Re: [emacs-bidi] Display routines

[Matan Ninio]

"Eli Zaretskii" <address@hidden> writes:

> > From: Michael Welsh Duggan <address@hidden>
> > Date: Mon, 05 Nov 2001 21:34:32 -0500
> > 
> > Would it make good sense to start work on modifying the display
> > routines to work with bidirectional text?
> 
> That's what I'm doing now.
> 
> > A prototype would be based on a embedded-level text property, which
> > would have a value consistant with that of UAX#9.
> 
> The current design does not call for text properties as the vehicle
> for storing the embedding level information.  This information is
> computed by the implementation of the UAX#9 algorithm, so it doesn't
> make much sense to store it as text properties.  Text properties are
> for something that is external to the text, and which cannot be
> computed from the text alone.
> 
> If I post the description of the design, this will become more
> clear.  If not, please ask specific questions later.
> 
> > I think this would be a good way to seperate work on the display
> > machanism itself from the actual handling of the texts themselves.
> 
> My work currently concentrates on the display; I think this is the
> first aspect of bidi support that should work.  It doesn't make sense
> to do anything else if we cannot even display the text.
> 
> However, perhaps it would be useful to have a list of tasks that are
> unrelated to display (what you seem to be calling ``the actual
> handling of the text'').  Could you suggest a list of these tasks?
> That list would probably include Emacs features which need to be
> modified to work correctly with bidirectional text, and missing Emacs
> features and infrastructure that needs to be added in order to
> support bidi.

here are some I can think of:
-> input and output for iso 8859-8 and 8859-8-i (visual and logical order
   Hebrew in the "European languages"/non Unicode standard. (and similarly
   8859-6 for Arabic).  (what is cp-1255 anyway?)
this needs to include some way to read and write such files, which are
common.  good examples can be found in Hebrew newspapers on the new
(www.haaretz.co.il for 8859-8 and www.ynet.co.il for 8859-8-i.  I'm
sorry, but I can't read Arabic or Persian, so I can't help much there)

-> printing of Hebrew files via ps-print or even print-buffer

-> auto detect of 8859-8 vise 8859-8-i
should probably be based on the frequency of pairs of letters.  I can
give you an algorithm that, given some body of texts, can give a
probability for each direction.  so if you open such a file, emacs could
either open in this way, the other way or, if it's not sure, ask the
user.

-> input keymaps and such for Hebrew input
including the two-way-cursor and the language swap (input and
paragraph's main direction)

-> visual2logical-region in case you pasted something in the wrong
   direction 

-> an algorithm for some support for "rectangle" in bidi texts
(may be a pain, that last one)

-> W3 mode ???

that's it for now, but there probably are many others (Word->displayable
form??)
                MN 


> 
> _______________________________________________
> emacs-bidi mailing list
> address@hidden
> http://mail.gnu.org/mailman/listinfo/emacs-bidi
> 

-- 
Matan Ninio             
--------------------------------+----------------------------------
Contrary to popular belief, Unix is user friendly.  It just happens
to be selective about who it makes friends with.  -- Dave Parnas