Complete.Org: Mailing Lists: Archives: freeciv-i18n: August 2002:
[freeciv-i18n] Re: Freeciv po files
Home

[freeciv-i18n] Re: Freeciv po files

[Top] [All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index] [Thread Index]
To: Freeciv i18n ML <freeciv-i18n@xxxxxxxxxxx>
Subject: [freeciv-i18n] Re: Freeciv po files
From: Davide Pagnin <nightmare@xxxxxxxxxx>
Date: 31 Aug 2002 08:33:51 +0200

On Wed, 2002-08-28 at 11:21, Christian Knoke wrote:
> 
> On Wed, Aug 28, 2002 at 10:06:11AM +0200, Davide Pagnin wrote:
> > 
> > On Fri, 2002-08-23 at 22:35, Christian Knoke wrote: 
> > > 
> > > On Wed, Aug 21, 2002 at 11:59:53AM +0200, Davide Pagnin wrote:
> > > > 
> > > > Thus I suggest to modify po/Makefile.in.in and add the -c check to the
> > > > relevant line.
> > > 
> > Well, can some of the maintainer or other involved people, spend some on
> > this? 
> > 
> > There is a problem with actually take this step, many .po files will be
> > broken if this -c parameter (but this is because they're wrong!) and
> > thus we need to handle this condition in some way. 
> > 
> > Contacting all translator is one possible thing, but for poorly
> > maintained .po files, we need something else. 
> 
> Well, is this such a big problem? I found:
> 
> File: fr.po
> 2639 translated messages, 206 fuzzy translations, 79 untranslated messages.
> found 2 fatal errors
> 
> File: fi.po
> 2437 translated messages, 309 fuzzy translations, 179 untranslated messages.
> found 7 fatal errors
> 
> File: hu.po
> 2589 translated messages, 69 fuzzy translations, 267 untranslated messages.
> found 36 fatal errors

What version of gettext are you using?

I suspect that you're using old 0.10.X gettext, because the errors
reported by my 0.11.1 gettext are many more.

Egbert has written that, even better, in 0.11.5 version, also other
errors are reported by msgfmt -c --statistics, thus perhaps we have to
found and change the appropriate translation instruction.

The use of a recent gettext is IMHO a necessity.

There are also stricter rules for the header part of the .po file.

> 
> only three translations. Just mail the translators the relevant entries,
> and then remove them. Most likely the errors are "fuzzy" anyway.

After reading this, my mind awakened... 

Initially I've thought: yes, he is right! we can remove the relevant
entries, so the problem is solved...

I've then taken the task of removing the relevant entries, and after
some removal, I've noticed: "strange, no one of this entries are fuzzy,
what's a pity..."

Then my mind really awakened and... 

I've thought: what happens if a make those entries "fuzzy"?

As someone has already imagined, a fuzzy entry will give no errors, thus
we have an easy way of reaching the goal (introduction of -c parameter)
and to preserve work of previous translators, even for unmaintained .po
files! 

I see no drawback of patching the .po files adding some fuzzy entries
and I hope that the maintainers will agree with this!

I've prepared the patch file for the .po files that do not compile on my
1.11.1 gettext machine:

da.po
fi.po
fr.po
et.po
hu.po
pt.po
pt_BR.po
ro.po
ru.po
sv.po

In the attached archive you will find the patches, against Freeciv cvs
of 30 August at 17:00 pm.

I have added "fuzzy" marker for the offending translation and "proper"
header definitions (the most part of those file, have some default
entries untouched that needs to be modified, and lacks of a proper
plural form entry, in this case I've added the suggested definition I've
found in the info related to gettext).

I hope that those patches will be included soon and that the -c
parameter for msgfmt program will hopefully found his way to the cvs!

> 
> > Gaute, what is your opinion, on this? 
> > Is there a configure option viable? 
> 
> Christian
> 
> -- 
> Christian Knoke     * * *      http://www.enter.de/~c.knoke/
> * * * * * * * * *  Ceterum censeo Microsoft esse dividendum.
> 
> 

        Ciao, Davide


-- Binary/unsupported file stripped by Ecartis --
-- Type: application/x-bzip
-- File: full-po-patch.tar.bz2




[Prev in Thread] Current Thread [Next in Thread]