From: Tom Lane <tgl@sss.pgh.pa.us>
To: Andres Freund <andres@anarazel.de>
cc: Peter Eisentraut <peter.eisentraut@2ndquadrant.com>,
        pgsql-hackers <pgsql-hackers@postgresql.org>
Subject: Re: SQL-standard function body
In-reply-to: <20200630192606.brfidtzwjrftcum2@alap3.anarazel.de>
References: <1c11f1eb-f00c-43b7-799d-2d44132c02d7@2ndquadrant.com>
 <20200630192606.brfidtzwjrftcum2@alap3.anarazel.de>
Comments: In-reply-to Andres Freund <andres@anarazel.de>
	message dated "Tue, 30 Jun 2020 12:26:06 -0700"
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-ID: <496793.1593546216.1@sss.pgh.pa.us>
Date: Tue, 30 Jun 2020 15:43:36 -0400
Message-ID: <496794.1593546216@sss.pgh.pa.us>
Precedence: bulk

Andres Freund <andres@anarazel.de> writes:
> On 2020-06-30 19:49:04 +0200, Peter Eisentraut wrote:
>> The function body is parsed at function definition time and stored as
>> expression nodes in probin.  So at run time, no further parsing is
>> required.

> Isn't a consequence of that that we'd get a lot more errors if any DDL
> is done to tables involved in the query? In contrast to other languages
> we'd not be able to handle column type changes etc, right?

I suppose it'd act like column references in a view, ie the dependency
mechanisms would forbid you from changing/dropping any column mentioned
in one of these functions.

			regards, tom lane