procyberian/emacs
2
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

(indian-font-foundry)

Browse source 
(indian-script-language-alist, indian-font-char-index-table)
(indian-font-char, indian-font-char-range, indian-script-table)
(indian-default-script, indian-composable-pattern): Delete them.
("Devanagari", "Kannada", "Malayalam", "Tamil"): Definitions of
language environments moved to here.
("Bengali", "Punjabi", "Gujarati", "Oriya", "Telugu"): New
language environments.
This commit is contained in:
Kenichi Handa 2008-02-22 01:49:10 +00:00
parent cc8d131137
commit cdbe6a039d
1 changed files with 83 additions and 113 deletions
Download patch file Download diff file Expand all files Collapse all files

196
lisp/language/indian.el
View file

@ -1,4 +1,4 @@
;;; indian.el --- Indian languages support -*- coding: iso-2022-7bit; -*-
;;; indian.el --- Indian languages support -*- coding: utf-8; -*-
;; Copyright (C) 1997, 1999, 2001, 2002, 2003, 2004, 2005, 2006, 2007, 2008
;; Free Software Foundation, Inc.
@ -6,7 +6,8 @@
;; National Institute of Advanced Industrial Science and Technology (AIST)
;; Registration Number H14PRO021
;; Maintainer: KAWABATA, Taichi <kawabata@m17n.org>
;; Maintainer: Kenichi Handa <handa@m17n.org>
;; KAWABATA, Taichi <kawabata@m17n.org>
;; Keywords: multilingual, i18n, Indian
;; This file is part of GNU Emacs.
@ -28,8 +29,8 @@
;;; Commentary:
;; This file defines in-is13194 coding system and relationship between
;; indian-glyph character-set and various CDAC fonts.
;; This file contains definitions of Indian language environments, and
;; setups for displaying the scrtipts used there.
;;; Code:
@ -44,120 +45,91 @@
(define-coding-system-alias 'devanagari 'in-is13194-devanagari)
(defvar indian-font-foundry 'cdac
"Font foundry for Indian characters.
Currently supported foundries are `cdac' and `akruti'.")
(set-language-info-alist
"Devanagari" '((charset unicode)
(coding-system utf-8)
(coding-priority utf-8)
(input-method . "dev-aiba")
(documentation . "\
Such languages using Devanagari script as Hindi and Marathi
are supported in this language environment."))
'("Indian"))
(defvar indian-script-language-alist
'((devanagari (hindi sanskrit) nil)
(bengali (bengali assamese) nil)
(gurmukhi (punjabi) nil)
(gujarati (gujarati) nil)
(oriya (oriya) nil)
(tamil (tamil) nil)
(telugu (telugu) nil)
(kannada (kannada) nil)
(malayalam (malayalam) nil))
"Alist of Indian scripts vs the corresponding language list and font foundry.
Each element has this form:
(set-language-info-alist
"Bengali" '((charset unicode)
(coding-system utf-8)
(coding-priority utf-8)
(input-method . "bengali-itrans")
(documentation . "\
Such languages using Bengali script as Bengali and Assamese
are supported in this language environment."))
'("Indian"))
(SCRIPT LANGUAGE-LIST FONT-FOUNDRY)
(set-language-info-alist
"Punjabi" '((charset unicode)
(coding-system utf-8)
(coding-priority utf-8)
(input-method . "punjabi-itrans")
(documentation . "\
North Indian language Punjabi is supported in this language environment."))
'("Indian"))
SCRIPT is one of Indian script names.
(set-language-info-alist
"Gujarati" '((charset unicode)
(coding-system utf-8)
(coding-priority utf-8)
(input-method . "gujarati-itrans")
(documentation . "\
North Indian language Gujarati is supported in this language environment."))
'("Indian"))
LANGUAGE-LIST is a list of Indian langauge names SCRIPT is used for.
The list is in the priority order.
(set-language-info-alist
"Oriya" '((charset unicode)
(coding-system utf-8)
(coding-priority utf-8)
(input-method . "oriya-itrans")
(documentation . "\
Such languages using Oriya script as Oriya, Khonti, and Santali
are supported in this language environment."))
'("Indian"))
FONT-FOUNDRY is a font foundry representing a group of Indian
fonts. If the value is nil, the value of `indian-font-foundry'
is used.")
(set-language-info-alist
"Tamil" '((charset unicode)
(coding-system utf-8)
(coding-priority utf-8)
(input-method . "tamil-itrans")
(documentation . "\
South Indian Language Tamil is supported in this language environment."))
'("Indian"))
