Re: Arabic letters separated by markup

From: Richard Wordingham (
Date: Fri Jun 17 2005 - 02:15:49 CDT

  • Next message: Asmus Freytag: "RE: Arabic letters separated by markup"

    Mark Shoulson wrote:

    > What about Devanagari K-SHA? I can't disassemble that at all. Nor

    There are, of course, claims that these should be letters, especially ja
    with its rather odd pronunciations. For ks.a I'd put the break at the upper
    point where the curve crosses itself. For ja I'd say the final downward
    stroke was the a part.

    > I think we simply cannot ever count on a computer being
    able mechanically to pick out which part of a ligature belongs to what

    No-one was suggesting this. OpenType layout controls may provide crude
    positioning information for cursor placement within a ligature, but 2-D
    placement is really needed, as for example in the Word equation editor.
    While ks.a shows the normal top-to-bottom ordering, this ja is almost
    right-to-left! The point is that the computer has to be told the analysis
    of the ligature.


    This archive was generated by hypermail 2.1.5 : Fri Jun 17 2005 - 02:19:07 CDT