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

Re: various fixes for stattrans.pl



Denis Barbier <barbier@linuxfr.org> (20/04/2005):
> >  	    $version{"$lang/$file"} = $altcvs->revision($f);
> >  	    if (!$transcheck->revision()) {
> > -		$original{"english/$file"} = $lang;
> > -		$transversion{"english/$file"} ||= "1.1";
> > +	        $transcheckenglish = Webwml::TransCheck->new("english/$file");
> > +		if (!$transcheckenglish->revision() and (-e "english/$file")) {
> > +		    $transversion{"$lang/$file"} = "1.1";
> > +		    $original{"$lang/$file"} = "english";
> > +		} else {
> > +		    $original{"english/$file"} = $lang;
> > +		    $transversion{"english/$file"} ||= "1.1";
> > +		}
> 
> English files are processed first,

I think it is not true.

> so there is IMO no need to introduce $transcheckenglish, you can
> instead try (not tested)
>   if (-e "english/$file" && !defined($transversion{"english/$file"})) {
>     $transversion{"$lang/$file"} = "1.1";
>     $original{"$lang/$file"} = "english";
>   } else {
>     $original{"english/$file"} = $lang;
>     $transversion{"english/$file"} ||= "1.1";
>   }

If this was true, the old version was enough. But I have already seen some
files without translation-check header beeing considered as the original
instead of English. I've checked with the -v switch, which confirm what
I'm saying.

> With both versions (yours and mine), $original{"english/$file"} is set if
> english/$file does not exist, I do not know whether this causes trouble.

No, no problems at all. It was already the case in the previous version.

Thanks for your comments.

Regards,

-- 
Thomas Huriaux

Attachment: signature.asc
Description: Digital signature


Reply to: