RE: Flex and 32-bits characters

[Hans Aberg]

At 13:16 +0100 2002/08/26, Mark Weaver wrote:
>Here we have a slightly greater problem anyway, which is not solved by
>UTF-32.
>Consider a g with an acute accent.  Now to a user, they might be trying to
>match a word containing this character.  Sounds reasonable.  However this in
>Unicode is a combining character sequence, which is not represented by a
>single code point.  Even in UTF-32.  So UTF-32 doesn't fix the fundamental
>problem.

Why would one not use:
01F4    LATIN CAPITAL LETTER G WITH ACUTE
        : 0047 0301
01F5    LATIN SMALL LETTER G WITH ACUTE
        * Macedonian and Serbian transliteration
        : 0067 0301

I think that the suggestions of the TeX successor in the LaTeX group
resulted in that one will have to attach an external code convertor that
identifies those combining character combinations and replaces it with
single Unicode character codes whenever possible. That should cover most
normal uses.

  Hans Aberg