Re: compatibility characters in Arabic block

From: Kairat A. Rakhim (rakhim@aport2000.ru)
Date: Tue Jul 10 2001 - 03:07:17 EDT


> On Tue, 5 Jun 2001 Peter_Constable@sil.org wrote:
>
> > In the Arabic block, there are four characters with compatibility
> > decompositions: 0675 - 0678. The names list says that these are for
Kazakh.
> > I'm wondering why these have compatibility decompositions and not
canonical
> > deompositions (possibly excluded from composition).
>
> I don't know about the roots, but I think this is because of the weird
> behaviour of U+0674. While not a non-spacing mark, it seems that it goes
> "before" the previous letter. This surely needs clarification by someone
> who knows about the usage in Kazakh.
>
> --roozbeh
>

I know only Kazakh (Arabic) alphabet which was used in Kazakhstan until
1929. Maybe, Kazakhs in China use another set of characters or spelling
rules? As far as I know there was a Proposal from Xinjiang Government to
ISO/IEC JTCI/SC2/WG2 with number 2027. Where can I find this paper?
But if the spelling rules are the same then it looks strange in my opinion.
I can present sources printed in Kazakhstan.

ARABIC LETTER HIGH HAMZA (U+0674) is called as 'D?jekshe belgisi' in Kazakh.
It goes before the whole word and means that all vowels in the word are
front. Else the same vowels are back. It doesn't mean any sound itself. It
is not used if the word contains such letters as KAF(U+0643), GAF (U+06AF),
AE (U+06D5). These letters always go with front vowels.
Thus, U+0674 refers to the _whole_ word, not only to _next_ letter. I
wonder, where U+0675 - U+0678 may be used except alphabet tables?

By the way, I can't find final form of AE in Unicode. It should look as
ARABIC LETTER HEH FINAL FORM (U+FFEA). This letter corresponds to 'E' in
Latin-based alphabet and to Cyrillic letter IE (U+0415/U+0435) in Cyrillic
one. Isolated form (U+FEE9) and final form (U+FEEA) of Arabic letter HEH are
not used in Kazakh. That is HEH in Kazakh has another contextual forms. Both
isolated and final forms should look like to initial form. Unfortunately, I
can't find any reliable sources with these rare glyphs.

Also, U+0678 is called as HIGH HAMZA YEH and decomposed to U+064A. In Kazakh
similar letter means front vowel and should be composed from back vowel ALEF
MAKSURA (U+0649) and U+0674.
That is Arabic contextual forms are language-dependent.

Kairat A. Rakhim,
rakhim@aport2000.ru



This archive was generated by hypermail 2.1.2 : Tue Jul 10 2001 - 02:18:03 EDT