RE: Arabic shaping behavior questions

From: Marco.Cimarosti@icl.com
Date: Wed Aug 09 2000 - 13:43:50 EDT


Bob Hallissy wrote:
> 1) Is the Arabic Joining Class [...] normative or informative?

Like it or not, it is normative. See
<http://www.unicode.org/Public/UNIDATA/UnicodeCharacterDatabase.html>, that
reads:

    ...
    ArabicShaping.txt (Section 8.2)

    Basic Arabic and Syriac character shaping properties, such as initial,
    medial and final shapes. These properties are *normative* for minimal
    shaping of Arabic and Syriac.
    ...

((( Personally, I think that this makes little sense. There is an even more
"minimal" shaping, that uses a single glyph for class "R" characters and 2
glyphs for most class "D" characters. Such an approach is used in the
Iranian standard charset, in some old DOS Arabic code pages, and in many
"font encodings". )))

> 2) U+0649 (ARABIC LETTER ALEF MAKSURA) has joining class "R",
> meaing in

I notice one more incongruity in entry U+0649: it is assigned to Link Group
"YEH". This means that it should have the same *four* glyphs as letter YEH,
with a different combination of dots.
And, thus, one more evidence that it should be in class "D".

_ Marco



This archive was generated by hypermail 2.1.2 : Tue Jul 10 2001 - 17:21:06 EDT