[Emacs-diffs] emacs/lisp/language indian.el

[Kenichi Handa]

CVSROOT:        /cvsroot/emacs
Module name:    emacs
Changes by:     Kenichi Handa <handa>   09/12/12 02:17:48

Modified files:
        lisp/language  : indian.el 

Log message:
        (indian-compose-regexp): New function.
        (malayalam-composable-pattern): Fix the pattern.
        (composition-function-table): Set malayalam-composable-pattern for
        Malayalam characters.

CVSWeb URLs:
http://cvs.savannah.gnu.org/viewcvs/emacs/lisp/language/indian.el?cvsroot=emacs&r1=1.39&r2=1.40

Patches:
Index: indian.el
===================================================================
RCS file: /cvsroot/emacs/emacs/lisp/language/indian.el,v
retrieving revision 1.39
retrieving revision 1.40
diff -u -b -r1.39 -r1.40
--- indian.el   2 Dec 2009 07:59:51 -0000       1.39
+++ indian.el   12 Dec 2009 02:17:48 -0000      1.40
@@ -129,6 +129,15 @@
 South Indian language Malayalam is supported in this language environment."))
  '("Indian"))
 
+;; Replace mnemonic characters in REGEXP according to TABLE.  TABLE is
+;; an alist of (MNEMONIC-STRING . REPLACEMENT-STRING).
+
+(defun indian-compose-regexp (regexp table)
+  (let ((case-fold-search nil))
+    (dolist (elt table)
+      (setq regexp (replace-regexp-in-string (car elt) (cdr elt) regexp t t)))
+    regexp))
+
 (defconst devanagari-composable-pattern
   (concat
    "\\([अ-औॠॡ][ँं]?\\)\\|[ः।]"
@@ -156,12 +165,27 @@
   "Regexp matching a composable sequence of Kannada characters.")
 
 (defconst malayalam-composable-pattern
+  (let ((table '(("V" . "[\u0D05-\u0D14\u0D60-\u0D61]") ; independent vowel
+                ("C" . "[\u0D15-\u0D39]")              ; consonant 
+                ("m" . "[\u0D46-\u0D48\u0D4A-\u0D4C]") ; prebase matra
+                ("p" . "[\u0D3E-\u0D44\u0D57]") ; postname matra
+                ("b" . "[\u0D62-\u0D63]")       ; belowbase matra
+                ("a" . "[\u0D02-\u0D03]")       ; abovebase sign
+                ("H" . "്")                   ; virama sign 
+                ("N" . "\u200D")                ; ZWJ
+                ("J" . "\u200C")                ; ZWNJ
+                ("X" . "[\u0D00-\u0D7F]")))) ; all coverage
+    (indian-compose-regexp
   (concat
-   "\\([അ-ഔ][ം]?\\)\\|ഃ"
-   "\\|\\("
-   
"\\(?:\\(?:[ക-ഹ]്\\)?\\(?:[ക-ഹ]്\\)?\\(?:[ക-ഹ]്\\)?[ക-ഹ]്\\)?"
-   "[ക-ഹ]\\(?:്\\|[ാ-ൃെേൈൊൊോൌ]?[ം്]?\\)?"
-   "\\)")
+      ;; consonant-based syllables
+      "\\(CJ?HJ?\\)*C\\(H[NJ]?\\|m?b?p?a?\\)\\|"
+      ;; syllables with an independent vowel
+      "V\\(J?HC\\)?m?b?p?n?a?\\|"
+      ;; special consonant form
+      "JHC\\|"
+      ;; any other singleton characters
+      "X")
+     table))
   "Regexp matching a composable sequence of Malayalam characters.")
 
 (let ((script-regexp-alist
@@ -173,7 +197,7 @@
         (tamil . "[\xB80-\xBFF\x200C\x200D]+")
         (telugu . "[\xC00-\xC7F\x200C\x200D]+")
         (kannada . "[\xC80-\xCFF\x200C\x200D]+")
-        (malayalam . "[\xD00-\xD7F\x200C\x200D]+"))))
+        (malayalam . ,malayalam-composable-pattern))))
   (map-char-table
    #'(lambda (key val)
        (let ((slot (assq val script-regexp-alist)))