lisp/language/japanese.el

   1 ;;; japanese.el --- support for Japanese -*- coding: iso-2022-7bit; no-byte-compile: t -*-
   2
   3 ;; Copyright (C) 1997, 2001, 2002, 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010
   4 ;;   Free Software Foundation, Inc.
   5 ;; Copyright (C) 1995, 1996, 1997, 1998, 1999, 2000, 2001, 2002, 2003, 2004,
   6 ;;   2005, 2006, 2007, 2008, 2009, 2010
   7 ;;   National Institute of Advanced Industrial Science and Technology (AIST)
   8 ;;   Registration Number H14PRO021
   9 ;; Copyright (C) 2003
  10 ;;   National Institute of Advanced Industrial Science and Technology (AIST)
  11 ;;   Registration Number H13PRO009
  12
  13 ;; Keywords: multilingual, Japanese
  14
  15 ;; This file is part of GNU Emacs.
  16
  17 ;; GNU Emacs is free software: you can redistribute it and/or modify
  18 ;; it under the terms of the GNU General Public License as published by
  19 ;; the Free Software Foundation, either version 3 of the License, or
  20 ;; (at your option) any later version.
  21
  22 ;; GNU Emacs is distributed in the hope that it will be useful,
  23 ;; but WITHOUT ANY WARRANTY; without even the implied warranty of
  24 ;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  25 ;; GNU General Public License for more details.
  26
  27 ;; You should have received a copy of the GNU General Public License
  28 ;; along with GNU Emacs.  If not, see <http://www.gnu.org/licenses/>.
  29
  30 ;;; Commentary:
  31
  32 ;; For Japanese, character sets JISX0201, JISX0208, JISX0212 are
  33 ;; supported.
  34
  35 ;;; Code:
  36
  37 ;;; Load translation tables for CP932.
  38 (load "international/cp51932")
  39 (load "international/eucjp-ms")
  40
  41 (define-coding-system 'iso-2022-jp
  42   "ISO 2022 based 7bit encoding for Japanese (MIME:ISO-2022-JP)."
  43   :coding-type 'iso-2022
  44   :mnemonic ?J
  45   :designation [(ascii japanese-jisx0208-1978 japanese-jisx0208
  46                        latin-jisx0201)
  47                 nil nil nil]
  48   :flags '(short ascii-at-eol ascii-at-cntl 7-bit designation)
  49   :charset-list '(ascii japanese-jisx0208
  50                         japanese-jisx0208-1978 latin-jisx0201)
  51   :mime-charset 'iso-2022-jp
  52   :suitable-for-keyboard t)
  53
  54 (define-coding-system-alias 'junet 'iso-2022-jp)
  55
  56 (define-coding-system 'iso-2022-jp-2
  57   "ISO 2022 based 7bit encoding for CJK, Latin-1, Greek (MIME:ISO-2022-JP-2)."
  58   :coding-type 'iso-2022
  59   :mnemonic ?J
  60   :designation [(ascii japanese-jisx0208-1978 japanese-jisx0208
  61                        latin-jisx0201 japanese-jisx0212
  62                        chinese-gb2312 korean-ksc5601)
  63                 nil
  64                 (nil latin-iso8859-1 greek-iso8859-7)
  65                 nil]
  66   :flags '(short ascii-at-eol ascii-at-cntl 7-bit designation single-shift
  67                  init-at-bol)
  68   :charset-list '(ascii japanese-jisx0208 japanese-jisx0212
  69                         latin-jisx0201 japanese-jisx0208-1978
  70                         chinese-gb2312 korean-ksc5601
  71                         latin-iso8859-1 greek-iso8859-7)
  72   :mime-charset 'iso-2022-jp-2
  73   :suitable-for-keyboard t)
  74
  75 (let ((map                      ; JIS           vs      CP932
  76        '((#x301C . #xFF5E)      ; WAVE DASH             FULLWIDTH TILDE
  77          (#x2014 . #x2015)      ; EM DASH               HORIZONTAL BAR
  78          (#x2016 . #x2225)      ; DOUBLE VERTICAL LINE  PARALLEL TO
  79          (#x2212 . #xFF0D)      ; MINUS SIGN            FULLWIDTH HYPHEN-MINUS
  80          (#x00A2 . #xFFE0)      ; CENT SIGN             FULLWIDTH CENT SIGN
