Re: Hebrew composition model, with cantillation marks

From: Philippe Verdy (verdy_p@wanadoo.fr)
Date: Wed Oct 29 2003 - 11:41:40 CST


> Normally meteg is positioned below and to the left of any other low
> centred mark. Less frequently it is positioned to the right of a low
> centred mark. But it is always to the left of a low right mark i.e.
> yetiv or dehi. It can also be centred within a hataf vowel. In
> http://www.qaya.org/academic/hebrew/Issues-Hebrew-Unicode.html section
> 3.4 I suggested using CGJ to encode right meteg. Jony Rosenne prefers a
> separate right meteg character. The best solution for right meteg would
> be a single meteg with the same combining class as all low vowels and
> all low centred accents. But this would not solve the medial meteg issue.

I'm not sure we need this: the distinction of right and left meteg only
occurs
with other vowels or cantillation marks positioned below.

As they will share the same combining class 220, the canonical ordering will
preserve their relative order, which is meant to be represented
left-to-right in
Hebrew if several points, vowels or mark collide at the same position.
So the input order makes this distinction, and we don't really need
then a right meteg.

The only thing that this proposal does not handle is the case of the medial
meteg.

It could be addressed by encoding either this medial meteg, or by
encoding the combination of the meteg (vertical stroke) with hataf vowels
on which this insertion occurs:

The first:
    o o o U+05B1 POINT HATAF SEGOL or
      o o SIL.org-proposed BIBLICAL VOWEL HATAF SEGOL

produces:
    o o | o BIBLICAL VOWEL HATAF SEGOL MEDIAL METEG
      o | o

The second:
    ---- o U+05B2 POINT HATAF PATAH or
            o SIL.org-proposed BIBLICAL VOWEL HATAF PATAH

produces:
    ---- | o BIBLICAL VOWEL HATAF PATAH MEDIAL METEG
           | o

The third:
    ----- o U+05B3 POINT HATAF QAMETS or
      | o SIL.org-proposed BIBLICAL VOWEL HATAF QAMETS

produces:
    ----- | o BIBLICAL VOWEL HATAF QAMETS MEDIAL METEG
      | | o

Another solution could be to define another character combining below, and
representig the HATAF right part.

In that case, the combination of <hataf segol>+<medial meteg> could be
represented as <hataf>,<meteg>,<segol> where all these characters have the
same combining class "below" (220 in the sil.org proposal)

This new HATAF point could be added with an additional decomposition for
HATAF vowels, but this would require adding the concept of required
composition
for hataf vowels for NFD and NFKD forms. That's why and would not like it,
and
would much prefer a medial meteg or precombined hataf vowels with medial
meteg.



This archive was generated by hypermail 2.1.5 : Thu Jan 18 2007 - 15:54:25 CST