Message-ID: <019ee93d8e7052b3a98fb46165e6c96fe312f338.camel@j-davis.com>
Subject: Re: Order changes in PG16 since ICU introduction
From: Jeff Davis <pgsql@j-davis.com>
To: Daniel Verite <daniel@manitou-mail.org>
Cc: Matthias van de Meent <boekewurm+postgres@gmail.com>, Andrew Gierth
 <andrew@tao11.riddles.org.uk>, Peter Eisentraut
 <peter.eisentraut@enterprisedb.com>, Sandro Santilli <strk@kbt.io>, Tom
 Lane <tgl@sss.pgh.pa.us>, Regina Obe <lr@pcorp.us>, 
 pgsql-hackers@lists.postgresql.org
Date: Fri, 26 May 2023 11:47:02 -0700
In-Reply-To: <a6204a46-c077-451b-8f9d-8965d95bb57c@manitou-mail.org>
References: <a6204a46-c077-451b-8f9d-8965d95bb57c@manitou-mail.org>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
User-Agent: Evolution 3.44.4-0ubuntu1 
MIME-Version: 1.0
Archived-At: <https://www.postgresql.org/message-id/019ee93d8e7052b3a98fb46165e6c96fe312f338.camel%40j-davis.com>
Precedence: bulk

On Fri, 2023-05-26 at 18:24 +0200, Daniel Verite wrote:
> To me initdb is wrong when coming up with the less precise "fr"
> instead of "fr-FR".
>=20
> I suggest the attached patch to call uloc_getDefault() instead of the
> current code that somehow leaves out the country/region component.

Thank you. I experimented with several ICU versions and different
environmental settings, and it does seem better at preserving the name
that comes from the environment.

There is a warning in the docs: "Do not use unless you know what you
are doing."[1] I don't see a reason for the warning or any major risk
from us using it. Perhaps it's because the result is affected by either
the environment or the last uloc_setDefault() call. We don't use
uloc_setDefault(), and we only call uloc_getDefault() once, so I don't
see a risk here.

The fix seems simple enough. Committed.

Regards,
	Jeff Davis

[1]
https://unicode-org.github.io/icu-docs/apidoc/dev/icu4c/uloc_8h.html#a4efa1=
6db7351e62293f8ef0c37aac8d2