mirror of
https://git.postgresql.org/git/postgresql.git
synced 2025-01-06 15:24:56 +08:00
Remove CTimeZone/HasCTZSet, root and branch.
These variables no longer have any useful purpose, since there's no reason to special-case brute force timezones now that we have a valid session_timezone setting for them. Remove the variables, and remove the SET/SHOW TIME ZONE code that deals with them. The user-visible impact of this is that SHOW TIME ZONE will now show a POSIX-style zone specification, in the form "<+-offset>-+offset", rather than an interval value when a brute-force zone has been set. While perhaps less intuitive, this is a better definition than before because it's actually possible to give that string back to SET TIME ZONE and get the same behavior, unlike what used to happen. We did not previously mention the angle-bracket syntax when describing POSIX timezone specifications; add some documentation so that people can figure out what these strings do. (There's still quite a lot of undocumented functionality there, but anybody who really cares can go read the POSIX spec to find out about it. In practice most people seem to prefer Olsen-style city names anyway.)
This commit is contained in:
parent
1c8a7f617f
commit
45f64f1bbf
@ -2419,8 +2419,11 @@ January 8 04:05:06 1999 PST
|
||||
optional daylight-savings zone abbreviation, assumed to stand for one
|
||||
hour ahead of the given offset. For example, if <literal>EST5EDT</>
|
||||
were not already a recognized zone name, it would be accepted and would
|
||||
be functionally equivalent to United States East Coast time. When a
|
||||
daylight-savings zone name is present, it is assumed to be used
|
||||
be functionally equivalent to United States East Coast time. In this
|
||||
syntax, a zone abbreviation can be a string of letters, or an
|
||||
arbitrary string surrounded by angle brackets (<literal><></>).
|
||||
When a daylight-savings zone abbreviation is present,
|
||||
it is assumed to be used
|
||||
according to the same daylight-savings transition rules used in the
|
||||
<literal>zoneinfo</> time zone database's <filename>posixrules</> entry.
|
||||
In a standard <productname>PostgreSQL</productname> installation,
|
||||
|
@ -243,7 +243,16 @@ SELECT setseed(<replaceable>value</replaceable>);
|
||||
</listitem>
|
||||
</varlistentry>
|
||||
</variablelist>
|
||||
</para>
|
||||
|
||||
<para>
|
||||
Timezone settings given as numbers or intervals are internally
|
||||
translated to POSIX timezone syntax. For example, after
|
||||
<literal>SET TIME ZONE -7</>, <command>SHOW TIME ZONE</> would
|
||||
report <literal><-07>+07</>.
|
||||
</para>
|
||||
|
||||
<para>
|
||||
See <xref linkend="datatype-timezones"> for more information
|
||||
about time zones.
|
||||
</para>
|
||||
|
@ -243,34 +243,17 @@ assign_datestyle(const char *newval, void *extra)
|
||||
* TIMEZONE
|
||||
*/
|
||||
|
||||
typedef struct
|
||||
{
|
||||
pg_tz *session_timezone;
|
||||
int CTimeZone;
|
||||
bool HasCTZSet;
|
||||
} timezone_extra;
|
||||
|
||||
/*
|
||||
* check_timezone: GUC check_hook for timezone
|
||||
*/
|
||||
bool
|
||||
check_timezone(char **newval, void **extra, GucSource source)
|
||||
{
|
||||
timezone_extra myextra;
|
||||
pg_tz *new_tz;
|
||||
long gmtoffset;
|
||||
char *endptr;
|
||||
double hours;
|
||||
|
||||
/*
|
||||
* Initialize the "extra" struct that will be passed to assign_timezone.
|
||||
* We don't want to change any of the three global variables except as
|
||||
* specified by logic below. To avoid leaking memory during failure
|
||||
* returns, we set up the struct contents in a local variable, and only
|
||||
* copy it to *extra at the end.
