Re: new language process (Quechua): 1st step adding a locale
On Thu, Feb 23, 2006 at 03:53:05PM +0200, Eddy Petrişor wrote:
> I have seen in your template the following section:
>
> # 4. Charsets
> # List of supported charsets. The first one is the default for
> # this locale. New locales should not need any other chanrset
> # than UTF-8.
> # Example:
> # charset "UTF-8"
> charset ""
>
> Is it possible to declare more than one charset? I was thinking of
> placing the other frequently used encodings for Romanian iso-8859-2
> and iso-8859-16 to the list. How should add that?
I will use this field to declare locales in /usr/share/i18n/SUPPORTED.
There is no syntax, I will edit this file by hand.
> The only thing I have seen related to encodings is:
> % Charset: UTF-8
>
> Should there be a line like the one below in the locale definition file?
> charset "UTF-8 ISO-8859-16 ISO-8859-2"
Hmm, maybe there is some misunderstanding here. The file I sent is a
human readable/editable file which can be converted into a locale
definition file. For now I will do this process by hand, so any
separator is fine by me. In /usr/share/i18n/locales/ro_RO you may add
% Charset: UTF-8 ISO-8859-16 ISO-8859-2
but I do not know if upstream will notice it, you should instead send
a patch against localedata/SUPPORTED.
Historically, locales had only one charset, eg. most western European
countries were defined with ISO-8859-1. AFAICT the
% Charset:
line was introduced to let glibc maintainers be aware of this encoding,
so that they can declare locales in the SUPPORTED file.
Denis
Reply to: