Re: glade localization woes

[Thierry Vignaud <tvignaud mandrakesoft com>]

Jan Hudec <bulb ucw cz> writes:

> > > It seems that there is a bug in gettext. Perl has two types of
> > > strings -- octet streams and unicode. It's a mess,
> > > sometimes. Read the perlunicode manpage to find more about
> > > it. Now what really should be done is forcing gettext to mark
> > > the string as unicode one if it is one.
> > >
> > > > Is it anything I'm doing wrong, or is it a Perl Glade bug (which
> > > > is what I'm inclined to believe)?
> > >
> > > No. It's a gettext bug. It returns unicode string, but does not mark
> > > it as such. And all the glib-perl/gtk2-perl/gtk2-glade-perl stuff
> > > honors that marks.
> >
> > gettext hasn't to tag a perl string as utf8 since it's not aware of
> > perl internals and you cannot expect it from doing so.
>
> The binding is aware of the internals and should properly recode and
> mark the string. The C interface to gettext function returns just
> character arrays.

nice but there's no gettext binding (at least not in perl-Gtk2).

> > you need to call "some_module::bind_textdomain_codeset("mydomain",
> > 'UTF8');" in order to get utf8 strings
>
> That's a bug in the bindings. The binding of gettext SHOULD
> *convert* the string to utf-8 from whatever gettext returns and mark
> it as utf-8 string. After all, in other languages that only have
> utf-8 strings it has to do the same.

no, read the gettext API: the program has to tell gettext in which
encoding it expects strings, else gettext assume the program will use
the encoding specified by the current locale.

then, if needed, gettext will do the conversion between the encoding
used by the translator and the one expected by the program

> > with this functions defined in a xs as:
> >
> > > ==================================================>
> > > char *                                           >
> > > bind_textdomain_codeset(domainname, codeset)     >
> > >    char * domainname                             >
> > >    char * codeset                                >
> > > ==================================================>
>
> That's nice. But:
>     1) Undocumented

just look at gettext doc

>     2) Shouldn't be -- the bindings should set utf-8 behind the
> scenes.

a) about common.pm, it had to cover both gtk+-1 and gtk+-2 since until
   a few monthes ago, we still had a few gtk+1 app using these
   modules.

b) gtk+2 perl bindinb does not include any gettext support. it just
   expect you to pass it unicode strings.

   BUT gettext will give it strings encoded in the *locale encoding*
   (the value of nl_langinfo (CODESET)', which depends on the
   `LC_CTYPE')

   just RTFM !

> > and then you've to tag strings as utf8
> >
> > see N() implementation in common.pm from drakx installer and tools:
>
> The c:: package does not exist. That's pretty useless.
> The code is damn lot complicated.

the c package is availlable in our cvs at
http://cvs.mandrakesoft.com/cgi-bin/cvsweb.cgi/gi/perl-install/