  81          (#x00A3 . #xFFE1)      ; POUND SIGN            FULLWIDTH POUND SIGN
  82          (#x00AC . #xFFE2)      ; NOT SIGN              FULLWIDTH NOT SIGN
  83          (#x00A6 . #xFFE4)      ; BROKEN LINE           FULLWIDTH BROKEN LINE
  84          )))
  85   (define-translation-table 'japanese-ucs-jis-to-cp932-map map)
  86   (mapc #'(lambda (x) (let ((tmp (car x)))
  87                         (setcar x (cdr x)) (setcdr x tmp)))
  88         map)
  89   (define-translation-table 'japanese-ucs-cp932-to-jis-map map))
  90
  91 ;; U+2014 (EM DASH) vs U+2015 (HORIZONTAL BAR)
  92 (define-translation-table 'japanese-ucs-glibc-to-jis-map '((#x2015 . #x2014)))
  93 (define-translation-table 'japanese-ucs-jis-to-glibc-map '((#x2014 . #x2015)))
  94
  95 (define-coding-system 'japanese-shift-jis
  96   "Shift-JIS 8-bit encoding for Japanese (MIME:SHIFT_JIS)"
  97   :coding-type 'shift-jis
  98   :mnemonic ?S
  99   :charset-list '(ascii katakana-jisx0201 japanese-jisx0208)
 100   :mime-charset 'shift_jis)
 101
 102 (define-coding-system-alias 'shift_jis 'japanese-shift-jis)
 103 (define-coding-system-alias 'sjis 'japanese-shift-jis)
 104
 105 (define-coding-system 'japanese-cp932
 106   "CP932 (Microsoft shift-jis)"
 107   :coding-type 'charset
 108   :mnemonic ?S
 109   :charset-list '(ascii katakana-sjis cp932-2-byte))
 110
 111 (define-coding-system-alias 'cp932 'japanese-cp932)
 112
 113 (define-coding-system 'japanese-iso-7bit-1978-irv
 114   "ISO 2022 based 7-bit encoding for Japanese JISX0208-1978 and JISX0201-Roman."
 115   :coding-type 'iso-2022
 116   :mnemonic ?j
 117   :designation [(latin-jisx0201 japanese-jisx0208-1978 japanese-jisx0208
 118                                 japanese-jisx0212 katakana-jisx0201)
 119                 nil nil nil]
 120   :flags '(short ascii-at-eol ascii-at-cntl 7-bit designation
 121                  use-roman use-oldjis)
 122   :charset-list '(ascii latin-jisx0201 japanese-jisx0208-1978 japanese-jisx0208
 123                         japanese-jisx0212))
 124
 125 (define-coding-system-alias 'iso-2022-jp-1978-irv 'japanese-iso-7bit-1978-irv)
 126 (define-coding-system-alias 'old-jis 'japanese-iso-7bit-1978-irv)
 127
 128 (define-coding-system 'japanese-iso-8bit
 129   "ISO 2022 based EUC encoding for Japanese (MIME:EUC-JP)."
 130   :coding-type 'iso-2022
 131   :mnemonic ?E
 132   :designation [ascii japanese-jisx0208 katakana-jisx0201 japanese-jisx0212]
 133   :flags '(short ascii-at-eol ascii-at-cntl single-shift)
 134   :charset-list '(ascii latin-jisx0201 japanese-jisx0208
 135                         katakana-jisx0201 japanese-jisx0212
 136                         japanese-jisx0208-1978)
 137   :mime-charset 'euc-jp)
 138
 139 (define-coding-system-alias 'euc-japan-1990 'japanese-iso-8bit)
 140 (define-coding-system-alias 'euc-japan 'japanese-iso-8bit)
 141 (define-coding-system-alias 'euc-jp 'japanese-iso-8bit)
 142
 143 (define-coding-system 'eucjp-ms
 144   "eucJP-ms (like EUC-JP but with CP932 extension).
 145 eucJP-ms is defined in <http://www.opengroup.or.jp/jvc/cde/appendix.html>."
 146   :coding-type 'iso-2022
 147   :mnemonic ?E
 148   :designation [ascii japanese-jisx0208 katakana-jisx0201 japanese-jisx0212]
 149   :flags '(short ascii-at-eol ascii-at-cntl single-shift)
 150   :charset-list '(ascii latin-jisx0201 japanese-jisx0208
 151                         katakana-jisx0201 japanese-jisx0212)
 152   :decode-translation-table 'eucjp-ms-decode
 153   :encode-translation-table 'eucjp-ms-encode)
 154
 155 (define-coding-system 'iso-2022-jp-2004
 156   "ISO 2022 based 7bit encoding for JIS X 0213:2004 (MIME:ISO-2022-JP-2004)."
