2009/1/26 Grzegorz Jaśkiewicz
<gryzman@gmail.com> Hey folks,
I have question really for all mighty developers, but don't want to
spam -hackers with it.
why :
select * from foo where X in (1,1,1,1,1,1,1,1) --- same values in search.
or select * from foo where (x,y) in ((1,2),(1,2),(1,2),(1,2),(1,2),(1,2),(1,2));
never gets optimized by planner, etc ?
I would guess that optimizing silly-written queries was always a low-priority task...
IMHO this is good topic for -hackers list.. and probably not so hard to implement :)
BTW, test on CVS HEAD:
CREATE TABLE atest(id integer primary key);
insert into atest select x from generate_series(1,100000) x(x);
ANALYZE atest;
EXPLAIN ANALYZE SELECT * FROM atest where id in (1,2,3,3,3,3,3,3,3,5,5,5,5,5,5,5,5,5,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1);
EXPLAIN ANALYZE SELECT * FROM atest where id in (1,2,3,5);
shows that second query is 2.5 times faster than the first ( 0.170 ms / 0.070 ms).
Is it just not worth optimizing from pg side? I am sure, it would make
sense to actually reorder these values, so that index/whatnot could
pick it up faster.
Just another one of those, 'why' (not) questions from my side.
thanks.
--
GJ
--
Sent via pgsql-general mailing list (pgsql-general@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-general