wide character support in glibc?

[Bruno Haible]

> I got the following message.  Would someone on this list be kind
> enough to reply?  Or else, where should I ideally send such requests?

I think the comp.unix.programmer newsgroup is the appropriate forum in
general, and linux-utf8 is the right mailing list for Linux specific
questions.

> I've been trying to internationalize some code, using egcs 2.95.1 with
> glibc 2.0, but I'm getting errors for many of the wide character funtions
> (wprintf, wsprintf, etc.). 
>
> I had thought this was part of the standard C library now - do I need to
> upgrade to a new glibc? What do I need to do to get the rest of the wide
> character functions for egcs on Linux? 

Yes, upgrading to glibc 2.1.95 is the easiest way to get the wide
character funtions.

> On a related note, do the other Gnu utilities - specifically GDBM - support
> wide characters in their API?

GDBM uses 'struct datum' for keys and values. A 'struct datum' can
hold any memory string - a regular locale dependent string, an UTF-8
string, a wide character string, or anything else.

> if/how program arguments in argv are passed as wide characters

No, argv is passed in the locale dependent encoding. You can pass
UTF-8 strings from one program to another through argv, if both the
caller and the receiver agree on the UTF-8 format - which is normally
the case only if the current LC_CTYPE locale is UTF-8.

Bruno