Re: First version of multi-key index support for GiST - Mailing list pgsql-hackers

From Tom Lane
Subject Re: First version of multi-key index support for GiST
Date
Msg-id 12959.991334124@sss.pgh.pa.us
Whole thread Raw
In response to First version of multi-key index support for GiST  (Oleg Bartunov <oleg@sai.msu.su>)
Responses Re: First version of multi-key index support for GiST
List pgsql-hackers
Oleg Bartunov <oleg@sai.msu.su> writes:
> We have implemented multi-key index support for GiST. Patch is available
> from  http://www.sai.msu.su/~megera/postgres/gist/code/7.1.2/patch_multikeygist.7.1.2.gz

I have committed these changes, along with your leak patch of 5/30.

> 1. initdb is required. But, it's possible just to execute update
>    update pg_am set amstrategies = 12 where amname = 'gist';

No initdb is needed --- I fixed the code instead ;-)

> 2. You have to recompile all gist_*_ops functions

I bit the bullet and fixed all the places that were using "char*" where
they should have been using "Datum".  This doesn't completely free GIST
from datatype assumptions: it still assumes that all datatypes it deals
with will be pass-by-reference.  But it's a step forward.  This means
not only a recompile but code changes for any user-supplied GIST ops.
I applied the appropriate changes to everything that's in contrib
(including your new RTREE emulation code).
        regards, tom lane


pgsql-hackers by date:

Previous
From: Luis Magaña
Date:
Subject: pg_dump & pg_dumpall problem.
Next
From: Tom Lane
Date:
Subject: Re: R-Tree implementation using GiST (compatible with multi-key GiST)