search

grigorig / ucdn

Unicode Database and Normalization
Other
85 stars 12 forks source link

East Asian width of U+30000 is wrong? #2

Closed behdad closed 12 years ago

behdad commented 12 years ago

I imported UCDN into HarfBuzz. Running the test suite shows that UCDN returns 1 for East Asian width of U+30000, while the test suite expects 2.

grigorig commented 12 years ago

Hmm... looks like U+30000-U+3FFFF is a special CJK range, although it doesn't contain any valid characters yet. Python, this is where I have the unicode database generation script from, just assigns full width to all unknown characters. This of course fixes the problem at hand, but I am not sure if it will cause any other issues.

How can I run the test suite?

behdad commented 12 years ago

You need to have glib to run the test suite. It's in test/api, and "make check" runs it. Right now you can't force it to use ucdn when glib is enabled, so you need to modify hb-unicode.cc at least.

grigorig commented 12 years ago

OK, this was harder than I thought, but it is fixed now. See the commit message for details.