|
||||
*/
|
||||
myextra.session_timezone = session_timezone;
|
||||
myextra.CTimeZone = CTimeZone;
|
||||
myextra.HasCTZSet = HasCTZSet;
|
||||
|
||||
if (pg_strncasecmp(*newval, "interval", 8) == 0)
|
||||
{
|
||||
/*
|
||||
@ -323,12 +306,11 @@ check_timezone(char **newval, void **extra, GucSource source)
|
||||
|
||||
/* Here we change from SQL to Unix sign convention */
|
||||
#ifdef HAVE_INT64_TIMESTAMP
|
||||
myextra.CTimeZone = -(interval->time / USECS_PER_SEC);
|
||||
gmtoffset = -(interval->time / USECS_PER_SEC);
|
||||
#else
|
||||
myextra.CTimeZone = -interval->time;
|
||||
gmtoffset = -interval->time;
|
||||
#endif
|
||||
myextra.session_timezone = pg_tzset_offset(myextra.CTimeZone);
|
||||
myextra.HasCTZSet = true;
|
||||
new_tz = pg_tzset_offset(gmtoffset);
|
||||
|
||||
pfree(interval);
|
||||
}
|
||||
@ -341,17 +323,14 @@ check_timezone(char **newval, void **extra, GucSource source)
|
||||
if (endptr != *newval && *endptr == '\0')
|
||||
{
|
||||
/* Here we change from SQL to Unix sign convention */
|
||||
myextra.CTimeZone = -hours * SECS_PER_HOUR;
|
||||
myextra.session_timezone = pg_tzset_offset(myextra.CTimeZone);
|
||||
myextra.HasCTZSet = true;
|
||||
gmtoffset = -hours * SECS_PER_HOUR;
|
||||
new_tz = pg_tzset_offset(gmtoffset);
|
||||
}
|
||||
else
|
||||
{
|
||||
/*
|
||||
* Otherwise assume it is a timezone name, and try to load it.
|
||||
*/
|
||||
pg_tz *new_tz;
|
||||
|
||||
new_tz = pg_tzset(*newval);
|
||||
|
||||
if (!new_tz)
|
||||
@ -367,40 +346,16 @@ check_timezone(char **newval, void **extra, GucSource source)
|
||||
GUC_check_errdetail("PostgreSQL does not support leap seconds.");
|
||||
return false;
|
||||
}
|
||||
|
||||
myextra.session_timezone = new_tz;
|
||||
myextra.HasCTZSet = false;
|
||||
}
|
||||
}
|
||||
|
||||
/*
|
||||
* Prepare the canonical string to return. GUC wants it malloc'd.
|
||||
*
|
||||
* Note: the result string should be something that we'd accept as input.
|
||||
* We use the numeric format for interval cases, because it's simpler to
|
||||
* reload. In the named-timezone case, *newval is already OK and need not
|
||||
* be changed; it might not have the canonical casing, but that's taken
|
||||
* care of by show_timezone.
|
||||
*/
|
||||
if (myextra.HasCTZSet)
|
||||
{
|
||||
char *result = (char *) malloc(64);
|
||||
|
||||
if (!result)
|
||||
return false;
|
||||
snprintf(result, 64, "%.5f",
|
||||
(double) (-myextra.CTimeZone) / (double) SECS_PER_HOUR);
|
||||
free(*newval);
|
||||
*newval = result;
|
||||
}
|
||||
|
||||
/*
|
||||
* Pass back data for assign_timezone to use
|
||||
*/
|
||||
*extra = malloc(sizeof(timezone_extra));
|
||||
*extra = malloc(sizeof(pg_tz *));
|
||||
if (!*extra)
|
||||
return false;
|
||||
memcpy(*extra, &myextra, sizeof(timezone_extra));
|
||||
*((pg_tz **) *extra) = new_tz;
|
||||
|
||||
return true;
|
||||
}
|
||||
@ -411,43 +366,19 @@ check_timezone(char **newval, void **extra, GucSource source)
|
||||
void
|
||||
assign_timezone(const char *newval, void *extra)
|
||||
{
|
||||
timezone_extra *myextra = (timezone_extra *) extra;
|
||||
|
||||
session_timezone = myextra->session_timezone;
|
||||
CTimeZone = myextra->CTimeZone;
|
||||
HasCTZSet = myextra->HasCTZSet;
|
||||
session_timezone = *((pg_tz **) extra);
|
||||
}
|
||||
|
||||
/*
|
||||
* show_timezone: GUC show_hook for timezone
|
||||
*
|
||||
* We wouldn't need this, except that historically interval values have been
|
||||
* shown without an INTERVAL prefix, so the display format isn't what would
|
||||
* be accepted as input. Otherwise we could have check_timezone return the
|
||||
* preferred string to begin with.
