- Add `=decomposition@cid' for some Adobe-Japan1 characters.

- Add `=decomposition@cid' for some Adobe-Japan1 characters.
- Add some Adobe-Japan1 code points.
(U+3789): Use `->subsumptive' to separate representative nodes.
(U-00021C09): Use `<-unknown' for JU+5C32; don't use
`morohashi-daikanwa'.
(U+378A): Use `->subsumptive' to separate representative nodes.
(U-00021C0E): Likewise.
(U-00021C10): Use `->subsumptive'.
(U-00021C11): Use `->subsumptive' to separate representative nodes.
(U-00021C12): Use `->subsumptive'.
(U-00021C14): Likewise.
(U-00021C0D): Use `->denotational' and `->subsumptive'.