MichaelL_at_frogware.com wrote:
> Note: Perhaps a solution is to have two variants of the procs, one for
> code points and one for code units. The code units variants would
> guarantee O(1) and the code point ones wouldn't.
That might make sense. (I assume you mean "Unicode scalar value"
rather than "code point". I checked
http://www.unicode.org/glossary/ )
The goal would be to permit at least UTF=8 and UTF-16 internal
representations. (Other representations should also be allowed,
but might be less useful.) The O(1) functions would get and set
the 8-bit or 16-bit values.
--
--Per Bothner
per_at_bothner.com http://per.bothner.com/
Received on Thu Mar 15 2007 - 13:15:20 UTC