Cc: pgsql-admin@postgresql.org, ashishka@synechron.com
Content-Type: text/plain; charset="iso-8859-1"
Date: Thu, 05 Mar 2009 17:23:01 +0100
From: Jan-Peter.Seifert@gmx.de
In-Reply-To: <6382.1236199119@sss.pgh.pa.us>
Message-ID: <20090305162301.81330@gmx.net>
MIME-Version: 1.0
References: <20090302122020.137490@gmx.net> <49ABE46D.3000501@synechron.com>
	<20090304173815.4090@gmx.net> <6382.1236199119@sss.pgh.pa.us>
Subject: Re: getting 'full' names of functions?
To: Tom Lane <tgl@sss.pgh.pa.us>
Content-Transfer-Encoding: 8bit

Hello Tom,

> > I combined your suggestions into this query I'll be using for now:
> 
> > SELECT DISTINCT n.nspname || '.' || p.oid::regprocedure::text FROM
> 
> This is flat *wrong*, as you'll soon find if you are working with
> functions in more than one schema.  regprocedure already puts a
> schema qualification on the name if one is needed.

You are right. I was more concerned with getting the same number of functions as in pgAdmin III. There were two 'missing' because of pg_function_is_visible:

public.plpgsql_call_handler()
public.plpgsql_validator(oid)

They are obviously relicts from an earlier PostgreSQL-version. Could they cause problems if left be although the new versions are now in pg_catalog which obviously is being searched before all other schemata?

But why is regprocedure not just addding the schema to all of the functions then? 

Thank you very much,

Peter
-- 
Psssst! Schon vom neuen GMX MultiMessenger gehört? Der kann`s mit allen: http://www.gmx.net/de/go/multimessenger01