Fix handling of BC years in to_date/to_timestamp.

Previously, a conversion such as
	to_date('-44-02-01','YYYY-MM-DD')
would result in '0045-02-01 BC', as the code attempted to interpret
the negative year as BC, but failed to apply the correction needed
for our internal handling of BC years.  Fix the off-by-one problem.

Also, arrange for the combination of a negative year and an
explicit "BC" marker to cancel out and produce AD.  This is how
the negative-century case works, so it seems sane to do likewise.

Continue to read "year 0000" as 1 BC.  Oracle would throw an error,
but we've accepted that case for a long time so I'm hesitant to
change it in a back-patch.

Per bug #16419 from Saeed Hubaishan.  Back-patch to all supported
branches.

Dar Alathar-Yemen and Tom Lane

Discussion: https://postgr.es/m/16419-d8d9db0a7553f01b@postgresql.org

diff --git a/doc/src/sgml/func.sgml b/doc/src/sgml/func.sgml
index 258c5bed21fc56f9a4e1dc2531b255d56db959b7..043c2f841e1fba2640c062f3f5ce8a1992d233d1 100644 (file)
--- a/doc/src/sgml/func.sgml
+++ b/doc/src/sgml/func.sgml
@@ -5933,8 +5933,18 @@ SELECT regexp_matches('abc01234xyz', '(?:(.*?)(\d+)(.*)){1,1}');
 
      <listitem>
       <para>
-       The <literal>YYYY</literal> conversion from string to <type>timestamp</type> or
-       <type>date</type> has a restriction when processing years with more than 4 digits. You must
+       In <function>to_timestamp</function> and <function>to_date</function>,
+       negative years are treated as signifying BC.  If you write both a
+       negative year and an explicit <literal>BC</literal> field, you get AD
+       again.  An input of year zero is treated as 1 BC.
+      </para>
+     </listitem>
+
+     <listitem>
+      <para>
+       In <function>to_timestamp</function> and <function>to_date</function>,
+       the <literal>YYYY</literal> conversion has a restriction when
+       processing years with more than 4 digits. You must
        use some non-digit character or template after <literal>YYYY</literal>,
        otherwise the year is always interpreted as 4 digits. For example
        (with the year 20000):

diff --git a/src/backend/utils/adt/formatting.c b/src/backend/utils/adt/formatting.c
index bb8d15a972260fa78a0632471a503c1cd7b6c843..abf8c050af577db64a31396cf04781f9372e7368 100644 (file)
--- a/src/backend/utils/adt/formatting.c
+++ b/src/backend/utils/adt/formatting.c
@@ -3678,8 +3678,11 @@ do_to_timestamp(text *date_txt, text *fmt,
                        /* If a 4-digit year is provided, we use that and ignore CC. */
                {
                        tm->tm_year = tmfc.year;
-                       if (tmfc.bc && tm->tm_year > 0)
-                               tm->tm_year = -(tm->tm_year - 1);
+                       if (tmfc.bc)
+                               tm->tm_year = -tm->tm_year;
+                       /* correct for our representation of BC years */
+                       if (tm->tm_year < 0)
+                               tm->tm_year++;
                }
        }
        else if (tmfc.cc)                       /* use first year of century */

diff --git a/src/test/regress/expected/horology.out b/src/test/regress/expected/horology.out
index 28ef3d2f405ea7079102070d42e407ead29c4cbe..699e254a21d49b4aad062afaaf307177e21a1ade 100644 (file)
--- a/src/test/regress/expected/horology.out
+++ b/src/test/regress/expected/horology.out
@@ -2957,6 +2957,45 @@ SELECT to_date('2458872', 'J');
  01-23-2020
 (1 row)
 
+--
+-- Check handling of BC dates
+--
+SELECT to_date('44-02-01 BC','YYYY-MM-DD BC');
+    to_date    
+---------------
+ 02-01-0044 BC
+(1 row)
+
+SELECT to_date('-44-02-01','YYYY-MM-DD');
+    to_date    
+---------------
+ 02-01-0044 BC
+(1 row)
+
+SELECT to_date('-44-02-01 BC','YYYY-MM-DD BC');
+  to_date   
+------------
+ 02-01-0044
+(1 row)
+
+SELECT to_timestamp('44-02-01 11:12:13 BC','YYYY-MM-DD HH24:MI:SS BC');
+          to_timestamp           
+---------------------------------
+ Fri Feb 01 11:12:13 0044 PST BC
+(1 row)
+
+SELECT to_timestamp('-44-02-01 11:12:13','YYYY-MM-DD HH24:MI:SS');
+          to_timestamp           
+---------------------------------
+ Fri Feb 01 11:12:13 0044 PST BC
+(1 row)
+
+SELECT to_timestamp('-44-02-01 11:12:13 BC','YYYY-MM-DD HH24:MI:SS BC');
+         to_timestamp         
+------------------------------
+ Mon Feb 01 11:12:13 0044 PST
+(1 row)
+
 --
 -- Check handling of multiple spaces in format and/or input
 --

diff --git a/src/test/regress/sql/horology.sql b/src/test/regress/sql/horology.sql
index a1cd2eea1742a1de4defd4e96448246051a856b9..b0402ae4fe203ee56a5da666b81e4fe7a5e8e411 100644 (file)
--- a/src/test/regress/sql/horology.sql
+++ b/src/test/regress/sql/horology.sql
@@ -455,6 +455,17 @@ SELECT to_date('1 4 1902', 'Q MM YYYY');  -- Q is ignored
 SELECT to_date('3 4 21 01', 'W MM CC YY');
 SELECT to_date('2458872', 'J');
 
+--
+-- Check handling of BC dates
+--
+
+SELECT to_date('44-02-01 BC','YYYY-MM-DD BC');
+SELECT to_date('-44-02-01','YYYY-MM-DD');
+SELECT to_date('-44-02-01 BC','YYYY-MM-DD BC');
+SELECT to_timestamp('44-02-01 11:12:13 BC','YYYY-MM-DD HH24:MI:SS BC');
+SELECT to_timestamp('-44-02-01 11:12:13','YYYY-MM-DD HH24:MI:SS');
+SELECT to_timestamp('-44-02-01 11:12:13 BC','YYYY-MM-DD HH24:MI:SS BC');
+
 --
 -- Check handling of multiple spaces in format and/or input
 --