Date: Mon, 5 Sep 2005 11:13:31 -0400
From: Alvaro Herrera <alvherre@alvh.no-ip.org>
To: Peter Eisentraut <peter_e@gmx.net>
Cc: Aldor <an@mediaroot.de>, pgsql-sql@postgresql.org
Subject: Re: POSIX Regular Expression question
Message-ID: <20050905151331.GA19192@surnet.cl>
Mail-Followup-To: Peter Eisentraut <peter_e@gmx.net>,
	Aldor <an@mediaroot.de>, pgsql-sql@postgresql.org
References: <431C4EB2.80401@mediaroot.de> <200509051619.29210.peter_e@gmx.net>
Mime-Version: 1.0
Content-Type: text/plain; charset=iso-8859-1
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
In-Reply-To: <200509051619.29210.peter_e@gmx.net>
User-Agent: Mutt/1.5.10i

On Mon, Sep 05, 2005 at 04:19:28PM +0200, Peter Eisentraut wrote:
> Am Montag, 5. September 2005 15:57 schrieb Aldor:
> > I want to get out a string only with characters A-Za-z.
> > Any idea how to do this in Postgres with POSIX Regex?
> 
> Presumably,
> 
>     colname ~ '^[A-Za-z]*$'
> 
> If you want to be independent of locale issues, then you'd have to enumerate 
> all the letters instead of using a range specification.

Note that [:alpha:] and such character classes are defined by POSIX to
be locale independent:

alvherre=# select 'á' ~ '[a-z]';
 ?column? 
----------
 f
(1 fila)

alvherre=# select 'á' ~ '[[:alpha:]]';
 ?column? 
----------
 t
(1 fila)


-- 
Alvaro Herrera -- Valdivia, Chile         Architect, www.EnterpriseDB.com
"En las profundidades de nuestro inconsciente hay una obsesiva necesidad
de un universo lógico y coherente. Pero el universo real se halla siempre
un paso más allá de la lógica" (Irulan)