making the wide character properties Unicode compatible

[Geoff Keating]

> From: Bruno Haible <haible@ilog.fr>
> Date: Mon, 25 Sep 2000 14:50:00 +0200 (CEST)

> The wide character classification functions in an UTF-8 locale currently
> have the following bugs and oddities:
> 
>   - print and graph: are lacking characters like
>       00A0;NO-BREAK SPACE;Zs;0;CS;<noBreak> 0020;;;;N;NON-BREAKING SPACE;;;;

I'd have thought of 00A0 as a control character, if it's not
whitespace.  It certainly doesn't seem to have a printable
representation...

Still, you probably know this better than I.

-- 
- Geoffrey Keating <geoffk@cygnus.com>