax allows inclusion of another file in the
|
|
<filename>.../share/timezonesets/</> directory. Inclusion can be nested,
|
|
to a limited depth.
|
|
</para>
|
|
|
|
<para>
|
|
The <literal>@OVERRIDE</> syntax indicates that subsequent entries in the
|
|
file can override previous entries (i.e., entries obtained from included
|
|
files). Without this, conflicting definitions of the same timezone
|
|
abbreviation are considered an error.
|
|
</para>
|
|
|
|
<para>
|
|
In an unmodified installation, the file <filename>Default</> contains
|
|
all the non-conflicting time zone abbreviations for most of the world.
|
|
Additional files <filename>Australia</> and <filename>India</> are
|
|
provided for those regions: these files first include the
|
|
<literal>Default</> file and then add or modify timezones as needed.
|
|
</para>
|
|
|
|
<para>
|
|
For reference purposes, a standard installation also contains files
|
|
<filename>Africa.txt</>, <filename>America.txt</>, etc, containing
|
|
information about every time zone abbreviation known to be in use
|
|
according to the <literal>zoneinfo</> timezone database. The zone name
|
|
definitions found in these files can be copied and pasted into a custom
|
|
configuration file as needed. Note that these files cannot be directly
|
|
referenced as <varname>timezone_abbreviations</> settings, because of
|
|
the dot embedded in their names.
|
|
</para>
|
|
|
|
<note>
|
|
<para>
|
|
If an error occurs while reading the time zone data sets, no new value is
|
|
applied but the old set is kept. If the error occurs while starting the
|
|
database, startup fails.
|
|
</para>
|
|
</note>
|
|
|
|
<caution>
|
|
<para>
|
|
Time zone abbreviations defined in the configuration file override
|
|
non-timezone meanings built into <productname>PostgreSQL</productname>.
|
|
For example, the <filename>Australia</> configuration file defines
|
|
<literal>SAT</> (for South Australian Standard Time). When this
|
|
file is active, <literal>SAT</> will not be recognized as an abbreviation
|
|
for Saturday.
|
|
</para>
|
|
</caution>
|
|
|
|
<caution>
|
|
<para>
|
|
If you modify files in <filename>.../share/timezonesets/</>,
|
|
it is up to you to make backups — a normal database dump
|
|
will not include this directory.
|
|
</para>
|
|
</caution>
|
|
|
|
</sect1>
|
|
|
|
<sect1 id="datetime-units-history">
|
|
<title>History of Units</title>
|
|
|
|
<indexterm zone="datetime-units-history">
|
|
<primary>Gregorian calendar</primary>
|
|
</indexterm>
|
|
<indexterm zone="datetime-units-history">
|
|
<primary>Julian date</primary>
|
|
</indexterm>
|
|
|
|
<para>
|
|
The SQL standard states that <quote>Within the definition of a
|
|
<quote>datetime literal</quote>, the <quote>datetime
|
|
values</quote> are constrained by the natural rules for dates and
|
|
times according to the Gregorian calendar</quote>.
|
|
<productname>PostgreSQL</> follows the SQL
|
|
standard's lead by counting dates exclusively in the Gregorian
|
|
calendar, even for years before that calendar was in use.
|
|
This rule is known as the <firstterm>proleptic Gregorian calendar</>.
|
|
</para>
|
|
|
|
<para>
|
|
The Julian calendar was introduced by Julius Caesar in 45 BC.
|
|
It was in common use in the Western world
|
|
until the year 1582, when countries started changing to the Gregorian
|
|
calendar. In the Julian calendar, the tropical year is
|
|
approximated as 365 1/4 days = 365.25 days. This gives an error of
|
|
about 1 day in 128 years.
|
|
</para>
|
|
|
|
<para>
|
|
The accumulating calendar error prompted
|
|
Pope Gregory XIII to reform the calendar in accordance with
|
|
instructions from the Council of Trent.
|
|
In the Gregorian calendar, the tropical year is approximated as
|
|
365 + 97 / 400 days = 365.2425 days. Thus it takes approximately 3300
|
|
years for the tropical year to shift one day with respect to the
|
|
Gregorian calendar.
