Thread: pgsql: arrays: tighten checks for multi-dimensional input
arrays: tighten checks for multi-dimensional input Previously an input array string that started with a single-element array dimension would then later accept a multi-dimensional segment. BACKWARD INCOMPATIBILITY Branch ------ master Details ------- http://git.postgresql.org/pg/commitdiff/d0ee93797d9c06e619b3f382705d99ee195bb573 Modified Files -------------- src/backend/utils/adt/arrayfuncs.c | 8 ++++---- src/test/regress/expected/arrays.out | 2 +- src/test/regress/sql/arrays.sql | 2 +- 3 files changed, 6 insertions(+), 6 deletions(-)
On Sat, 2014-02-01 at 15:49 +0000, Bruce Momjian wrote:
> arrays: tighten checks for multi-dimensional input
>
> Previously an input array string that started with a single-element
> array dimension would then later accept a multi-dimensional segment.
>
> BACKWARD INCOMPATIBILITY
-select cardinality('{{{1}},{{2,3},{3,4}}}'::int[]);
+select cardinality('{{{1,9},{5,6}},{{2,3},{3,4}}}'::int[]);
cardinality
-------------
8
I was a little confused by this change: the query changed, but the
results did not. Was that a bugfix, or were the original semantics
intentional?
I didn't immediately find the thread where this was discussed, so if you
send a link that might answer my question.
Regards,
Jeff Davis
On Tue, Feb 4, 2014 at 11:21:24AM -0800, Jeff Davis wrote:
> On Sat, 2014-02-01 at 15:49 +0000, Bruce Momjian wrote:
> > arrays: tighten checks for multi-dimensional input
> >
> > Previously an input array string that started with a single-element
> > array dimension would then later accept a multi-dimensional segment.
> >
> > BACKWARD INCOMPATIBILITY
>
> -select cardinality('{{{1}},{{2,3},{3,4}}}'::int[]);
> +select cardinality('{{{1,9},{5,6}},{{2,3},{3,4}}}'::int[]);
> cardinality
> -------------
> 8
>
> I was a little confused by this change: the query changed, but the
> results did not. Was that a bugfix, or were the original semantics
> intentional?
Yes. the problem was that the new checking would throw an error for that
misformed array.
> I didn't immediately find the thread where this was discussed, so if you
> send a link that might answer my question.
Here is the spot in the thread where I replied:
http://www.postgresql.org/message-id/20140201011125.GG19957@momjian.us
--
Bruce Momjian <bruce@momjian.us> http://momjian.us
EnterpriseDB http://enterprisedb.com
+ Everyone has their own god. +