From: | Andrew Gierth <andrew(at)tao11(dot)riddles(dot)org(dot)uk> |
---|---|
To: | pgsql-hackers(at)postgresql(dot)org |
Subject: | Aggregate ORDER BY patch |
Date: | 2009-11-13 03:16:04 |
Message-ID: | 87tywzqi13.fsf@news-spur.riddles.org.uk |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-hackers |
Herewith a patch to implement agg(foo ORDER BY bar) with or without
DISTINCT, etc. No artificial restrictions are imposed on what
syntactical combinations are allowed. However, ORDER BY is not allowed
with aggregates used as window functions (as per the existing
restriction on DISTINCT).
Included are regression tests but not docs yet, those will follow
shortly in a separate patch (to save having to keep redoing the code
patch like last time).
Caveat: as discussed earlier, this patch changes the behaviour of
array_agg(DISTINCT x) when applied to NULL inputs. Formerly, the NULLs
were unconditionally skipped; now, they are treated just like DISTINCT
or GROUP BY normally do.
The previous restriction of agg(DISTINCT ...) to single-argument
aggregates is removed. However, there is still a separate code path
for aggregates that use DISTINCT or ORDER BY with only one input
column, for performance reasons.
If a non-default ordering operator is used in combination with
DISTINCT, then the notion of "equality" used for the DISTINCT
comparisons is the one that belongs to the ordering, rather than the
default.
--
Andrew.
Attachment | Content-Type | Size |
---|---|---|
aorder-20091113.patch.gz | application/octet-stream | 17.5 KB |
From | Date | Subject | |
---|---|---|---|
Next Message | Itagaki Takahiro | 2009-11-13 03:22:36 | Re: TRIGGER with WHEN clause |
Previous Message | Robert Haas | 2009-11-13 03:10:24 | Re: next CommitFest |