Japanese JSON available

For Japanese only, we have character groups classified according to Heisig Vol. II (pure, semi-pure, etc.), along with all the on'yomi, plus the highest frequency vocab we could find in EDICT for each on'yomi, plus the historical on'yomi spelling, the kyuu- and shinjitai forms of each character, plus of course the phonetic component used by each character in the group. The output is ordered by group regularity, then number of chars in the group, then frequency of most frequent word used in the pronunciation examples in the group.