Re: stand-alone composite types patch (was [HACKERS] Proposal: stand-alone composite types) - Mailing list pgsql-patches

From Tom Lane
Subject Re: stand-alone composite types patch (was [HACKERS] Proposal: stand-alone composite types)
Date
Msg-id 29946.1028784103@sss.pgh.pa.us
Whole thread Raw
In response to Re: stand-alone composite types patch (was [HACKERS] Proposal:  (Joe Conway <mail@joeconway.com>)
List pgsql-patches
Joe Conway <mail@joeconway.com> writes:
> Tom Lane wrote:
>> Also, I'm not following the point of the separation between
>> DefineCompositeType and DefineCompositeTypeRelation; nor do I see a need
>> for a CommandCounterIncrement call in there.

> Well the next thing I was going to work on after this was an implicitly
> created composite type when creating a function. I thought maybe the
> CommandCounterIncrement would be needed so that the type could be
> created and then immediately used by the function.

Hm.  Maybe --- it would depend on whether the function-creating code
actually tried to look at the type definition, as opposed to just using
its OID.  (You'll probably want DefineCompositeType to return the type
OID, btw.)  In any case, I'd be inclined to put the CCI call in the
caller not the callee, so it's only done when actually needed.  It's
surely not needed for a standalone CREATE TYPE command.

            regards, tom lane

pgsql-patches by date:

Previous
From: Joe Conway
Date:
Subject: Re: stand-alone composite types patch (was [HACKERS] Proposal:
Next
From: Alvaro Herrera
Date:
Subject: Re: CLUSTER patch and regression test