Den 4. jan. 2012 kl. 01:07 skrev Ferren MacIntyre:

> Hi guys,
> 
> I have a problem.
> 
> Qur'an, for instance, does not really have an apostrophe in the middle, but, as I understand it, a hamza, which should look rather like a small 'c'.

A reversed small c superscript; the right-way c is used for 'ayn, a different character.


> Character Viewer shows a 'high hamza', a 'c' with a little left tail, and although it claims that I have fonts that include the right glyph, none of them can be 'inserted with font'. I have faked it by using U+7580, which is literally a small high 'c', and while it looks OK in text in HTML, it will not appear if preceded by <li>, as in a table of contents. 
> 
> Surely some of you who work with Arabic have solved this problem? I would like to get rid of the apostrophe, because for obscure reasons I have to keep switching between 'Educate Quotes' and 'Straighten Quotes' in Text Wrangler,
> where I repair the dog's breakfast NWP produces when it exports HTML, and the fake hamza gets caught up in this and confused.

There are indeed two ways to produce hamza (and its reverse, ayn). The "old" and cop-out one, is as you say, to use "smart quotes", a single opening quote for ayn, a closing quote for hamza. Apart from often being mangled by word processors, it is also a cop-out, because it can be confused visually with a real smart quote or apostrophe, when it occurs at the end of the word (like in the common word ‘ulamā’, "scholars"; it looks like the word is put in quotes).

The "good" way is to use the proper Unicode character for these two, which is found in "Character Viewer" as decimal 702 (U+02BE, your hamza) and 703 (U+02BF, 'ayn). They are called "modifier letter right/left half ring", but are indicated to be used for these two Arabic characters. Unfortunately, these two are not among the characters you can type with the US Extended keyboard, so you are forced to use Character Viewer, unless you hav PopChar or one of the more spcialized keyboard layouts that let you type it directly (as my American Diacs, incidentally: http://www.smi.uib.no/files/diacs.zip). But for the occasional usage Character Viewer should be OK.

There you can also see listed which fonts you have that contain your character. Not all fonts contain these two (they are more rare than "regular" diacritics) but on recent Macs, there will be about half a dozen or so built-on fonts with both.
Notice that most fonts present these characters just as it is described, a "half-ring". In my view, it should look like any other character, so it should be serifed in a serif font. I have them thus in my Jaghbub font, a few others follow this view, but in Times New Roman and other standard fonts, the characters look fairly measly in small sizes. 

"Modifier small c" (7580, a.k.a U+1D9C) is a separate character, used for certain phonetic transcription purposes. It is not commonly used for hamza/ayn. High hamza (1652 / U+0674) is a real hamza for Arabic script, it is used in some Central Asian languages when written in Arabic script, Kazakh (and/or Uighur?), I believe. 

Can real hamza/ayn be used on web pages / HTML? Yes, as long as the reader's computer has these characters, which we can mostly assume today that they have, in Macs, Lucida Grande will at least provide, in WIndows either Times New Roman or Arial Unicode with these characters will most often be present on the reader's computer. If by "list" you mean that they will be sorted correctly in automatic sorting (i.e. ignored), that is generally not the case, they will almost always be considered as separate characters after z, even in software that sorts other diacritics as a/line above and s/dot under correctly.


> 
> So, does anyone know how to get a hamza to work in an HTML5 list? (Rumor going back to Ubay ibn Kab claims that Mohammed pronounced Quran *without* the hamza, but I need at least one to show what I'm talking about.)

There is also the debate about whether "history" is correctly transcribed tārīkh, as Wehr's Dictionary has it, or taʾrīkh, as the grammar would indicate.... But we need the hamza. And the 'ayn.

Knut
(see also http://www.smi.uib.no/ksv/diacs.html )
> 
> Happy New Year, and TIA.
> 
> Ferren
> ===================
> (Dr) Ferren MacIntyre
> 1 ch. des Echarts
> Campagne sur Aude
> 11260 France
> +33 (0) 468 748870
> ===================
> MacBook Pro 5,1; 2.4GHz, 4GB
> OS X 10.6.8 Snow Leopard
>