BH> ave you explored various ways of doing the conversion, including
BH> iconv?
Nope. But I've modularised the actual conversion code (hiding it
behing a ``charset'' object) with just such alternatives in mind.
BH> Now that it is in libc, it is desirable for all programs to use
BH> it, as far as possible, for the following reasons:
I fully agree with your points 1. and 3. However,
BH> 2) Ease of upgrading.
For me, upgrading fontenc is very simple: add a single new mapping
file in lib/X11/fonts/encodings, rebuild the encodings directory. I
don't think it would be easy to get e.g. DEC Special Graphics into
every single variant of libc around.
BH> it seems you could [use] your existing infrastructure for
BH> determining the character boundaries and for resynchronization,
BH> but [use] iconv() for doing the conversion from/to Unicode.
Technically, you are absolutely right, as usual. But esthetically...
well...
Juliusz
-
Linux-UTF8: i18n of Linux on all levels
Archive: http://mail.nl.linux.org/linux-utf8/
