Re: Question about an old i18n bug in MC

[Pavel Roskin <proski gnu org>]

Hi, Leonard!

Quoting Leonard den Ottolander <leonard den ottolander nl>:

> Hi Rostislav,
>
> On Sun, 2005-07-10 at 00:12, Rostislav Krasny wrote:
> > It cannot display the help and hints properly in any russian localized
> > system that is using any cyrillic codepage, except a KOI8-R.
>
> Problem probably is that there is only one .po file per language.

That's not a problem.  The text from .po is shown correctly in all encodings
that are standard and properly specified.  The complaint is about help and
hints, which are shown without any translation.  Fixing that would require mc
to use charset translation functions, either from gettext or from libc.  In
either case, it would require changes in the configure script and lots of
testing.

> You
> might try to convert po/ru.po to the charset you'd like to use using
> iconv and put the translated page back in your build tree. You might
> also have to update the meta data in the header for this to work
> correctly. Also I wouldn't know if it is possible to use multiple
> translations in one tree.

It's possible to use multiple translations for .po (see gettext manual), but
it's not what would help.  We could add similar mechanism to find hints and
help files matching the exact locale, but I think it should be reserved for
non-trivial cases, like dialects or different scripts (e.g. in Serbian). 
Encodings in Russian can be trivially recoded to each other.

--
Regards,
Pavel Roskin