Re: non-positional numerals in Unicode?

From: Antoine Leca (Antoine10646@leca-marti.org)
Date: Fri Feb 17 2006 - 04:18:04 CST

  • Next message: Kit Peters: "Re: Unihan numeric support in Java?"

    Not sure this is the best forum.

    On Thursday, February 16th, 2006 16:40Z, Kit Peters wrote:

    > What numerals are encoded in Unicode that represent non-positional
    > numbering systems (e.g. Egyptian hieroglyphs, Babylonian - I know
    > those aren't implemented in Unicode, but they're examples of
    > non-positional numbering systems)?

    For me, Babylonian is known as one of the best example for _positional_
    numbering systems. I know Babylonians and their neighbours used also
    non-positional systems, as the civilizations known under those names span
    millenaries. It makes me wonder if you did not make a mistake while asking;
    in such a case, then you should filter the Unicode Character Database for
    General Category being "Nl" or "No". You can answer several such queries in
    a similar way. Even more, some are already answered, and the result kept as
    "derived properties" and the like. Read with attention
    http://www.unicode.org/Public/UNIDATA/UCD.html, there is many information
    there.

    The problem with your initial query is that a lot of such numerals are not
    encoded per se, but rather the character primarily "designed" for another
    use was reused as numeral. A typical example would be the letters of the
    Semitic alphabets, like Hebrew, and derivations like Greek or Gothic. The
    letters of the Brahmi syllabaries also enter this category ("cabalistic"
    uses), particularly before Indians use decimal digits; Nagari letters are
    usually used these days to show this kind of uses for easiness, but of
    course the real letters used predate those.

    > - I know those aren't implemented in Unicode,

    Depend on your definition of implemented. If you have a good grasp at them,
    download the current beta version of 5.0.0 (d4 in online as I write this)
    and analyse the datas there. If you find something incorrect, report it (it
    is on public review), I guess such work will be very much appreciated.

    Hope it helps,

    Antoine



    This archive was generated by hypermail 2.1.5 : Fri Feb 17 2006 - 13:27:42 CST