On 07/21/11 15:27, Evan Martin wrote: > On Wed, Jul 20, 2011 at 8:21 PM, Behdad Esfahbod <[email protected]> wrote: >> I'm wondering: should harfbuzz try compatibility composition/decomposition >> (NFKC/NFKD) if a font doesn't support a character? >> >> It's easy to do so, but I'm sure how desirable that is. > > In general, the less harfbuzz does, the better; makes it more > pluggable into weird environments.
The current callback design makes it quite portable. If a callback implementation doesn't implement, say, decompose(), then NFD wouldn't work. So that's not an issue. > If it's not possible to do this outside of harfbuzz or if having > harfbuzz do it somehow affects performance in an important way, then > maybe I'd say it's a better idea. What I can do is to leave it to the callback implementation to decide whether to return canonical or compatibility decomposition. For now I'm working with 1:2 decompositions which only support canonical. Want to get this working first... Thanks for all the feedback. behdad _______________________________________________ HarfBuzz mailing list [email protected] http://lists.freedesktop.org/mailman/listinfo/harfbuzz
