Text shaping: support glyph decomposition
lojjic opened this issue · comments
Some languages require decomposition of characters into multiple glyphs, either by Unicode Canonical Normalization or from GSUB multiple substitutions.
To implement this we'll need a better way to track the original character indexes; the current algorithm only allows a fewer total number of glyphs than string characters, and we'll have to support the possibility of more glyphs.