lisp/language/thai-util.el

   1 ;;; thai-util.el --- utilities for Thai -*- coding: utf-8; -*-
   2
   3 ;; Copyright (C) 1995, 1996, 1997, 1998, 1999, 2000, 2001, 2002, 2003, 2004,
   4 ;;   2005, 2006, 2007, 2008, 2009, 2010
   5 ;;   National Institute of Advanced Industrial Science and Technology (AIST)
   6 ;;   Registration Number H14PRO021
   7 ;; Copyright (C) 2000, 2001, 2002, 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010
   8 ;;   Free Software Foundation, Inc.
   9
  10 ;; Keywords: mule, multilingual, Thai, i18n
  11
  12 ;; This file is part of GNU Emacs.
  13
  14 ;; GNU Emacs is free software: you can redistribute it and/or modify
  15 ;; it under the terms of the GNU General Public License as published by
  16 ;; the Free Software Foundation, either version 3 of the License, or
  17 ;; (at your option) any later version.
  18
  19 ;; GNU Emacs is distributed in the hope that it will be useful,
  20 ;; but WITHOUT ANY WARRANTY; without even the implied warranty of
  21 ;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  22 ;; GNU General Public License for more details.
  23
  24 ;; You should have received a copy of the GNU General Public License
  25 ;; along with GNU Emacs.  If not, see <http://www.gnu.org/licenses/>.
  26
  27 ;;; Commentary:
  28
  29 ;;; Code:
  30
  31 (defvar thai-auto-composition-mode)
  32
  33 ;; Setting information of Thai characters.
  34
  35 (defconst thai-category-table (make-category-table))
  36 (define-category ?c "Thai consonant" thai-category-table)
  37 (define-category ?v "Thai upper/lower vowel" thai-category-table)
  38 (define-category ?t "Thai tone mark" thai-category-table)
  39 (define-category ?u "Thai tone mark and upper sign" thai-category-table)
  40 (define-category ?I "THAI CHARACTER SARA I" thai-category-table)
  41 (define-category ?U "THAI CHARACTER THANTHAKHAT" thai-category-table)
  42
  43 ;; The general composing rules are as follows:
  44 ;;
  45 ;;                          T
  46 ;;       V        U         V                  U
  47 ;; CV -> C, CU -> C, CVT -> C, Cv -> C, CvU -> C
  48 ;;                                   v         v
  49 ;;
  50 ;; where C: consonant, V: vowel upper, v: vowel lower,
  51 ;;       T: tone mark, U: tone mark and upper sign.
  52 ;; Special rule: The sign `์' can be put on the vowel `ิ'.
  53
  54
  55 (defvar thai-composition-pattern
  56   "\\cc\\(\\cu\\|\\cI\\cU\\|\\cv\\ct?\\)\\|\\cv\\ct\\|\\cI\\cU"
  57   "Regular expression matching a Thai composite sequence.")
  58
  59 (let ((l '((?ก consonant "LETTER KO KAI")                             ; 0xA1
  60            (?ข consonant "LETTER KHO KHAI")                           ; 0xA2
  61            (?ฃ consonant "LETTER KHO KHUAT")                          ; 0xA3
  62            (?ค consonant "LETTER KHO KHWAI")                          ; 0xA4
  63            (?ฅ consonant "LETTER KHO KHON")                           ; 0xA5
  64            (?ฆ consonant "LETTER KHO RAKHANG")                                ; 0xA6
  65            (?ง consonant "LETTER NGO NGU")                            ; 0xA7
  66            (?จ consonant "LETTER CHO CHAN")                           ; 0xA8
  67            (?ฉ consonant "LETTER CHO CHING")                          ; 0xA9
  68            (?ช consonant "LETTER CHO CHANG")                          ; 0xAA
  69            (?ซ consonant "LETTER SO SO")                              ; 0xAB
  70            (?ฌ consonant "LETTER CHO CHOE")                           ; 0xAC
  71            (?ญ consonant "LETTER YO YING")                            ; 0xAD
  72            (?ฎ consonant "LETTER DO CHADA")                           ; 0xAE
  73            (?ฏ consonant "LETTER TO PATAK")                           ; 0xAF
  74            (?ฐ consonant "LETTER THO THAN")                           ; 0xB0
  75            (?ฑ consonant "LETTER THO NANGMONTHO")                     ; 0xB1
  76            (?ฒ consonant "LETTER THO PHUTHAO")                                ; 0xB2
  77            (?ณ consonant "LETTER NO NEN")                             ; 0xB3
