unicode identifiers

Hi!

I would even suggest using my unicode patches for clang even if it's not
the final and standard conformant way to do it because of the following
reason: some os's (such as windows) still use encodings such as
windows-1252 or shift-jis. with my patches variables can be words in these
encodings too. of course the backends have to care about encodings, but
then the problem is where it belongs to and rejecting there every string with
msb set in a character is ok at first.

please tell me which action you took (accept/reject)

-Jochen