Re: [emacs-bidi] Arabic in Unicode

[TAKAHASHI Naoto]

Eli Zaretskii writes:

> So that means we cannot support Arabic presentation forms in the current 
> Emacs except in the mule-unicode-* character sets, is that correct?  

That is correct.

> Because IIRC, the Mule Arabic charset is based on iso8859-6, right?

No.  Although Mule preserves a leading-code for iso8859-6, it is not
used at all.  The Arabic support is based on private characeter sets;
they are completely different from iso8859-6.

>> To release the modification, we need to get an official permission
>> from the responsible section of our institute.  A necessary procedure
>> has just begun, but nobody knows how long it will take.

> Perhaps in the meantime, you could publish a textual description of what 
> the code does.  Then someone else could implement that without bumping 
> into copyright issues.

For bidi support, there is not so much to describe.  Basically, Handa
ported the bidi display routines in Mule-2.3 to Emacs-21.0.4.  He did
it in a day or two.  (But you know, it is Handa. :-)

The font-lock like mechanism and others were written to select and
display the appropriate glyph when a basic codepoint and the textual
information (word beginning, middle, end, etc.) are given.  So they
concern only Arabic and Indian scripts.

-- 
TAKAHASHI Naoto
address@hidden
http://www.m17n.org/ntakahas/