(defconst indian-font-char-index-table
'( ; for which language(s)
;; CDAC fonts
(#x0000 . cdac:dv-ttsurekh) ; hindi, etc
(#x0100 . cdac:sd-ttsurekh) ; sanskrit
(#x0200 . cdac:bn-ttdurga) ; bengali
(#x0300 . cdac:tm-ttvalluvar) ; tamil
(#x0400 . cdac:tl-tthemalatha) ; telugu
(#x0500 . cdac:as-ttdurga) ; assamese
(#x0600 . cdac:or-ttsarala) ; oriya
(#x0700 . cdac:kn-ttuma) ; kannada
(#x0800 . cdac:ml-ttkarthika) ; malayalam
(#x0900 . cdac:gj-ttavantika) ; gujarati
(#x0A00 . cdac:pn-ttamar) ; punjabi
(set-language-info-alist
"Telugu" '((charset unicode)
(coding-system utf-8)
(coding-priority utf-8)
(input-method . "telugu-itrans")
(documentation . "\
South Indian Language Telugu is supported in this language environment."))
'("Indian"))
;; AKRUTI fonts
(#x0B00 . akruti:dev) ; hindi, etc
(#x0C00 . akruti:bng) ; bengali
(#x0D00 . akruti:pnj) ; punjabi
(#x0E00 . akruti:guj) ; gujarati
(#x0F00 . akruti:ori) ; oriya
(#x1000 . akruti:tml) ; tamil
(#x1100 . akruti:tlg) ; telugu
(#x1200 . akruti:knd) ; kannada
(#x1300 . akruti:mal) ; malayalam
)
"Alist of indices of `indian-glyph' character vs Indian font identifiers.
Each element has this form: (INDEX . FONT-IDENTIFIER)
(set-language-info-alist
"Kannada" '((charset unicode)
(coding-system mule-utf-8)
(coding-priority mule-utf-8)
(input-method . "kannada-itrans")
(sample-text . "Kannada (ಕನ್ನಡ) ನಮಸ್ಕಾರ")
(documentation . "\
Kannada language and script is supported in this language
environment."))
'("Indian"))
INDEX is an index number of the first character in the charset
`indian-glyph' assigned for glyphs in the font specified by
FONT-IDENTIFIER. Currently FONT-IDENTIFIERs are defined for CDAC
and AKRUTI font groups.")
(defun indian-font-char (index font-identifier)
"Return character of charset `indian-glyph' made from glyph index INDEX.
FONT-IDENTIFIER is an identifier of an Indian font listed in the
variable `indian-font-char-index-table'. It specifies which
font INDEX is for."
(if (or (< index 0) (> index 255))
(error "Invalid glyph index: %d" index))
(let ((start (car (rassq font-identifier indian-font-char-index-table))))
(if (not start)
(error "Unknown font identifier: %s" font-identifier))
(setq index (+ start index))
(make-char 'indian-glyph (+ (/ index 96) 32) (+ (% index 96) 32))))
;; Return a range of characters (cons of min and max character) of the
;; charset `indian-glyph' for displaying SCRIPT in LANGUAGE by a font
;; of FOUNDRY.
(defun indian-font-char-range (font-identifier)
(cons (indian-font-char 0 font-identifier)
(indian-font-char 255 font-identifier)))
(defvar indian-script-table
'[
devanagari
sanskrit
bengali
tamil
telugu
assamese
oriya
kannada
malayalam
gujarati
punjabi
]
"Vector of Indian script names.")
(let ((len (length indian-script-table))
(i 0))
(while (< i len)
(put (aref indian-script-table i) 'indian-glyph-code-offset (* 256 i))
(setq i (1+ i))))
(defvar indian-default-script 'devanagari
"Default script for Indian languages.
Each Indian language environment sets this value
to one of `indian-script-table' (which see).
The default value is `devanagari'.")
(defvar indian-composable-pattern
(make-char-table nil)
"Char table of regexps for composable Indian character sequence.")
(set-language-info-alist
"Malayalam" '((charset unicode)
(coding-system utf-8)
(coding-priority utf-8)
(input-method . "malayalam-itrans")
(documentation . "\
South Indian language Malayalam is supported in this language environment."))
'("Indian"))
(let ((script-regexp-alist
'((devanagari . "[\x900-\x9FF\x200C\x200D]+")
@ -168,8 +140,7 @@ The default value is `devanagari'.")
(tamil . "[\xB80-\xBFF\x200C\x200D]+")
(telugu . "[\xC00-\xC7F\x200C\x200D]+")
(kannada . "[\xC80-\xCFF\x200C\x200D]+")
(malayalam . "[\xD00-\xD7F\x200C\x200D]+")
(sinhala . "[\xD80-\xDFF\x200C\x200D]+"))))
(malayalam . "[\xD00-\xD7F\x200C\x200D]+"))))
(map-char-table #'(lambda (key val)
(let ((slot (assq val script-regexp-alist)))
(if slot
@ -177,7 +148,6 @@ The default value is `devanagari'.")
composition-function-table key
(list (cons (cdr slot) 'font-shape-text))))))
char-script-table))
(provide 'indian)