Re: Questions about yeh, hamzah on yeh, alef maksura and dotless ba

[Gregg Reynolds <gar at arabink dot com>]

Meor Ridzuan Meor Yahaya wrote:
> Dear all,
...
> 2. Seems like most referrence I could find saying that alef maksura
> occurs only in final forms (for arabic, other languages does have
> medial and initial form). I think the word ila (to) , uses alef
> maksura (correct me if I'm wrong). However, when the words become
> ilaina ( to us), it uses yeh, right? So, may I know which words uses
> alef maksura, or it's rule?

Rule number one:  disregard Unicode naming and go by character 
properties.  "Alef maksura" is incorrect, but Unicode refuses to correct 
names; instead they added an annotation indicating (incorrectly) that 
alef maksura "represents YEH-shaped letter with no dots in any 
positional form".  I.e. dotless yeh.  If you look at the bidi properties 
file, 0649 is classed as dual-joining.  So the *codepoint* represents 
dotless yeh; but it does *not* represent alef maksura.

Alef maksura is not a letter form, it is a grammatical concept; it 
refers to a short vowel /a/.  Sometimes a dotless ya in final from is 
called alef maksura (e.g. ila إلى, rama رمى); sometimes a standard alef 
in final form is called alef maksura (da'a دعا for example).

Which means "alef maksura" is not a good candidate for a code point; 
instead it should be "dotless yeh", but Unicode never changes names once 
they are approved.

The general rule is that a suffix added to a final ya-shaped alef 
maksura causes it to take the shape of a standard alef.  Words like 'ala 
على and ila إلى are exceptions to this rule.  It's important to 
understand that the rule reflects phonology, not the written forms.  In 
other words, rama رمى becomes رماه ramAhu because the vowel changes from 
short to long.

> 3. There are instances in quran where the small alef have chair. Last
> time I encode it as alef maksura + small alef. I think for final alef
> maksura is correct, but  it is wrong when it occurs in initial or
> medial forms. Example: Sura 104 aya 5(initial) and sura 102 aya 1
> (medial). I chose alefmaksura because it seems like it is alef.
> However, from above, seems like alef maksura only occur in final form.
> So, it should be a different character. So, what character it should
> be? Someone suggest it is dotless ba. If it is true, can someone
> explain, the use (description) of dotless ba? Unicode document does
> not say much.

dotless yeh 0649 is what you want.

> 4. Hamza on yeh (U+0626). Should I use this character or not?Mr Yousif
> (Quran Project leader) against the use of it. Instead, suggest to use
> yeh + hamza above/below. Because I want to make it compatible with
> him, I uses alef maksura + hamza above/below. However, seems like it
> is wrong. I choose alef maksura over yeh because to make it compatible
> with other font (other font most likely to render the yeh with dots).

I think he is right; use dotless ya + hamza above/below.  Unfortunately, 
you can't be certain that all fonts will include all four forms for 
dotless ya (a/k/a "alef maksura"), since that was a Unicode change; 
through Unicode v 3.0 "alef maksura" was classified as right-joining. 
Starting with update 1 to 3.0, it has been classified as dual-joining 
(http://www.unicode.org/Public/).  But any fonts that have been made 
after 3.0 update 1 should work.

> The other problem is, in unicode document, it specifically mentioned
> hamza "above". However, in the mushaf, it render the hamza depending
> on context, either above or below. In my font implementation, the
> hamza will appear depending on the mark, so no problem. However, other
> font might not be "intelligent" enough. So, the question again, should
> 0626 be use? I personally think it should be use. The problem of above

My own preference is to have above/below codepoints for hamza and 
shadda; it will make searching a bit easier.  But I don't think there's 
anything wrong with using 0626.


-gregg