Date: Thu, 24 Jul 2003 12:03:51 -0700
From: elein <elein@varlena.com>
To: Tom Lane <tgl@sss.pgh.pa.us>
Cc: elein <elein@varlena.com>,
	Robert Treat <xzilla@users.sourceforge.net>, pgsql-sql@postgresql.org
Subject: Re: obtuse plpgsql function needs
Message-ID: <20030724120351.D26401@cookie>
References: <1058913112.22259.271.camel@camel> <20030722163335.A18881@cookie>
	<1058965609.22259.292.camel@camel>
	<20030723111201.D20411@cookie> <15642.1058987750@sss.pgh.pa.us>
	<20030723122723.C26401@cookie> <15510.1059023238@sss.pgh.pa.us>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <15510.1059023238@sss.pgh.pa.us>
User-Agent: Mutt/1.3.22.1i

Bruce--
Something for the todo list. This would be extremely
handy.  At minimum C functions should be able to 
ask the type of thing that was actually passed in and get
a legitimate answer even if the type were a 
rowtype.

This will also lead to the need for unnamed rowtypes,
sooner or later.

I know, I know, send a patch.

--elein

On Thu, Jul 24, 2003 at 01:07:18AM -0400, Tom Lane wrote:
> elein <elein@varlena.com> writes:
> > So, other than C, plperl or pltcl is the way to go.
> > As long as they can input generic composite types
> > (I wasn't sure of that, but I should have known), 
> 
> Come to think of it, that is a problem: we don't have any way to declare
> a function as taking "any tuple type".  So even though pltcl or plperl
> functions could be written to work with such input, we can't declare them.
> This is a problem even for C functions.  You could declare a C function
> as taking "any", but then you can't even check that what you got was a
> tuple ...
> 
> Something to work on for 7.5, I suppose.
> 
> 			regards, tom lane
> 
> ---------------------------(end of broadcast)---------------------------
> TIP 6: Have you searched our list archives?
> 
>                http://archives.postgresql.org
>