  78            (?ด consonant "LETTER DO DEK")                             ; 0xB4
  79            (?ต consonant "LETTER TO TAO")                             ; 0xB5
  80            (?ถ consonant "LETTER THO THUNG")                          ; 0xB6
  81            (?ท consonant "LETTER THO THAHAN")                         ; 0xB7
  82            (?ธ consonant "LETTER THO THONG")                          ; 0xB8
  83            (?น consonant "LETTER NO NU")                              ; 0xB9
  84            (?บ consonant "LETTER BO BAIMAI")                          ; 0xBA
  85            (?ป consonant "LETTER PO PLA")                             ; 0xBB
  86            (?ผ consonant "LETTER PHO PHUNG")                          ; 0xBC
  87            (?ฝ consonant "LETTER FO FA")                              ; 0xBD
  88            (?พ consonant "LETTER PHO PHAN")                           ; 0xBE
  89            (?ฟ consonant "LETTER FO FAN")                             ; 0xBF
  90            (?ภ consonant "LETTER PHO SAMPHAO")                                ; 0xC0
  91            (?ม consonant "LETTER MO MA")                              ; 0xC1
  92            (?ย consonant "LETTER YO YAK")                             ; 0xC2
  93            (?ร consonant "LETTER RO RUA")                             ; 0xC3
  94            (?ฤ vowel-base "LETTER RU (Pali vowel letter)")            ; 0xC4
  95            (?ล consonant "LETTER LO LING")                            ; 0xC5
  96            (?ฦ vowel-base "LETTER LU (Pali vowel letter)")            ; 0xC6
  97            (?ว consonant "LETTER WO WAEN")                            ; 0xC7
  98            (?ศ consonant "LETTER SO SALA")                            ; 0xC8
  99            (?ษ consonant "LETTER SO RUSI")                            ; 0xC9
 100            (?ส consonant "LETTER SO SUA")                             ; 0xCA
 101            (?ห consonant "LETTER HO HIP")                             ; 0xCB
 102            (?ฬ consonant "LETTER LO CHULA")                           ; 0xCC
 103            (?อ consonant "LETTER O ANG")                              ; 0xCD
 104            (?ฮ consonant "LETTER HO NOK HUK")                         ; 0xCE
 105            (?ฯ special "PAI YAN NOI (abbreviation)")                  ; 0xCF
 106            (?ะ vowel-base "VOWEL SIGN SARA A")                                ; 0xD0
 107            (?ั vowel-upper "VOWEL SIGN MAI HAN-AKAT N/S-T")           ; 0xD1
 108            (?า vowel-base "VOWEL SIGN SARA AA")                               ; 0xD2
 109            (?ำ vowel-base "VOWEL SIGN SARA AM")                               ; 0xD3
 110            (?ิ vowel-upper "VOWEL SIGN SARA I N/S-T")                 ; 0xD4
 111            (?ี vowel-upper "VOWEL SIGN SARA II N/S-T")                        ; 0xD5
 112            (?ึ vowel-upper "VOWEL SIGN SARA UE N/S-T")                        ; 0xD6
 113            (?ื vowel-upper "VOWEL SIGN SARA UEE N/S-T")                       ; 0xD7
 114            (?ุ vowel-lower "VOWEL SIGN SARA U N/S-B")                 ; 0xD8
 115            (?ู vowel-lower "VOWEL SIGN SARA UU N/S-B")                        ; 0xD9
 116            (?ฺ vowel-lower "VOWEL SIGN PHINTHU N/S-B (Pali virama)")  ; 0xDA
 117            (?฻ invalid nil)                                           ; 0xDA
 118            (?฼ invalid nil)                                           ; 0xDC
 119            (?฽ invalid nil)                                           ; 0xDC
 120            (?฾ invalid nil)                                           ; 0xDC
 121            (?฿ special "BAHT SIGN (currency symbol)")                 ; 0xDF
 122            (?เ vowel-base "VOWEL SIGN SARA E")                                ; 0xE0
 123            (?แ vowel-base "VOWEL SIGN SARA AE")                               ; 0xE1
 124            (?โ vowel-base "VOWEL SIGN SARA O")                                ; 0xE2
 125            (?ใ vowel-base "VOWEL SIGN SARA MAI MUAN")                 ; 0xE3
 126            (?ไ vowel-base "VOWEL SIGN SARA MAI MALAI")                        ; 0xE4
 127            (?ๅ vowel-base "LAK KHANG YAO")                            ; 0xE5