|
||||
*/
|
||||
const char *
|
||||
show_timezone(void)
|
||||
{
|
||||
const char *tzn;
|
||||
|
||||
if (HasCTZSet)
|
||||
{
|
||||
Interval interval;
|
||||
|
||||
interval.month = 0;
|
||||
interval.day = 0;
|
||||
#ifdef HAVE_INT64_TIMESTAMP
|
||||
interval.time = -(CTimeZone * USECS_PER_SEC);
|
||||
#else
|
||||
interval.time = -CTimeZone;
|
||||
#endif
|
||||
|
||||
tzn = DatumGetCString(DirectFunctionCall1(interval_out,
|
||||
IntervalPGetDatum(&interval)));
|
||||
}
|
||||
else
|
||||
tzn = pg_get_timezone_name(session_timezone);
|
||||
/* Always show the zone's canonical name */
|
||||
tzn = pg_get_timezone_name(session_timezone);
|
||||
|
||||
if (tzn != NULL)
|
||||
return tzn;
|
||||
@ -497,7 +428,7 @@ check_log_timezone(char **newval, void **extra, GucSource source)
|
||||
*extra = malloc(sizeof(pg_tz *));
|
||||
if (!*extra)
|
||||
return false;
|
||||
memcpy(*extra, &new_tz, sizeof(pg_tz *));
|
||||
*((pg_tz **) *extra) = new_tz;
|
||||
|
||||
return true;
|
||||
}
|
||||
@ -519,6 +450,7 @@ show_log_timezone(void)
|
||||
{
|
||||
const char *tzn;
|
||||
|
||||
/* Always show the zone's canonical name */
|
||||
tzn = pg_get_timezone_name(log_timezone);
|
||||
|
||||
if (tzn != NULL)
|
||||
|
@ -93,8 +93,6 @@ bool ExitOnAnyError = false;
|
||||
int DateStyle = USE_ISO_DATES;
|
||||
int DateOrder = DATEORDER_MDY;
|
||||
int IntervalStyle = INTSTYLE_POSTGRES;
|
||||
bool HasCTZSet = false;
|
||||
int CTimeZone = 0;
|
||||
|
||||
bool enableFsync = true;
|
||||
bool allowSystemTableMods = false;
|
||||
|
@ -219,15 +219,6 @@ extern int DateOrder;
|
||||
|
||||
extern int IntervalStyle;
|
||||
|
||||
/*
|
||||
* HasCTZSet is true if user has set timezone as a numeric offset from UTC.
|
||||
* If so, CTimeZone is the timezone offset in seconds (using the Unix-ish
|
||||
* sign convention, ie, positive offset is west of UTC, rather than the
|
||||
* SQL-ish convention that positive is east of UTC).
|
||||
*/
|
||||
extern bool HasCTZSet;
|
||||
extern int CTimeZone;
|
||||
|
||||
#define MAXTZLEN 10 /* max TZ name len, not counting tr. null */
|
||||
|
||||
extern bool enableFsync;
|
||||
|
@ -2941,9 +2941,9 @@ DETAIL: Value must be in the range -2147483648 to 2147483647.
|
||||
SET TIME ZONE 'America/New_York';
|
||||
SET TIME ZONE '-1.5';
|
||||
SHOW TIME ZONE;
|
||||
TimeZone
|
||||
----------------------
|
||||
@ 1 hour 30 mins ago
|
||||
TimeZone
|
||||
----------------
|
||||
<-01:30>+01:30
|
||||
(1 row)
|
||||
|
||||
SELECT '2012-12-12 12:00'::timestamptz;
|
||||
|
Loading…
Reference in New Issue
Block a user