Re: [Texi2html-bug] Problems with texi2html and Japanese navigation bars

[Patrice Dumas]

On Tue, Feb 03, 2009 at 01:09:12AM +0100, John Mandereau wrote:
> Karl Berry a écrit :
>>     for package maintainers who use the TP is to convert the PO files 
>> they get     to UTF-8 charset when committing to the sources.
>>
>> The goal of the TP is for package maintainers to *copy* the
>> translation files, unaltered.  Doing anything else is asking for trouble.
>
> IMHO you should ask translators to provide translations exclusively 
> encoded in UTF-8, as it makes converting the encoding when writing 
> Texinfo output in texi2html a little easier: there would be only 
> conversions from UTF-8 to some other encoding to do.

I don't think it really matters. Indeed perl converts everything 
to an internal encoding anyway, so I guess that a gettext-like 
module in perl would accept any kind of encoding.

Currently, in texi2html it is possible to have the out encoding different 
from the input texi file encoding, and both can be any encoding
perl knows about, and it is perl that does the decoding and recoding
once the encoding name has been associated with a filehandle with 
binmode.

--
Pat