Re: ngettext revisited

уторак, 30. септембар 2003. 11:04:20 CEST — Christian Rose написа:


This seems to be troublesome because there are strings in XML files
using "%d".

AFAICS, intltool doesn't yet support plural forms with XML files.
Actually, this doesn't seem easily solveable at all, because there are
some strings which are loaded from XML files, and since they store the
messages at runtime (PO files are just used for initial editting, and
the strings from there are put into XML files at compile time), there's
no way to get the correct strings in running programs.

I don't know much about XML localization issues (yes, I've noticed that
some kind of 'xml:lang' attribute is used for translations, but this is
not enough for plural forms), so any hint would be helpful.

Btw, this could probably be solved more easily if GDM would actually
pass these strings to (n)gettext (depending on whether %d is used).
Still, there would need to be some more custom markup for (like
<N_and then the tag name>singular english form<n_>plural english form</
N_closing tag>

But this would still bring in many complications.


This one will has one tricky for fixing, like where there are several
"%d"'s in one string.

It would be nice to get some input from other translation teams if it's
ok to use concatenation, i.e. can they translate correctly the string:

 msgid "Uptime is %d days, %d hours, %d minutes"

if it's split into:

 msgid "Uptime is %d days"
 msgid "%d hours"
 msgid "%d minutes"

and later concatenated into
 "%s, %s, %s"

There's possibly also more instances where ngettext might be needed
but that haven't been reported yet. But this list should be a start.

Some of the Serbian translators didn't pay close attention when
translating, so I'll look where are some left. Most of the base desktop
and developer platform I translated or updated the translation, so I
usually noticed all the occurrences.

All the others modules which I didn't mention in here already contain
some sort of patch, so they can be fixed in no time; or there are no issues which would require input from other translators (like this concatenation issue, or strings from XML files).


[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]