GNOME Bugzilla – Bug 393307
shows U+200D boxes
Last modified: 2012-08-18 17:49:31 UTC
When looking for instance at the Thai unicode block, using the Norasi font, some characters like U+0E31, U+0E34, etc. are rendered with a U+200D box. This is probably because the Norasi doesn't provide a glyph for U+200D, but since it is a formating character, nothing should be rendered at all, the font shouldn't have to provide a glyph. Maybe it is just a pango bug, but I can't reproduce the behavior in gedit for instance.
Created attachment 79486 [details] snapshot
Is this with Pango HEAD? Pango-1.15.2? Or an older version?
Happens with pango HEAD too.
Not a hangul module bug. It doesn't handle thai script and it never renders zero width characters as hex boxes. BTW, I think adding pango_is_zero_width() check to thai module should fix it.
Not a hangul module bug. Assigning it to general.
gucharmap tries to render individual combining character by putting ZWJ before it. But that is not currently supported by Thai module. I think we should add ZWJ support to it.
As it was suggested, you should use pango_is_zero_width() to filter out such characters. Please see how other modules use it.
We've merged the HarfBuzz branch. Closing fixed.