CharacterDB is a collaborative effort for gathering semantic data on the
structure of Chinese characters' appearances for Hanzi, Kanji and Hanja. One
goal is to gather data that can be used by the cjklib project.

See <http://characterdb.cjklib.org/>.
