Dear users,
I have used font creator before and created few font with it; have got a question,
in all arabic TTF fonts when we type these letters one by one “A” “L” “L” “H” it automatically creates a special combined word for “Allah”.
What is this feature called, and how can i use it in font creator to make my own special words?
if my question is not understood; I would explain again.
Waiting for your help.
Regards
Taha
It’s called a ligature, and ligation is a fairly complex feature that is not natively supported in FCP. This is not a lost cause, though. If you are sticking to Arabic, the Allah ligature actually has a Unicode code point (for compatibility), at U+FDF2, and I would assume that most word processing software will automatically replace (autocorrect?) the A+L+L+H with U+FDF2, much like they will replace f+i with the fi ligature. There are a LOT of Arabic ligatures in Unicode, but also a few Latin ones ff, fi, fl, ffi, ffl, ſt, and st (ff, fi, fl, ffi, ffl, ſt, and st). I am assuming that you are looking to apply this to other situations as well, and here it is a bit more complicated. What you need to do is add OpenType substitution features to your font. With OpenType, you do not have to map your ligatures to the Private Use Area and then try to figure out how to access those code points, you simply declare a ligature glyph as equivalent to some series of characters. There are a number of (free!) ways to do this, and a lot of people have written about it in this forum.
http://forum.high-logic.com:9080/t/adding-opentype-features/1904/15
is a fairly comprehensive explanation of hand coding OpenType features. You can also download Microsoft’s Visual Open Type Layout program (VOLT), which has an instructional video, and documentation. You can download VOLT at http://www.microsoft.com/typography/volt.mspx . This is not a project for the feint of heart, though.
Now, the last problem is that only a few high-end programs fully support the full range of OpenType glyph processing. Uniscribe (the Windows glyph processor) only supports a minimum of complex rendering, limiting the features that can be used with a given script. So while Arabic text has a whole slew of positioning and substitution features enabled, allowing adjacent characters to link together, change forms contextually, and position marks, Latin text is limited to the placement of diacritical marks. Devanagari has a complex set of half-forms and ligatures that aren’t accessible to Greek text - or Kharoșți. Check out those links, explore the Microsoft typography site, and see if that doesn’t get you off on the right track.
-Van
Dear vanisaac
Tnx alot for your post, wow it seems much complicated than i thought, I would defiantly go through the link you gave me, so now i should understand that this thing can not be done using FCP alone,
Got another question.
When i open any arabic font in FCP, and I look at the properties of the Glyph “ALLAH” i find that it is mapped to something and that something is a big list of the section “Arabic Presentation A” there i find at the end “$FDF2 Allah word complete”, I mean it is mentioned that this Glyph is mapped to Allah word, so why cant I create other Glyphs and use the same mapping list which is 95% unused.
Now whenever I type the word Allah is shows a unique shape for it, whatever font , software or operating system it is, so what is making this change? the true type font itself or some other coding which is common in computer world?
Now whenever I type the word Allah is shows a unique shape for it, whatever font , software or operating system it is, so what is making this change? the true type font itself or some other coding which is common in computer world?
The font itself could be making that change, if the software you are using supports OpenType. It is possible that all programs that run Uniscribe (the Windows native character engine) will automatically change A+L+L+H into U+FDF2, and that it would automatically transform many of the other words found in Arabic Presentation Forms - A. It is also possible that the autocorrect feature in a word processor would do this, much like f+i becomes fi - not having an Arabic language pack, I can’t say how many of these things are handled exactly.
When i open any arabic font in FCP, and I look at the properties of the Glyph “ALLAH” i find that it is mapped to something and that something is a big list of the section “Arabic Presentation A” there i find at the end “$FDF2 Allah word complete”, I mean it is mentioned that this Glyph is mapped to Allah word, so why cant I create other Glyphs and use the same mapping list which is 95% unused.
When I opened up Adobe Arabic - Regular, the Allah ligature was only mapped to U+FDF2. What other code points is it mapped to in on your fonts?
-Van
Now whenever I type the word Allah is shows a unique shape for it, whatever font , software or operating system it is, so what is making this change? the true type font itself or some other coding which is common in computer world?
I believe the operating system does it automatically, as well as connecting Arabic letters correctly. Since (in my experience) it happens in software that does not have OpenType support, it can’t be software dependent.
Which, counterintuitively, means that it probably is OpenType features that are being automatically handled by Uniscribe. So you’re going to have to do OpenType substitution and figure out which feature to bind it to - there should be a standard substitution feature for Arabic documented somewhere.