Horton

Re: pg_get_functiondef forgets about most GUC_LIST_INPUT GUCs

Michael Paquier <michael@paquier.xyz>

View thread View on postgr.es
From: Michael Paquier <michael@paquier.xyz>
To: Pavel Stehule <pavel.stehule@gmail.com>
Cc: Kyotaro HORIGUCHI <horiguchi.kyotaro@lab.ntt.co.jp>, Tom Lane <tgl@sss.pgh.pa.us>, Artur Zakirov <a.zakirov@postgrespro.ru>, Michael Paquier <michael.paquier@gmail.com>, PostgreSQL Hackers <pgsql-hackers@postgresql.org>
Date: 2018-03-16T12:15:54Z
Lists: pgsql-hackers 
On Fri, Mar 16, 2018 at 09:40:18AM +0100, Pavel Stehule wrote:
> 2018-03-16 9:34 GMT+01:00 Kyotaro HORIGUCHI <horiguchi.kyotaro@lab.ntt.co.jp
>> That's true, but doesn't the additional rule make it more
>> bothersome to maintain the list?
> 
> Hard to say. I have not opinion. But just when I see in this context
> variables like listen_address, shared_preload_libraries, then it looks
> messy.

Let's be clear.  I have listed all the variables in the patch to gather
more easily opinions, and because it is easier to review the whole stack
this way. I personally think that the only variables where the patch
makes sense are:
- DateStyle
- search_path
- plpgsql.extra_errors
- plpgsql.extra_warnings
- wal_consistency_checking
So I would be incline to drop the rest from the patch.  If there are
authors of popular extensions willing to get this support, let's update
the list once they argue about it and only if it makes sense.  However,
as far as I can see, there are no real candidates.  So let's keep the
list simple.
--
Michael

Commits

  1. Prevent extensions from creating custom GUCs that are GUC_LIST_QUOTE.

  2. Fix mishandling of quoted-list GUC values in pg_dump and ruleutils.c.