Re: Remaining dependency on setlocale() - Mailing list pgsql-hackers

From Thomas Munro
Subject Re: Remaining dependency on setlocale()
Date
Msg-id CA+hUKG+BHraKYVVOzaQ+-5c5pdYZJgo2GiVx3-daUj92RrPK5w@mail.gmail.com
Whole thread Raw
In response to Re: Remaining dependency on setlocale()  (Peter Eisentraut <peter@eisentraut.org>)
Responses Re: Remaining dependency on setlocale()
List pgsql-hackers
On Fri, Aug 16, 2024 at 1:25 AM Peter Eisentraut <peter@eisentraut.org> wrote:
> On 15.08.24 00:43, Thomas Munro wrote:
> > "" is a problem however... the special value for "native environment"
> > is returned as a real locale name, which we probably still need in
> > places.  We could change that to newlocale("") + query instead, but
>
> Where do we need that in the server?

Hmm.  Yeah, right, the only way I've found so far to even reach that
code and that captures that result is:

create database db2 locale = '';

Thats puts 'en_NZ.UTF-8' or whatever in pg_database.  In contrast,
create collation will accept '' but just store it verbatim, and the
GUCs for changing time, monetary, numeric accept it too and keep it
verbatim.  We could simply ban '' in all user commands.  I doubt
they're documented as acceptable values, once you get past initdb and
have a running system.  Looking into that...

> It should just be initdb doing that and then initializing the server
> with concrete values based on that.

Right.

> I guess technically some of these GUC settings default to the
> environment?  But I think we could consider getting rid of that.

Yeah.



pgsql-hackers by date:

Previous
From: Nathan Bossart
Date:
Subject: Re: pgsql: Introduce hash_search_with_hash_value() function
Next
From: Peter Geoghegan
Date:
Subject: Re: Showing primitive index scan count in EXPLAIN ANALYZE (for skip scan and SAOP scans)