Date: Mon, 24 Sep 2012 10:31:04 -0400
From: Bruce Momjian <bruce@momjian.us>
To: Tom Lane <tgl@sss.pgh.pa.us>
Cc: Peter Eisentraut <peter_e@gmx.net>, Rural Hunter <ruralhunter@gmail.com>,
	PostgreSQL-development <pgsql-hackers@postgresql.org>
Subject: Re: [ADMIN] pg_upgrade from 9.1.3 to 9.2 failed
Message-ID: <20120924143104.GD21242@momjian.us>
References: <20120917210723.GA30394@momjian.us>
	<20120918232239.GA2014@momjian.us> <505C307E.20501@gmail.com>
	<1348403628.18490.1.camel@vanquo.pezone.net>
	<505F1A98.9090901@gmail.com>
	<1348440393.18490.2.camel@vanquo.pezone.net>
	<20120924125500.GA21242@momjian.us> <50605ABC.1000304@gmx.net>
	<20120924130820.GB21242@momjian.us>
	<14327.1348496025@sss.pgh.pa.us>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <14327.1348496025@sss.pgh.pa.us>
User-Agent: Mutt/1.5.20 (2009-06-14)

On Mon, Sep 24, 2012 at 10:13:45AM -0400, Tom Lane wrote:
> Bruce Momjian <bruce@momjian.us> writes:
> >>> I can confirm that pg_upgrade does case-insensitive comparisons of
> >>> encoding/locale names:
> 
> > Or we could just remove dashes from the name before comparisons.
> 
> That would merely move the breakage somewhere else.  I think you are
> already assuming far too much about the OS' interpretation of locale
> names by assuming they are case-insensitive.  Assuming that dashes
> aren't significant seems 100% wrong.
> 
> FWIW, what I found out last time I touched this code is that on many
> systems setlocale doesn't bother to return a canonicalized spelling;
> it just gives back the string you gave it.  It might be worth doing
> what Peter suggests, just to be consistent with what we are doing
> elsewhere, but I'm not sure how much it will help.

This comment in initdb.c doesn't sound hopeful:

 * If successful, and canonname isn't NULL, a malloc'd copy of the locale's
 * canonical name is stored there.  This is especially useful for figuring out
 * what locale name "" means (ie, the environment value).  (Actually,
 * it seems that on most implementations that's the only thing it's good for;
 * we could wish that setlocale gave back a canonically spelled version of
 * the locale name, but typically it doesn't.)

-- 
  Bruce Momjian  <bruce@momjian.us>        http://momjian.us
  EnterpriseDB                             http://enterprisedb.com

  + It's impossible for everything to be true. +