2011/8/23 Heiko Oberdiek <[email protected]>:
> * The latter composition "o+dieresis" is not replaced by the
>  glyph "odieresis". In the documentation of LuaTeX I found:
> | 2.3 UNICODE text support
> | ...
> | Normalization of the Unicode input can be handled by a macro package
> | during callback processing (this will be explained in section 4.1.2).
>
>  This seems not be done by fontspec. Can this enabled/supported by fontspec
>  or does another package exists for Unicode normalization?

Such a package would have to include an extension module to wrap the
ICU library (unless somebody wants to reinvent the wheel and implement
the normalization algorithms in Lua). In fact, it is somewhat common
to normalize to form D on input to simplify processing. This would
break \catcode and the likes (which are rather ill-designed anyway
because they act on single code points instead of sequences). But even
without normalization decomposed sequences should just work. I think
this is an engine/backend problem.

Reply via email to