 128            (?ๆ special "MAI YAMOK (repetion)")                                ; 0xE6
 129            (?็ sign-upper "VOWEL SIGN MAI TAI KHU N/S-T")             ; 0xE7
 130            (?่ tone "TONE MAI EK N/S-T")                              ; 0xE8
 131            (?้ tone "TONE MAI THO N/S-T")                             ; 0xE9
 132            (?๊ tone "TONE MAI TRI N/S-T")                             ; 0xEA
 133            (?๋ tone "TONE MAI CHATTAWA N/S-T")                                ; 0xEB
 134            (?์ sign-upper "THANTHAKHAT N/S-T (cancellation mark)")    ; 0xEC
 135            (?ํ sign-upper "NIKKHAHIT N/S-T (final nasal)")            ; 0xED
 136            (?๎ sign-upper "YAMAKKAN N/S-T")                           ; 0xEE
 137            (?๏ special "FONRMAN")                                     ; 0xEF
 138            (?๐ special "DIGIT ZERO")                                  ; 0xF0
 139            (?๑ special "DIGIT ONE")                                   ; 0xF1
 140            (?๒ special "DIGIT TWO")                                   ; 0xF2
 141            (?๓ special "DIGIT THREE")                                 ; 0xF3
 142            (?๔ special "DIGIT FOUR")                                  ; 0xF4
 143            (?๕ special "DIGIT FIVE")                                  ; 0xF5
 144            (?๖ special "DIGIT SIX")                                   ; 0xF6
 145            (?๗ special "DIGIT SEVEN")                                 ; 0xF7
 146            (?๘ special "DIGIT EIGHT")                                 ; 0xF8
 147            (?๙ special "DIGIT NINE")                                  ; 0xF9
 148            (?๚ special "ANGKHANKHU (ellipsis)")                               ; 0xFA
 149            (?๛ special "KHOMUT (beginning of religious texts)")               ; 0xFB
 150            (?๜ invalid nil)                                           ; 0xFC
 151            (?๝ invalid nil)                                           ; 0xFD
 152            (?๞ invalid nil)                                           ; 0xFE
 153            ))
 154       elm)
 155   (while l
 156     (setq elm (car l) l (cdr l))
 157     (let ((char (car elm))
 158           (ptype (nth 1 elm)))
 159       (put-char-code-property char 'phonetic-type ptype)
 160       (cond ((eq ptype 'consonant)
 161              (modify-category-entry char ?c thai-category-table))
 162             ((memq ptype '(vowel-upper vowel-lower))
 163              (modify-category-entry char ?v thai-category-table)
 164              (if (= char ?ิ)
 165                  ;; Give category `I' to "SARA I".
 166                  (modify-category-entry char ?I thai-category-table)))
 167             ((eq ptype 'tone)
 168              (modify-category-entry char ?t thai-category-table)
 169              (modify-category-entry char ?u thai-category-table))
 170             ((eq ptype 'sign-upper)
 171              (modify-category-entry char ?u thai-category-table)
 172              (if (= char ?์)
 173                  ;; Give category `U' to "THANTHAKHAT".
 174                  (modify-category-entry char ?U thai-category-table))))
 175       (put-char-code-property char 'name (nth 2 elm)))))
 176
 177 (defun thai-compose-syllable (beg end &optional category-set string)
 178   (or category-set
 179       (setq category-set
 180             (char-category-set (if string (aref string beg) (char-after beg)))))
 181   (if (aref category-set ?c)
 182       ;; Starting with a consonant.  We do relative composition.
 183       (if string
 184           (compose-string string beg end)
 185         (compose-region beg end))
 186     ;; Vowel tone sequence.
 187     (if string
 188         (compose-string string beg end (list (aref string beg) '(Bc . Bc)
 189                                              (aref string (1+ beg))))
 190       (compose-region beg end (list (char-after beg) '(Bc . Bc)
 191                                     (char-after (1+ beg))))))
 192   (- end beg))
 193
 194 ;;;###autoload
 195 (defun thai-compose-region (beg end)
 196   "Compose Thai characters in the region.
