Received: from sss.pgh.pa.us ([192.204.191.242])
	by postgresql.org (8.11.3/8.11.1) with ESMTP id f5JEdJa09773
	for <pgsql-sql@postgresql.org>; Tue, 19 Jun 2001 10:39:19 -0400 (EDT)
	(envelope-from tgl@sss.pgh.pa.us)
Received: from sss2.sss.pgh.pa.us (tgl@localhost [127.0.0.1])
	by sss.pgh.pa.us (8.11.4/8.11.4) with ESMTP id f5JEck322744;
	Tue, 19 Jun 2001 10:38:46 -0400 (EDT)
To: David BOURIAUD <david.bouriaud@ac-rouen.fr>
cc: "pgsql-sql@postgresql.org" <pgsql-sql@postgresql.org>
Subject: Re: About primary keys. 
In-reply-to: <3B2F4A8E.B6E07DBB@ac-rouen.fr> 
References: <20010619123625.F01A21A78@druid.net>
	<3B2F4A8E.B6E07DBB@ac-rouen.fr>
Comments: In-reply-to David BOURIAUD <david.bouriaud@ac-rouen.fr>
	message dated "Tue, 19 Jun 2001 14:50:22 +0200"
Date: Tue, 19 Jun 2001 10:38:45 -0400
Message-ID: <22741.992961525@sss.pgh.pa.us>
From: Tom Lane <tgl@sss.pgh.pa.us>
X-Archive-Number: 200106/291
X-Sequence-Number: 3554

David BOURIAUD <david.bouriaud@ac-rouen.fr> writes:
>> Caveat:  Does not work for complex primary keys.

> Thanks, but that is my problem actually !

Here's a rather brute-force approach:

select relname,
	(select attname from pg_attribute where
	 attrelid = c.oid and attnum = indkey[0]),
	(select attname from pg_attribute where
	 attrelid = c.oid and attnum = indkey[1]),
	(select attname from pg_attribute where
	 attrelid = c.oid and attnum = indkey[2]),
	(select attname from pg_attribute where
	 attrelid = c.oid and attnum = indkey[3])
from pg_class c, pg_index i
where c.oid = indrelid and indisprimary;

You can carry this out to however many key columns you want to
deal with.  The sub-selects will yield NULLs for the columns after
the last key column of a particular index, which is just what you
want here.

			regards, tom lane