Michael (michka) Kaplan wrote:

> Actually, most "internal" mechanisms do not need cch calculations; for them
> the count of code points is fine. Since you do not work much with platforms
> that use it, I guess you would not have run across this fact?

IMO if you need just code points which are not characters, that code can 
use raw bytes from UTF-8 as well.

Regards, Igor






Reply via email to