From: Kent Karlsson (kent.karlsson14@comhem.se)
Date: Fri Jun 09 2006 - 13:56:03 CDT
Philippe Verdy wrote:
> Note that if you compare case insensitively and don't care
> about other variations [...] eliminates all combinining
> characters with combining class greater than zero. In fact
> the filter(NFKD()) composition can be precomputed easily by
> computing the closure of all decomposition mappings from the
> UCD main file, and eliminating all combining characters with
> non-zero combining classes.
Whatever gave you the idea that "characters with combining class
greater than zero" would be used only for "variations"? That is
just plainly wrong. (Not just for Tibetan.)
/kent k
This archive was generated by hypermail 2.1.5 : Fri Jun 09 2006 - 14:15:08 CDT