Hi all,

thinking a bit about support for different character sets for
CDText raised a more general question about how strings should be
handled in libcdio.

The current status is, that the character set is always ignored
except in iso9660_fs.c, where strings are converted to the locale
charset (obtained by nl_langinfo(CODESET)). With this approach,
reading e.g. a Japanese CDROM on a system with Latin-1 locale will
be doomed to failure.

My idea would be, to define strings (like filenames, CDText strings)
to be *always* UTF-8. The advantage is obvious since UTF-8 is the
default locale charset on many modern Linux distributions and in
gtk-2.x. Furthermore, any conversion problems (like Japanese -> Latin-1)
would be moved out of libcdio.

But since I live in a Linux-only ecosystem, I don't know if such a
move would break other applications and/or OSes.
Would anyone see a problem with defining libcdio strings UTF-8 only?

Thanks

Burkhard


_______________________________________________
Libcdio-devel mailing list
[email protected]
http://lists.gnu.org/mailman/listinfo/libcdio-devel

Reply via email to