Re: [edict-jmdict] Chinese names within headwords

[Paul Upchurch <upchurch.paul@*********>]

I'm fine with including both the hybrid (e.g. かみチャ)/hiragana (e.g.
さんしょくどうじゅん) and the katakana for these terms, provided we don't put
[nokanji] on the katakana. As an example:
上家(かみチャ;カミチャ[nokanji]): You'd normally expect to see this in kanji, and if
it has furigana, it'd be かみチャ. You wouldn't expect to see 上家(カミチャ).
But "上家(かみチャ)" gets 9 hits, whereas "上家(カミチャ)" gets 2.2k.
If we add [uk] to the above, now you'd normally expect to see カミチャ without
any kanji.
But "カミチャ" gets 5k hits, while "上家" "麻雀" gets 173k.

In order to reflect the way the term is actually used, as well as its
etymology, I think our best option is to accept a certain degree of
exception for these Japanese/hybrid-origin mahjong terms.
上家(カミチャ;かみチャ) [lsrc=chi/p:]
And for 三色同順 (based on hits):
三色同順(サンショクドウジュン;さんしょくどうじゅん;サンショクドージュン)
If we wanted to really reinforce that despite the katakana, it's a
Japanese-origin term, I suppose it could have [lsrc=jpn:] (which I don't
think is used at all currently).