This is the mail archive of the
libc-alpha@sources.redhat.com
mailing list for the glibc project.
Re: es_AR locale patch
- From: Keld Jørn Simonsen <keld at dkuug dot dk>
- To: Jeff Bailey <jbailey at nisa dot net>
- Cc: libc-alpha at sources dot redhat dot com
- Date: Tue, 17 Sep 2002 13:20:08 +0200
- Subject: Re: es_AR locale patch
- References: <20020916023451.GH11743@nisa.net>
On Sun, Sep 15, 2002 at 07:34:51PM -0700, Jeff Bailey wrote:
> Please accept this patch to correctly format numbers for the es_AR
> locale. For verification we looked at:
>
> A major Argentinian newspaper:
> http://www.clarin.com/diario/hoy/s-04201.htm
> A ministry of economy chart:
> http://www.mecon.gov.ar/boletin/2dotrim01/capitulos/capitulo62.htm
> And finally, confirmed that Microsoft Windows 2000 shows the same
> behaviour.
>
> 2002-09-15 Jeff Bailey <jbailey@debian.org>
>
> * localedata/locales/es_AR: Fix es_AR number formatting.
> Thanks to Nicolas Lichtmaier for the patch, and Jordi
> Mallach for verifying it.
>
> --- localedata/locales/es_AR~ 2000/10/26 23:22:55 1.10
> +++ localedata/locales/es_AR 2001/10/07 20:56:18
> @@ -80,8 +80,8 @@
>
> LC_NUMERIC
> decimal_point "<U002C>"
> -thousands_sep ""
> -grouping 0;0
> +thousands_sep "<U002E>"
> +grouping 3;3
> END LC_NUMERIC
>
> LC_TIME
The problem this tries to remedy is, whether to use the "." as
the thousands separator or not, in normal numbers. And I think
that is a question of convenience. In money amounts I think it
is more clear that it is convenient to use thousand separators,
but in normal numbers I think it is very customary just to leave out
the thousands separator. Consider
123456,43
123.456,43
The former format probably is easier to scan in again, if the
output is to be used again, which is often the case with normal
numbers. I am not sure whether the glibc string scanning routines
or other libc packages take account of decimal_point and/or
thousands_sep, could somebody enlighten me? And also I think
these scanning functions, if able to be locale dependent, are very little
used... My recommendation would be to keep the old dot-less
notation for the regular numbers, but it should of cause be up to
the Argentines.
BTW, I would prefer if we use character literals in the locales
like <.> and <,> - it makes it much easier to read in many cases.
reading 10646 hex codes is like reading binary programs in hex to
me (it can be done but is very cumbersome).
Kind regards
keld