 197 When called from a program, expects two arguments,
 198 positions (integers or markers) specifying the region."
 199   (interactive "r")
 200   (let ((pos (point)))
 201     (save-restriction
 202       (narrow-to-region beg end)
 203       (goto-char (point-min))
 204       (with-category-table thai-category-table
 205         (while (re-search-forward thai-composition-pattern nil t)
 206           (setq beg (match-beginning 0) end (match-end 0))
 207           (if (and (> pos beg) (< pos end))
 208               (setq pos end))
 209           (thai-compose-syllable beg end
 210                                  (char-category-set (char-after beg))))))
 211     (goto-char pos)))
 212
 213 ;;;###autoload
 214 (defun thai-compose-string (string)
 215   "Compose Thai characters in STRING and return the resulting string."
 216   (with-category-table thai-category-table
 217     (let ((idx 0))
 218       (while (setq idx (string-match thai-composition-pattern string idx))
 219         (thai-compose-syllable idx (match-end 0) nil string)
 220         (setq idx (match-end 0)))))
 221   string)
 222
 223 ;;;###autoload
 224 (defun thai-compose-buffer ()
 225   "Compose Thai characters in the current buffer."
 226   (interactive)
 227   (thai-compose-region (point-min) (point-max)))
 228
 229 ;;;###autoload
 230 (defun thai-composition-function (gstring)
 231   (if (= (lgstring-char-len gstring) 1)
 232       (compose-gstring-for-graphic gstring)
 233     (or (font-shape-gstring gstring)
 234         (let ((glyph-len (lgstring-glyph-len gstring))
 235               (last-char (lgstring-char gstring
 236                                         (1- (lgstring-char-len gstring))))
 237               (i 0)
 238               glyph)
 239           (while (and (< i glyph-len)
 240                       (setq glyph (lgstring-glyph gstring i)))
 241             (setq i (1+ i)))
 242           (if (= last-char ?ำ)
 243               (setq i (1- i)))
 244           (compose-glyph-string-relative gstring 0 i 0.1)))))
 245
 246 ;; Thai-word-mode requires functions in the feature `thai-word'.
 247 (require 'thai-word)
 248
 249 (defvar thai-word-mode-map
 250   (let ((map (make-sparse-keymap)))
 251     (define-key map [remap forward-word] 'thai-forward-word)
 252     (define-key map [remap backward-word] 'thai-backward-word)
 253     (define-key map [remap kill-word] 'thai-kill-word)
 254     (define-key map [remap backward-kill-word] 'thai-backward-kill-word)
 255     (define-key map [remap transpose-words] 'thai-transpose-words)
 256     map)
 257   "Keymap for `thai-word-mode'.")
 258
 259 (define-minor-mode thai-word-mode
 260   "Minor mode to make word-oriented commands aware of Thai words.
 261 The commands affected are \\[forward-word], \\[backward-word], \\[kill-word], \\[backward-kill-word], \\[transpose-words], and \\[fill-paragraph]."
 262   :global t :group 'mule
 263   (cond (thai-word-mode
 264          ;; This enables linebreak between Thai characters.
 265          (modify-category-entry (make-char 'thai-tis620) ?|)
 266          ;; This enables linebreak at a Thai word boundary.
 267          (put-charset-property 'thai-tis620 'fill-find-break-point-function
 268                                'thai-fill-find-break-point))
 269         (t
 270          (modify-category-entry (make-char 'thai-tis620) ?| nil t)
 271          (put-charset-property 'thai-tis620 'fill-find-break-point-function
 272                                nil))))
 273
 274 ;; Function to call on entering the Thai language environment.
 275 (defun setup-thai-language-environment-internal ()
 276   (thai-word-mode 1))
 277
 278 ;; Function to call on exiting the Thai language environment.
 279 (defun exit-thai-language-environment-internal ()
 280   (thai-word-mode -1))
 281
 282 ;;
 283 (provide 'thai-util)
 284
 285 ;; arch-tag: 59425d6a-8cf9-4e06-a6ab-8ab7dc7a7a97
 286 ;;; thai-util.el ends here