Re: [PATCHes] Add basic multibyte charset handling to makeinfo

[Karl Berry]

    Sorry, I don't think I understand.  Can you provide a few examples of
    such non-portability?

Miloslav can correct me, but my impression is that there's no way to
look up a name like "ISO-8859-1", because the system operates on the
basis of locales, not encodings.

    all we need to do is query them for a locale
    that supports a given encoding.

That would be nice, but I do not know if it's possible.

    Finally, we could even redesign @documentencoding to require a valid
    locale name, not just an encoding name, if that would make the
    difference.

I don't think that would be so easy.  The reason @documentencoding takes
names like "ISO-8859-1" now is that those names work (as is) in the HTML
output, and (I believe) with Emacs.

Perhaps the right place to add locale information would be in
@documentlanguage, which right now is just a two-letter abbreviation.
There have been previous discussions about making that be a locale
instead (and do some mapping for backward compatibility), to make the
document/program translation work better (as I recall?), and/or to help
with Oleg's project of getting better font/language support in
texinfo.tex.  Unfortunately my brain is mush on the details.

In any case, whether that helps with the encoding issue, though, I do
not know.

    you could simply switch locales dynamically, 

If it's possible to figure out the right locale from the encoding.
That's the basic question, I guess.

Best,
Karl