From owner-cvs-lib  Tue Aug 13 07:08:05 1996
Return-Path: owner-cvs-lib
Received: (from root@localhost)
          by freefall.freebsd.org (8.7.5/8.7.3) id HAA14968
          for cvs-lib-outgoing; Tue, 13 Aug 1996 07:08:05 -0700 (PDT)
Received: from sovcom.kiae.su (sovcom.kiae.su [193.125.152.1])
          by freefall.freebsd.org (8.7.5/8.7.3) with SMTP id HAA14917;
          Tue, 13 Aug 1996 07:07:53 -0700 (PDT)
Received: by sovcom.kiae.su id AA06023
  (5.65.kiae-1 ); Tue, 13 Aug 1996 16:58:48 +0300
Received: by sovcom.KIAE.su (UUMAIL/2.0); Tue, 13 Aug 96 16:58:48 +0300
Received: (from ache@localhost) by nagual.ru (8.7.5/8.7.3) id RAA01737; Tue, 13 Aug 1996 17:54:01 +0400 (MSD)
Message-Id: <199608131354.RAA01737@nagual.ru>
Subject: Re: cvs commit:  src/lib/libc/locale collcmp.c
In-Reply-To: <199608131101.VAA10191@godzilla.zeta.org.au> from "Bruce Evans" at "Aug 13, 96 09:01:20 pm"
To: bde@zeta.org.au (Bruce Evans)
Date: Tue, 13 Aug 1996 17:54:00 +0400 (MSD)
Cc: CVS-committers@freefall.freebsd.org, cvs-all@freefall.freebsd.org,
        cvs-lib@freefall.freebsd.org
From: =?KOI8-R?Q?=E1=CE=C4=D2=C5=CA_=FE=C5=D2=CE=CF=D7?= (Andrey A. Chernov) <ache@nagual.ru>
Organization: self
X-Class: Fast
X-Mailer: ELM [version 2.4ME+ PL24 (25)]
Mime-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Sender: owner-cvs-lib@FreeBSD.ORG
X-Loop: FreeBSD.org
Precedence: bulk

> >  Modified:    lib/libc/locale  collcmp.c
> >  Log:
> >  There is so many places where range comparation (using collate)
> >  needed (much more than I think initially), so I forced to add
> >  new user-visible non-standard function to libc.
> 
> This is probably OK for applications, but please try not to introduce
> more nonstandard namespace-polluting names that are referenced from
> separate libraries (collate_range_cmp is referenced from libgnuregex).

I don't plan to introduce more, one function is enough.
It needed for _all_ cases where [a-z]-type national letters ranges
used, including gnuregex, of course. The reason is simple:
letters in the top half of code table usually not sorted
alphabetically. Using exact code table layout ranges make
additional portablity troubles, i.e. I can covert shell script
which use [<Russian A>-<Russian YA>] range to another
Russian code table easily, but all non-alphabetical ranges
(exact layout ranges f.e.)
needs additional porter attention.
POSIX 1003.2 also says that collate can be used in [a-z]-type ranges.

-- 
Andrey A. Chernov
<ache@nagual.ru>
http://www.nagual.ru/~ache/