|
|
</para>
|
|
|
|
<para>
|
|
The approximation 365+97/400 is achieved by having 97 leap years
|
|
every 400 years, using the following rules:
|
|
|
|
<simplelist>
|
|
<member>
|
|
Every year divisible by 4 is a leap year.
|
|
</member>
|
|
<member>
|
|
However, every year divisible by 100 is not a leap year.
|
|
</member>
|
|
<member>
|
|
However, every year divisible by 400 is a leap year after all.
|
|
</member>
|
|
</simplelist>
|
|
|
|
So, 1700, 1800, 1900, 2100, and 2200 are not leap years. But 1600,
|
|
2000, and 2400 are leap years.
|
|
|
|
By contrast, in the older Julian calendar all years divisible by 4 are leap
|
|
years.
|
|
</para>
|
|
|
|
<para>
|
|
The papal bull of February 1582 decreed that 10 days should be dropped
|
|
from October 1582 so that 15 October should follow immediately after
|
|
4 October.
|
|
This was observed in Italy, Poland, Portugal, and Spain. Other Catholic
|
|
countries followed shortly after, but Protestant countries were
|
|
reluctant to change, and the Greek Orthodox countries didn't change
|
|
until the start of the 20th century.
|
|
|
|
The reform was observed by Great Britain and its dominions (including what
|
|
is now the USA) in 1752.
|
|
Thus 2 September 1752 was followed by 14 September 1752.
|
|
|
|
This is why Unix systems have the <command>cal</command> program
|
|
produce the following:
|
|
|
|
<screen>
|
|
$ <userinput>cal 9 1752</userinput>
|
|
September 1752
|
|
S M Tu W Th F S
|
|
1 2 14 15 16
|
|
17 18 19 20 21 22 23
|
|
24 25 26 27 28 29 30
|
|
</screen>
|
|
|
|
But, of course, this calendar is only valid for Great Britain and
|
|
dominions, not other places.
|
|
Since it would be difficult and confusing to try to track the actual
|
|
calendars that were in use in various places at various times,
|
|
<productname>PostgreSQL</> does not try, but rather follows the Gregorian
|
|
calendar rules for all dates, even though this method is not historically
|
|
accurate.
|
|
</para>
|
|
|
|
<para>
|
|
Different calendars have been developed in various parts of the
|
|
world, many predating the Gregorian system.
|
|
|
|
For example,
|
|
the beginnings of the Chinese calendar can be traced back to the 14th
|
|
century BC. Legend has it that the Emperor Huangdi invented that
|
|
calendar in 2637 BC.
|
|
|
|
The People's Republic of China uses the Gregorian calendar
|
|
for civil purposes. The Chinese calendar is used for determining
|
|
festivals.
|
|
</para>
|
|
|
|
<para>
|
|
The <firstterm>Julian Date</firstterm> system is another type of
|
|
calendar, unrelated to the Julian calendar though it is confusingly
|
|
named similarly to that calendar.
|
|
The Julian Date system was invented by the French scholar
|
|
Joseph Justus Scaliger (1540-1609)
|
|
and probably takes its name from Scaliger's father,
|
|
the Italian scholar Julius Caesar Scaliger (1484-1558).
|
|
In the Julian Date system, each day has a sequential number, starting
|
|
from JD 0 (which is sometimes called <emphasis>the</> Julian Date).
|
|
JD 0 corresponds to 1 January 4713 BC in the Julian calendar, or
|
|
24 November 4714 BC in the Gregorian calendar. Julian Date counting
|
|
is most often used by astronomers for labeling their nightly observations,
|
|
and therefore a date runs from noon UTC to the next noon UTC, rather than
|
|
from midnight to midnight: JD 0 designates the 24 hours from noon UTC on
|
|
24 November 4714 BC to noon UTC on 25 November 4714 BC.
|
|
</para>
|
|
|
|
<para>
|
|
Although <productname>PostgreSQL</> supports Julian Date notation for
|
|
input and output of dates (and also uses Julian dates for some internal
|
|
datetime calculations), it does not observe the nicety of having dates
|
|
run from noon to noon. <productname>PostgreSQL</> treats a Julian Date
|
|
as running from midnight to midnight.
|
|
</para>
|
|
|
|
</sect1>
|
|
</appendix>
|