X-Original-To: pgsql-sql@postgresql.org
Received: from localhost (unknown [64.117.224.193])
	by developer.postgresql.org (Postfix) with ESMTP id 3EC65924E3E
	for <pgsql-sql@postgresql.org>; Tue, 20 May 2003 12:47:27 -0400 (EDT)
Received: from developer.postgresql.org ([64.117.224.193])
	by localhost (developer.postgresql.org [64.117.224.193:10024])
	(amavisd-new) with ESMTP id 19542-05 for <pgsql-sql@postgresql.org>;
	Tue, 20 May 2003 12:47:22 -0400 (EDT)
Received: from megazone.bigpanda.com (megazone.bigpanda.com [63.150.15.178])
	by developer.postgresql.org (Postfix) with ESMTP id 99304925044
	for <pgsql-sql@postgresql.org>; Tue, 20 May 2003 12:46:38 -0400 (EDT)
Received: by megazone.bigpanda.com (Postfix, from userid 1001)
	id 730A3D622; Tue, 20 May 2003 09:46:37 -0700 (PDT)
Received: from localhost (localhost [127.0.0.1])
	by megazone.bigpanda.com (Postfix) with ESMTP
	id 68C475C0C; Tue, 20 May 2003 09:46:37 -0700 (PDT)
Date: Tue, 20 May 2003 09:46:37 -0700 (PDT)
From: Stephan Szabo <sszabo@megazone23.bigpanda.com>
To: ww zz <ts_999@yahoo.com>
Cc: <pgsql-sql@postgresql.org>
Subject: Re: how to do this query?
In-Reply-To: <20030520160011.81254.qmail@web41803.mail.yahoo.com>
Message-ID: <20030520094321.T68048-100000@megazone23.bigpanda.com>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
X-Virus-Scanned: by amavisd-new
X-Archive-Number: 200305/391
X-Sequence-Number: 13556

On Tue, 20 May 2003, ww zz wrote:

> New to sql and may be a simple question for you:
>
> I have four tables
>
> sample:   sid, sname
> make:     sid, mdate
> measure:  sid, mdate, rid
> result:   rid, resultdata
>
> for a gaven sample it always has a sid and sname, but
> it may or may not have any entries in the other
> tables, the sample could be measured more than once
> and each measure give a result row.
> question:
> how do I do a query that given a sid, it will retrieve
> all the information available in the four tables, if
> there are available. If, for example, all tables have
> some entry for a gave sample it will retreve all the
> information, if the "measure" and "result" are empty
> for that sample it will give only the 'sample' and
> 'make' information....

Well, you're not going to get a limited column set, but you can
get NULLs in the non-applicable table columns by using outer joins.

Maybe something like something like
 select * from
  sample left outer join make using (sid)
   left outer join measure using (sid)
   left outer join result using (rid)
 where sid=<value>;