Currently they all do their own parsing/merging which is getting quite ugly. Do a "base parse" per node or something?
Currently they all do their own parsing/merging which is getting quite ugly. Do a "base parse" per node or something?