GNOME Bugzilla – Bug 123406
need a way to do unicode normalization in atomic steps
Last modified: 2011-09-05 02:00:49 UTC
Need a way to do unicode normalization in one normalization unit at a time.
Pango wants this so that it can render canonically equivalent strings
One of our assumptions was wrong-- a normalization unit isn't
necessarily the same as a combining character sequence. For instance,
U+09CB = U+09C7 + U+09BE, but U+09C7 and U+09BE both have combining
class zero. I'm not sure there's a simpler way to figure out the
atomic normalization units than to see what happens when you do the
Created attachment 21459 [details]
the patch that led me to make this discovery (so it doesn't get lost)
Adding the PATCH keyword and marking priority as high.
The patch is informative, not meant to be used.
We have this now.