emacs and indian scripts

[François Patte <francois.patte@xxxxxxxxxxxxxxxxxxxxxxxx>]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Bonjour,

I have a problem with mule: I want to build multilingual texts to be
processed by xelatex.

I want 3 wrintings (+ latin): devanagari, bengali, tamil. I will
describe the problem for the two first, for I think that if someone
helps me to solve the bengali problem, the solution will apply to tamil.

Firstly, I had some problems with devanagari which has been solved
installing cdac fonts; intlfonts are of no use and create problems for
indian scripts as far as I could see....

So exit devanagari from the list of problems but, I keep it for comparison.

One can read in
~ /usr/share/emacs/22.1/lisp/international/fontset.el.gz:

~             ;; Indian CDAC
~             (,(indian-font-char-range 'cdac:dv-ttsurekh)
~              . (nil . "Devanagari-CDAC"))
~             (,(indian-font-char-range 'cdac:sd-ttsurekh)
~              . (nil . "Sanskrit-CDAC"))
~             (,(indian-font-char-range 'cdac:bn-ttdurga)
~              . (nil . "Bengali-CDAC"))

And in
/usr/share/emacs/22.1/leim/quail/indian.el.gz:

(if nil
~    (quail-define-package "devanagari-itrans" "Devanagari" "DevIT" t
"Devanagari ITRANS"))
(quail-define-indian-trans-package
~ indian-dev-itrans-v5-hash "devanagari-itrans" "Devanagari" "DevIT"
~ "Devanagari transliteration by ITRANS method.")

(if nil
~    (quail-define-package "bengali-itrans" "Bengali" "BngIT" t "Bengali
ITRANS"))
(quail-define-indian-trans-package
~ indian-bng-itrans-v5-hash "bengali-itrans" "Bengali" "BngIT"
~ "Bengali transliteration by ITRANS method.")



So, if I select devanagari-itrans input method, the devanagari cdac
fonts are automatically loaded, and I can display devanagari char in my
emacs window. Using describe-char fonction on a devanagari char, I get:

~  character: अ (334533, #o1215305, #x51ac5, U+0905)
~    charset: mule-unicode-0100-24ff
~         (Unicode characters of the range U+0100..U+24FF.)
~ code point: #x35 #x45
~     syntax: w     which means: word
~   category: 1:base (independent) vowel i:Indian
~ Properties: script: devanagari; iscii: अ;
~   to input: type "a" with devanagari-itrans
buffer code: #x9C #xF4 #xB5 #xC5
~  file code: #xE0 #xA4 #x85 (encoded by coding system mule-utf-8)
~    display: composed to form "अ" (see below)

Composed by the rule:
The component character(s) are displayed by these fonts (glyph codes):
~ �: -FreeType-DV
TTSurekh-Medium-R-Normal--33-240-100-100-P-119-Devanagari-CDAC (#x2B)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

That is exactly what is expected. But, if I select bengali-itrans input
method, I cannot get any display of the bengali glyph in emacs (only an
empty rectangle) and describe-char gives:

~  character: অ (334693, #o1215545, #x51b65, U+0985)
~    charset: mule-unicode-0100-24ff
~         (Unicode characters of the range U+0100..U+24FF.)
~ code point: #x36 #x65
~     syntax: w     which means: word
~   category: 1:base (independent) vowel
~   to input: type "a" with bengali-itrans
buffer code: #x9C #xF4 #xB6 #xE5
~  file code: #xE0 #xA6 #x85 (encoded by coding system mule-utf-8)
~    display: by this font (glyph code)
~     -Misc-Fixed-Medium-R-Normal--24-240-72-72-C-120-iso10646.indian-1
(#x985)

Last line shows that, in that case, cdac fonts are not loaded.


My question: howto have cdac fonts loaded in any cas when I select an
indian script input method?


Thanks for help and sorry for a long post.


- --
François Patte
UFR de mathématiques et informatique
Université Paris Descartes
45, rue des Saints Pères
F-75270 Paris Cedex 06
Tél. +33 (0)1 44 55 35 61
http://www.math-info.univ-paris5.fr/~patte
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.7 (GNU/Linux)
Comment: Using GnuPG with Fedora - http://enigmail.mozdev.org

iD8DBQFH6+hfdE6C2dhV2JURAtcYAJ0YMnRLeoZ25iSz7uoBpqW9Cvig+ACeJmIi
Xr6U2EeOOSJJux7EEbDIOgs=
=JjPf
-----END PGP SIGNATURE-----