[XeTeX] Converting legacy encodings to utf-8

[Peter Heslin]

Will Robertson <wspr81 at gmail.com> writes:

> On 12/07/2006, at 8:10 , Jonathan Kew wrote:
>
>> Really, though, all this language support stuff needs to be updated
>> for Unicode. Who's ready to write the unibabel package? :)
>
> Personally, I'm hoping for a port of mem:
>    <http://mem-latex.sourceforge.net/>

I'm wondering about a more light-weight solution.  This unibabel.sty
could be just like babel.sty, except that for "problematic" languages it
would load language-uni.ldf instead of language.ldf.  Problematic
languages would be those for which a distinct set of utf-8 hyphenation
patterns are required, or for which the old babel sets up legacy 8-bit
encodings.  The new language-uni.ldf files could be copied from the
respective language.ldf, and then have the code for legacy encodings
ripped out and the new hyphenation patterns substituted.

It's been a while since I looked at the internals of babel, so maybe I'm
not seeing pitfalls in this simplistic approach.  As Ralf says, this
unibabel package will need to know what the hyphenation patterns will
be.

-- 
Peter Heslin (http://www.dur.ac.uk/p.j.heslin)