emacs-bidi
[Top][All Lists]
Advanced

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

Re: [emacs-bidi] Thank you Eli for your work on bidi!


From: Eli Zaretskii
Subject: Re: [emacs-bidi] Thank you Eli for your work on bidi!
Date: Sat, 24 Jul 2010 12:56:51 +0300

> Date: Wed, 07 Jul 2010 14:40:46 +0300
> From: Eli Zaretskii <address@hidden>
> Cc: address@hidden, address@hidden
> 
> > I can give a hand, What else do you think is missing at the lisp level?
> 
> Thanks.
> 
> I don't have a full list or a detailed plan.  What's below is just a
> knee-jerk reaction, plus bits and pieces I gathered from my notes.
> Note that most of the suff needs at least some discussion and then
> design, before it can be coded.
> 
> For now, there's only one feature implemented on this level: the
> effect of left and right arrow keys that is sensitive to the current
> paragraph direction.  But this is just a bare minimum.
> 
> With those caveats out of my way, here's what I've got, in no
> particular order:

Here's one more:

 . Provide a command to convert a region of text from English to a
   non-English script, using input method key remapping.

Explanation: I guess each one of us had an annoying experience whereby
we forget to switch on the input method, and get a bunch of gibberish
with ASCII letters instead of some script we wanted to type.  Then you
need to remove everything you typed and retype it again after
switching on the input method.

The command I have in mind would prompt for an input method (using the
current default one as the default value), and then convert every
ASCII character in the region to the corresponding character that
would have been inserted had the input method been switched on.

This, of course, is useful for non-bidi input methods as well.

Note that, with some input methods, a key can start a key sequence
that is translated as a whole.  So this command would need to add
characters to the sequence until a complete translatable sequence is
found.  It could ask the user for guidance, if this algorithm is not
guaranteed to produce the correct result.

Any takers?



reply via email to

[Prev in Thread] Current Thread [Next in Thread]