Re: Best way to scan on-disk bitmaps
Tom Lane <tgl@sss.pgh.pa.us>
From: Tom Lane <tgl@sss.pgh.pa.us>
To: Teodor Sigaev <teodor@sigaev.ru>
Cc: Greg Stark <gsstark@mit.edu>, Bruce Momjian <pgman@candle.pha.pa.us>, "Victor Y. Yegorov" <viy@mits.lv>, Postgres Hackers <pgsql-hackers@postgresql.org>
Date: 2005-06-13T23:20:21Z
Lists: pgsql-hackers
Teodor Sigaev <teodor@sigaev.ru> wrote: > ... So, if index is defined as 'using gist (a,b,c)' then, in > principle, GiST index can speed up queries like 'a=V1 and c=V2'. But > it will not usable for queries ( b=V3 and c=V2 ). By the way, instead > of '=' operation may be used other operations. Number of supported > operations by GiST is indefinite unlike, for example, btree which > supported only five: <, <=, =, =>, >. I have committed changes to the planner to arrange that a GiST indexscan must supply at least one restriction clause for the first index column, and can supply restriction clauses for any, all, or none of the remaining columns; the old left-to-right heuristic is gone. As far as I can tell, this doesn't require any changes to the GiST code, but please take another look if you aren't too sure about it. regards, tom lane