Future-proof regression tests against possibly-missing posixrules file.

The IANA time zone folk have deprecated use of a "posixrules" file in
the tz database.  While for now it's our choice whether to keep
supplying one in our own builds, installations built with
--with-system-tzdata will soon be needing to cope with that file not
being present, at least on some platforms.

This causes a problem for the horology test, which expected the
nonstandard POSIX zone spec "CST7CDT" to apply pre-2007 US daylight
savings rules.  That does happen if the posixrules file supplies such
information, but otherwise the test produces undesired results.
To fix, add an explicit transition date rule that matches 2005 practice.
(We could alternatively have switched the test to use some real time
zone, but it seems useful to have coverage of this type of zone spec.)

While at it, update a documentation example that also relied on
"CST7CDT"; use a real-world zone name instead.  Also, document why
the zone names EST5EDT, CST6CDT, MST7MDT, PST8PDT aren't subject to
similar failures when "posixrules" is missing.

Back-patch to all supported branches, since the hazard is the same
for all.

Discussion: https://postgr.es/m/1665379.1592581287@sss.pgh.pa.us

4 files changed, 29 insertions, 11 deletions

diff --git a/doc/src/sgml/datetime.sgml b/doc/src/sgml/datetime.sgml
index 7da4d0b7789..71fbf842cca 100644
--- a/doc/src/sgml/datetime.sgml
+++ b/doc/src/sgml/datetime.sgml
@@ -758,6 +758,18 @@
   </para>
 
   <para>
+   The four timezone names <literal>EST5EDT</literal>,
+   <literal>CST6CDT</literal>, <literal>MST7MDT</literal>,
+   and <literal>PST8PDT</literal> look like they are POSIX zone
+   specifications.  However, they actually are treated as named time zones
+   because (for historical reasons) there are files by those names in the
+   IANA time zone database.  The practical implication of this is that
+   these zone names will produce valid historical USA daylight-savings
+   transitions, even when a plain POSIX specification would not due to
+   lack of a suitable <filename>posixrules</filename> file.
+  </para>
+
+  <para>
    One should be wary that it is easy to misspell a POSIX-style time zone
    specification, since there is no check on the reasonableness of the
    zone abbreviation(s).  For example, <literal>SET TIMEZONE TO
diff --git a/doc/src/sgml/func.sgml b/doc/src/sgml/func.sgml
index a8d57f4e39a..aa235f139e3 100644
--- a/doc/src/sgml/func.sgml
+++ b/doc/src/sgml/func.sgml
@@ -9114,18 +9114,22 @@ SELECT (DATE '2001-10-30', DATE '2001-10-30') OVERLAPS
    When adding an <type>interval</type> value to (or subtracting an
    <type>interval</type> value from) a <type>timestamp with time zone</type>
    value, the days component advances or decrements the date of the
-   <type>timestamp with time zone</type> by the indicated number of days.
+   <type>timestamp with time zone</type> by the indicated number of days,
+   keeping the time of day the same.
    Across daylight saving time changes (when the session time zone is set to a
    time zone that recognizes DST), this means <literal>interval '1 day'</literal>
    does not necessarily equal <literal>interval '24 hours'</literal>.
-   For example, with the session time zone set to <literal>CST7CDT</literal>,
-   <literal>timestamp with time zone '2005-04-02 12:00-07' + interval '1 day'</literal>
-   will produce <literal>timestamp with time zone '2005-04-03 12:00-06'</literal>,
-   while adding <literal>interval '24 hours'</literal> to the same initial
-   <type>timestamp with time zone</type> produces
-   <literal>timestamp with time zone '2005-04-03 13:00-06'</literal>, as there is
-   a change in daylight saving time at <literal>2005-04-03 02:00</literal> in time zone
-   <literal>CST7CDT</literal>.
+   For example, with the session time zone set
+   to <literal>America/Denver</literal>:
+<screen>
+SELECT timestamp with time zone '2005-04-02 12:00:00-07' + interval '1 day';
+<lineannotation>Result: </lineannotation><computeroutput>2005-04-03 12:00:00-06</computeroutput>
+SELECT timestamp with time zone '2005-04-02 12:00:00-07' + interval '24 hours';
+<lineannotation>Result: </lineannotation><computeroutput>2005-04-03 13:00:00-06</computeroutput>
+</screen>
+   This happens because an hour was skipped due to a change in daylight saving
+   time at <literal>2005-04-03 02:00:00</literal> in time zone
+   <literal>America/Denver</literal>.
   </para>
 
   <para>
diff --git a/src/test/regress/expected/horology.out b/src/test/regress/expected/horology.out
index f67d624ad5d..c8c33a0fc06 100644
--- a/src/test/regress/expected/horology.out
+++ b/src/test/regress/expected/horology.out
@@ -652,7 +652,8 @@ SELECT (timestamp with time zone 'tomorrow' > 'now') as "True";
 (1 row)
 
 -- timestamp with time zone, interval arithmetic around DST change
-SET TIME ZONE 'CST7CDT';
+-- (just for fun, let's use an intentionally nonstandard POSIX zone spec)
+SET TIME ZONE 'CST7CDT,M4.1.0,M10.5.0';
 SELECT timestamp with time zone '2005-04-02 12:00-07' + interval '1 day' as "Apr 3, 12:00";
          Apr 3, 12:00         
 ------------------------------
diff --git a/src/test/regress/sql/horology.sql b/src/test/regress/sql/horology.sql
index d1f6d5bfcd3..c464e6766c6 100644
--- a/src/test/regress/sql/horology.sql
+++ b/src/test/regress/sql/horology.sql
@@ -122,7 +122,8 @@ SELECT (timestamp with time zone 'tomorrow' = (timestamp with time zone 'yesterd
 SELECT (timestamp with time zone 'tomorrow' > 'now') as "True";
 
 -- timestamp with time zone, interval arithmetic around DST change
-SET TIME ZONE 'CST7CDT';
+-- (just for fun, let's use an intentionally nonstandard POSIX zone spec)
+SET TIME ZONE 'CST7CDT,M4.1.0,M10.5.0';
 SELECT timestamp with time zone '2005-04-02 12:00-07' + interval '1 day' as "Apr 3, 12:00";
 SELECT timestamp with time zone '2005-04-02 12:00-07' + interval '24 hours' as "Apr 3, 13:00";
 SELECT timestamp with time zone '2005-04-03 12:00-06' - interval '1 day' as "Apr 2, 12:00";