xgettext (was: more Fun With Encoding)

[Christian Stimming]

Am Sonntag, 9. Oktober 2005 03:46 schrieb Chris Shoemaker:
> > > xgettext appears to now support finding strings in Lisp files.  Would
> > > be worth checking to see if this works with the scheme variant of lisp.
> > >  I would expect it does.  If so we could dump the intl-scm directory
> > > altogether and use xgettext directly.
> >
> > I noticed that, too.  And that's what I was planning on looking into.
>
> Indeed, xgettext does seem to correctly parse our scheme files.
>
> From looking at xgettext's NEWS file, I gather that this has been
> supported since Version 0.14.2 - February 2005.  IIUC, the only people
> who need to run xgettext are the packagers, i.e. the developers.  So,
> is this version sufficiently old enough that people are in favor of
> dropping our "use perl to collect Scheme strings into C file so
> xgettext can find them"-trick?

As for my development system, which is suse9.3 so already quite recent (its 
successor suse10.0 was just released this week), the bundled xgettext is 
0.14.1. I would therefore prefer sticking with the old method for something 
like the upcoming 6 more months.

And except for this very weird encoding error, this "trick" for collecting the 
strings from non-C-files is working really well and is done for other things 
as well (glade, for example, although already automated by xgettext itself). 
It is a workaround, but IMHO not one of the worse ones but a rather elegant 
one.

Christian