Re: Unihan: kTraditionalVariant vs. kZVariant

From: John H. Jenkins (
Date: Tue Dec 08 2009 - 13:04:57 CST

  • Next message: Christopher Key: "[OT] New character request"

    It's an error in the kZVariant data. I'm currently in the process of fixing errors in the kZVariant field and will take care of this one right away.

    To add a variant relationship of any other type, we'd need an authoritative source. I did a quick check of the Hanyu Da Zidian, and it identifies them as semantic variants, so we'll go with that for the moment.

    On Dec 6, 2009, at 12:29 AM, Ben Monroe wrote:

    > Greetings.
    > I was recently going over Unihan_Variants.txt and noticed the following entry:
    > U+5909 kZVariant U+8B8A
    > The corresponding Unihan entry is here:
    > I would have expected the relationship to be kTraditionalVariant
    > rather than kZVariant.
    > I reviewed TR38:
    > The issue I assume comes down to "abstract shape". And yet I have a
    > hard time concluding that these two have the same abstract shape.
    > Beyond the apparent visual differences, they both have different
    > radicals as well as significantly different stroke counts as well. And
    > even if the kZVariant relationship holds, I would still expect and
    > desire a kTraditionalVariant mapping as well.
    > Would someone please explain this to me?
    > Regards,
    > Ben Monroe
    > Tōkyō, Japan

    John H. Jenkins

    This archive was generated by hypermail 2.1.5 : Tue Dec 08 2009 - 13:13:33 CST