Re: XFree86, UTF-8 and gtk+

Alexander Voropay writes:

>  Unfortunately, the current "POSIX locale" concept is not multilingual.
>  ...
> I can't "stack-up" a set of required locales (for Russian, Greek and French):
>  setlocale(LC_CTYPE,"ru_RU");
>  setlocale(LC_CTYPE,"el_GR");
>  setlocale(LC_CTYPE,"fr_FR");

It's really the encodings which are implicitly designated by these
locales (KOI8-R, ISO-8859-7, ISO-8859-1) which are not
multilingual. When you are in the ru_RU.UTF-8 locale, you can display
multilingual text using xterm-patch#8 and (to a more limited extent)
XFree86 4.0.1's XmbDrawString function.

So, it's not the fault of the POSIX locale concept.

> Otherwise, with UNICODE POSIX loceles (like en_US.UTF-8)
> I don't need LC_CTYPE and LC_COLLATE POSIX categories at all.

You need LC_CTYPE - otherwise the iswgraph etc. functions will not
give you correct information about the characters. And you need
LC_COLLATE - because collation order in de_DE.UTF-8 is different from
the one in es_ES.UTF-8.

> (most of characters is isalphanum() and due to Generic UNICODE
> collation : UTR#10)

UTF#10 allows for "tailoring". Here the LC_COLLATE locale facet comes
in.

Bruno
-
Linux-UTF8:   i18n of Linux on all levels
Archive:      http://mail.nl.linux.org/lists/