Content-Type: text/plain;
  charset="iso-8859-1"
From: Richard Huxton <dev@archonet.com>
Organization: Archonet Ltd
To: sammynash@uboot.com (Sammy), pgsql-sql@postgresql.org
Subject: Re: Chain/Thread Problem
Date: Fri, 16 May 2003 18:22:37 +0100
User-Agent: KMail/1.4.3
References: <6729af8d.0305160646.d9ab2d2@posting.google.com>
In-Reply-To: <6729af8d.0305160646.d9ab2d2@posting.google.com>
MIME-Version: 1.0
Content-Transfer-Encoding: quoted-printable
Message-Id: <200305161822.37391.dev@archonet.com>

On Friday 16 May 2003 3:46 pm, Sammy wrote:
> Hi,
>
> I have a SQL problem for you all. What SQL qurey would I have to make
> on the following table to extract the chain (thread) of emails in a
> conversation.
>
>
>          id|emailid|referenceid
>          -----------------------
>           1|  e125 |   Null             (Start of chain)
>           2|  r1fe |   e125             (2nd letter in chain)
>           ...
>           ...
>           n|  e4ju |   r1fe             (3rd in chain)

Search the archives for "connect by", "hierarchy", "nested" and "celko" - t=
his=20
is a limitation of SQL. In your particular case, I'd add a SERIAL "thread_i=
d"=20
column which is autogenerated for the first msg in a thread and set to the=
=20
first message for all others.

You might want to check the contrib/ directory, I've got a feeling there's =
an=20
example table-function to do something like this.
--=20
  Richard Huxton