Computer programming languages aren't really human languages at all. Going the reverse direction, I once came across a book in Japanese where the amongst all the katakana and the pictograms would occasionally appear words like "GOTO" and "IF/THEN".
If you want an extreme case, take a look at the APL programming language. Back when it was first invented and graphics terminals were virtually non-existent, you had to actually replace the typing element on your keyboard/printer device in order to get the characters that the language used.
So, Arabic is no real problem. You need to do your lexical scanning right-to-left, of course, but that's trivial. Just be glad you're not doing Mongolian (vertical text)!
An IDE is no substitute for an Intelligent Developer.
Joined: Mar 18, 2009
I thought to create a Java compiler that takes as parameters the code written in Arabic but converted into ASCI code. so I can benefit from the rich existing that has Java.
but above all, in your opinion, is it a good idea to reproduce the wheel and create a new programming language?
Is it not a waste of time?
ASCII stands for American Standard for Information Interchange. The Arabic abjad doesn't map very well onto it, especially when you consider letters like 'ayn, ghayn, and the two versions of "t" and "d". To say nothing of the fact that waw is used for "w", "o" and "u". Then there's the question of whether or not to vowelize and refinement such as shadda.
Fortunately, Java works in Unicode, so you're actually better off just using an Arabic text editor and the actual symbols.
But it seems that you're less interesting in creating an actual compiler as in making Java be more Arabic-friendly. Probably the simplest way to do that would be to run a pre-processing program that does code-page translation. As I mentioned above, it's not going to be totally faithful, but then I've seen similar issues in Spanish, since the letter "ñ" isn't part of the ASCII set, either. The stock code page translator we were using to convert EBCDIC to ASCII rendered it as the 2-character sequence "n~". Similar issues occur with languages like German umlaut, although they've been dealing with that for a long time.
Joined: Mar 18, 2009
I think at this level, there will be no problem, because Java 5.2 contains already the codes all abjads even same characters of the Qur'an.
Thank you very much for your help