[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

VIM-6.0 & Arabic



Salam all,

On the heals of semi-success with emacs-21 (can display Arabic pre-unicode   
Mule characters backwards :-/) - I have some more good news to note.   

I downloaded the latest vim (VI Improved (www.vim.org) - if you are a vi
user, this package is a must for ya; it has multiple-split-window support,  
color, multi-undo/redo, etc -- really, really cool stuff -- one can almost
compare it to emacs) and saw that alot has been done in terms of
Bi-directional (Bidi) as well as Right-to-Left (R2L) support.

The latest beta (6.0an - you'll need the src, runtime, misc & lang tar files)
already has hebrew support and I was fortunate enough to add Arabic with one
major hurdle left - shaping.  With that said, I remember Chahine saying that  
he has some cool fonts that are position agnostic -- could you point me to
them please ?  We also need to add the necessary code to do shaping to VIM, the
maintainer (Bram and Sven) are very prompt responders and are rather helpful, 
I told Bram to expect some questions from us :-)

In terms of VIM, here's what's left to be done:

 1. Find some decent ISO8859-6 Arabic fonts (position agnostic ?) just to test.
    --> [help, Chahine ?]
 2. Modify VIM code so that the R2L support is not noted strictly for hebrew, 
    but for Arabic/Farsi, etc. (currently the mode line assumes its hebrew,  
    plus there is a bunch of code that notes ONLY hebrew which needs to be 
    modified - I think mostly minor stuff).
    --> [I can look into this and talk to the maintainer]
 3. We need to add shaping code to the source for Arabic shaping - we'll
    need to talk to the VIM maintainers about their thoughts before starting
    anything so that we're all on the same footing.
    --> [help !!]
 4. We need to add support for Windows Arabic fonts (CP-1256).  Does anyone
    have their encodings and fonts ?
    --> [help !!]
 5. To my surprise, the microsoft keyboard notes english-letter-b as Arabic
    "LAM ALAF" (as in the arabic word "LAA" - english NO) which is not part
    of ISO-8859-6 encodings -- what can we do about that (can one key produce
    two consecutive codes so as to fake the key) ? Can we standardize on a
    keyboard for ALL of our projects..
    --> [help !!]

(*NOTE: for the record, TAKAHASHI Naoto <ntakahas at m17n dot org> has noted as of
 17.July.2001 that he's reworked the Multibyte support improving and
 correcting Arabic support which ought to be incorporated into emacs-21.2
 and he also promised a patch for emacs-21.1 as soon as he gets a chance -
 I'll post it to the mailing-list when I get it).

PS: please reply to the 'general' list :-)

 - Nadim