RE: Using BOOL in indexes - Mailing list pgsql-hackers

From Hiroshi Inoue
Subject RE: Using BOOL in indexes
Date
Msg-id 000201bfcaa2$e631cf80$2801007e@tpf.co.jp
Whole thread Raw
In response to Using BOOL in indexes  (Andrew McMillan <Andrew@catalyst.net.nz>)
List pgsql-hackers
> -----Original Message-----
> From: pgsql-hackers-owner@hub.org [mailto:pgsql-hackers-owner@hub.org]On
> Behalf Of Andrew McMillan
> 
> Hi,
> 
> I'm trying to convert an application from MS SQL / ASP / IIS to
> PostgreSQL / PHP / Apache.  I am having trouble getting efficient
> queries on one of my main tables, which tends to have some fairly large
> records in it.  Currently there are around 20000 records, and it looks
> like they average around 500 bytes from the VACUUM ANALYZE statistics
> below.
> 
> I don't really want any query on this table to return more than about 20
> records, so it seems to me that indexed access should be the answer, but
> I am having some problems with indexes containing BOOLEAN types.
> 
> I can't see any reason why BOOL shouldn't work in an index, and in other
> systems I have commonly used them as the first component of an index,
> which is what I want to do here.
>
> Also, I can't see why the estimator should see a difference between
> "WHERE head1" and "WHERE head1=TRUE".
> 
> 
> newsroom=# explain SELECT DISTINCT story.story_id, written, released,
> title, precis, author, head1 FROM story WHERE head1 ORDER BY written

Please add head1 to ORDER BY clause i.e. ORDER BY head1,written.

Regards.

Hiroshi Inoue
Inoue@tpf.co.jp



pgsql-hackers by date:

Previous
From: "Mitch Vincent"
Date:
Subject: Re: Full text indexing preformance! (long)
Next
From: Thomas Lockhart
Date:
Subject: Re: Re: [SQL] aliases break my query