From: John H. Jenkins (jenkins@apple.com)
Date: Mon Jan 25 2010 - 11:37:58 CST
On Jan 24, 2010, at 10:59 AM, Ed Trager wrote:
> What is the romanization system used for kJapaneseOn and kJapaneseKun
> readings in the Unihan_Readings database?
>
> I think that UAX#38 does not say ...
>
kJapaneseOn and kJapaneseKun are two of the oldest fields in the Unihan database, with data provided initially by RLG and/or Xerox, and they've hardly been touched since. It's old enough that we don't even have adequate internal documentation, which is why the documentation in UAX#38 is so sparse.
Personally, I feel that the best course in the long term would be to do what we did with the kKorean field: create a new field with known provenance and documented contents and deprecate the current fields. (By using katakana and hiragana, we could distinguish on and kun readings and use only one field.) I haven't had time to do this myself, but if someone were to volunteer data from a source we could use in the Unihan database, I'm sure the UTC would be willing to consider making the transition.
=====
John H. Jenkins
jenkins@apple.com
This archive was generated by hypermail 2.1.5 : Mon Jan 25 2010 - 11:40:29 CST