Re: [AUCTeX-devel] Bulgarian and math mode

[Georgi Boshnakov]

Quoting Ralf Angeli <address@hidden>:

> > I suggest that this be extended to Bulgarian and Russian. This is
> > certainly the right thing to do for Bulgarian and I am almost sure
> > about Russian.
>
> Wouldn't you nowadays want to use Cyrillic characters in your LaTeX
> source rather than the respective LaTeX commands?

This has always been the case.

In maths mode however cyrillic letters are not used. Therefore it is  
natural that switching to maths mode should also switch off  the  
cyrillic alphabet. This is exactly what happens when  one starts maths  
mode with (an ellectric) `$' but only for CJK languages.   Here is the  
place in tex.el I was referring to in my previous email.


(defun TeX-toggle-off-input-method ()
  "Toggle off input methods.
Currently, only support CJK input methods provided by LEIM package."
  (cond
   ;; LEIM Package
   ((and (boundp 'current-input-method) current-input-method
         (string-match "^\\(chinese\\|japanese\\|korean\\)" 
current-input-method))
    (inactivate-input-method))
   (t );; do nothing
   ))

To achieve the desired effect it is sufficient to change the above regexp to

"^\\(chinese\\|japanese\\|korean\\|bulgarian\\|russian\\)"

The user may independently disable the automatic switch-off of input  
method by using a variable defined earlier in tex.el, which is also  
used in the decision to switch off or not to switch off the input  
method. This is:

(defcustom TeX-math-toggle-off-input-method t
  "*If non-nil, auto toggle off CJK input methods when entering math mode."
  :group 'TeX-macro
  :type 'boolean)

The documentation of this variable may be changed accordingly to  
reflect other relevant languages.

The above change seems trivial but
if you need a volunteer to look into this more deeply, I will be happy  
to do it.

Thanks,
Georgi