From: Denis J. <mo...@gm...> - 2005-10-24 21:43:02
|
Hi, I've figured out how to use the Contextual or Chain substitution for Latin script. So I can add two features: * 'i' and 'j' being substituted with their dotless components when followed by a diacritic above. * substitute the diacritics by upper case size diacritics when preceded by a capital. I've also started adding above mark and mkmk so depending on your font renderer you should be able to stack diacritics on the characters that have anchors. I'll need to either shift the uppercase diacritics U+10001-1000C range and thus all their dependencies, or make reference and add the shifted ones somewhere (in the PUA ?). They need to be in the negative on the X-axis so font renderer don't give them unwanted width. I was also thinking about adding the contour tone marks. These aren't defined in Unicode but are generally use in IPA as ligatures of the different U+02E5-02E9, there are many possible ligature. About Mono and diacritics, I have no idea what' s the best way to go about this. Some languages can only form glyphs with composition. Maybe we could support some precomposition in the PUA? We could use the Unicode Named Sequences list as a reference: http://www.unicode.org/Public/UNIDATA/NamedSequences.txt even if it is missing many glyphs. Here's a new claim for more diacritics: 0305 0309 030E-030D 0310-0311 0313-0317 031B 031F 0324-0326 0329-0333 0339-033F 0361 For some reason using mkmk was tricky, I don't know it is my patched Pango or FF that is picky, but mkmk only work on a diacritic if the base-mark was defined before the mark-mark. Cheers! -- Denis Moyogo Jacquerye --- http://home.sus.mcgill.ca/~moyogo |