I'm also curious to know the answer to this. I'm currently using HB_BUFFER_CLUSTER_LEVEL_MONOTONE_CHARACTERS as I read somewhere that it produces finer grained cluster data.
I'm currently adding the remaining bits of editable text support that I'd skipped over before, and part of this is dealing with picking characters via a cursor hit-test. This is mostly working now, but I still need a way to handle picking characters that have become ligatures during shaping (eg, "fi"), as they exist as a single glyph in my shaped data. Would I re-shape them using different settings? Would the cluster level settings help here? -Jamie.
_______________________________________________ HarfBuzz mailing list [email protected] http://lists.freedesktop.org/mailman/listinfo/harfbuzz
