09-24-2008, 10:20 PM
erdalronahi Wrote:I have set up a wiki page to discuss the initial hamza issue. It is found here:Erdal,
http://wiki.ferheng.org/doku.php/initial..._in_sorani
I have found that Unicode has some of the needed forms. I would much prefer a Unicode-conform solution.
Erdal
I opened a new topic to be more specific about the subject.
Your wiki page is very interesting but I think what you are looking in Unicode is not what you need. You must forget about all ligatures and presentation forms, even those already encoded in Unicode because Unicode doesn't do this anymore. That is the reason you don't see the presentation forms of newer encoded characters such as yeh with small v above. So think as if Unicode blocks of Arabic Presentation Forms A and B never existed (for any language).
For compliance with Unicode, your Kurdish keyboard should only carry character codes from U+0600 Arabic block (and of-course punctuation marks and other non Arabic script specific codes). Unicode has delegated the presentation forms of the characters to the font and text engine technologies. The keyboard (and text) code just tells the font which character should be properly presented in the context.
The issue about proper encoding of Kurdish text here is, whether hamza before vowel letters is a character in itself or just another presentation form of the same vowel character. Is ئا at the beginning of a Kurdish word the same alef or something else? Is it a single code or a composite code? This is the real issue about those vowels in Kurdish language. If ئا at the beginning of a word, will reapear in a derivative of the same word, but not at the beginning of the word and not with hamza, then it seems to me that we are simply dealing with a single character and a single code, but with different presentation form.
XB Shafigh Kurd is made based on this concept.
In any rate, the first step in resolving the problem is not Unicode, but Kurdish language itself that should give a clear definition of how it is structured. Language structure doesn't conform itself to Unicode. Unicode does. If the issue of hamza in Kurdish vowels is well defined, then font technology will be able to produce these vowels correctly within the context. The same way that font technology (and not Unicode) produce the correct shape of 'Seen' within the context.
I think experimenting with XB Shafigh Kurd can open up the issue and help in finding a sound solution.