On Wed, Jun 16, 2004 at 16:50:17 +0200, A. Pagaltzis wrote:
* Jan Hudec <bulb ucw cz> [2004-06-16 13:36]:On Wed, Jun 16, 2004 at 13:02:58 +0200, Thierry Vignaud wrote:last but not least, since Locale::Gettext has no knowledge of gtk2-perl and cannot and shouldn't know whether gtk+2 is used or not, there's no reason that this package should enforce utf8You are right here. Perl works just right with all string in locale encoding. Touching Locale::gettext is not a good idea.May I interrupt? If I understood correctly, Locale::gettext does not mark the string as UTF-8 for the perl internals, even if you request UTF-8 encoding from gettext, right? In that case, that's definitely a bug. I don't think it's a good idea to have Locale::gettext *force* UTF-8, but when the wrapper is told to return strings encoded that way, it should certainly set the corresponding flag on the string so things will Just Work in Perl.
Yes, I agree with that. Though it would have some tracking to do as to
what domain is used and what charset is selected for that domain. It's,
however, doable and not too hard. The idea is:
- Binding 'textdomain' should set own variable and not call textdomain
- Binding 'gettext' should use that variable and call our version of
dgettext
- Our version of dgettext should call bind_textdomain_codeset with NULL
to get the charset. If it is utf-8, or if it is NULL and
nl_langinfo(CODESET) returned utf-8, mark the string as utf-8.
It can even be done in perl.
-------------------------------------------------------------------------------
Jan 'Bulb' Hudec <bulb ucw cz>
Attachment:
signature.asc
Description: Digital signature