 157   :coding-type 'iso-2022
 158   :mnemonic ?J
 159   :designation [(ascii japanese-jisx0208 japanese-jisx0213.2004-1
 160                        japanese-jisx0213-1 japanese-jisx0213-2)
 161                 nil nil nil]
 162   :flags '(short ascii-at-eol ascii-at-cntl 7-bit designation)
 163                  ;; init-at-bol)
 164   :charset-list '(ascii japanese-jisx0208 japanese-jisx0213.2004-1
 165                         japanese-jisx0213-1 japanese-jisx0213-2)
 166   :mime-charset 'iso-2022-jp-2004
 167   :suitable-for-keyboard t)
 168
 169 (define-coding-system-alias 'iso-2022-jp-3 'iso-2022-jp-2004)
 170
 171 (define-coding-system 'euc-jis-2004
 172   "ISO 2022 based EUC encoding for JIS X 0213 (MIME:EUC-JIS-2004)."
 173   :coding-type 'iso-2022
 174   :mnemonic ?E
 175   :designation [ascii japanese-jisx0213.2004-1 katakana-jisx0201
 176                       japanese-jisx0213-2]
 177   :flags '(short ascii-at-eol ascii-at-cntl single-shift)
 178   :charset-list '(ascii latin-jisx0201 japanese-jisx0213.2004-1
 179                         japanese-jisx0213-1 katakana-jisx0201
 180                         japanese-jisx0213-2)
 181   :mime-charset 'euc-jis-2004)
 182
 183 (define-coding-system-alias 'euc-jisx0213 'euc-jis-2004)
 184
 185 (define-coding-system 'japanese-shift-jis-2004
 186   "Shift_JIS 8-bit encoding for Japanese (MIME:SHIFT_JIS-2004)"
 187   :coding-type 'shift-jis
 188   :mnemonic ?S
 189   :charset-list '(ascii katakana-jisx0201
 190                         japanese-jisx0213.2004-1 japanese-jisx0213-2))
 191
 192 (define-coding-system-alias 'shift_jis-2004 'japanese-shift-jis-2004)
 193
 194 (set-language-info-alist
 195  "Japanese" '((setup-function . setup-japanese-environment-internal)
 196               (exit-function . use-default-char-width-table)
 197               (iso639-language . ja)
 198               (tutorial . "TUTORIAL.ja")
 199               (charset japanese-jisx0208
 200                        japanese-jisx0212 latin-jisx0201 katakana-jisx0201
 201                        japanese-jisx0213.2004-1 japanese-jisx0213-1
 202                        japanese-jisx0213-2 japanese-jisx0208-1978)
 203               (coding-system iso-2022-jp japanese-iso-8bit
 204                              japanese-shift-jis japanese-iso-7bit-1978-irv
 205                              iso-2022-jp-2004 japanese-shift-jis-2004
 206                              euc-jis-2004)
 207               (coding-priority iso-2022-jp japanese-iso-8bit
 208                                japanese-shift-jis
 209                                iso-2022-jp-2004 euc-jis-2004
 210                                japanese-shift-jis-2004
 211                                iso-2022-jp-2)
 212               (input-method . "japanese")
 213               (features japan-util)
 214               (sample-text . "Japanese (\e$BF|K\8l\e(B)   \e$B$3$s$K$A$O\e(B, \e(I:]FAJ\e(B")
 215               (documentation . t)))
 216
 217 (let ((map
 218        ;; JISX0213-1 vs Unicode
 219        '((#x2477 . [#x304B #x309A])
 220          (#x2478 . [#x304D #x309A])
 221          (#x2479 . [#x304F #x309A])
 222          (#x247a . [#x3051 #x309A])
 223          (#x247b . [#x3053 #x309A])
 224          (#x2577 . [#x30AB #x309A])
 225          (#x2578 . [#x30AD #x309A])
 226          (#x2579 . [#x30AF #x309A])
 227          (#x257a . [#x30B1 #x309A])
 228          (#x257b . [#x30B3 #x309A])
 229          (#x257c . [#x30BB #x309A])
 230          (#x257d . [#x30C4 #x309A])
 231          (#x257e . [#x30C8 #x309A])
 232          (#x2678 . [#x31F7 #x309A])
 233          (#x2b44 . [#x00E6 #x0300])
 234          (#x2b48 . [#x0254 #x0300])
 235          (#x2b49 . [#x0254 #x0301])
 236          (#x2b4a . [#x028C #x0300])
 237          (#x2b4b . [#x028C #x0301])
 238          (#x2b4c . [#x0259 #x0300])
 239          (#x2b4d . [#x0259 #x0301])
 240          (#x2b4e . [#x025A #x0300])
 241          (#x2b4f . [#x025A #x0301])
 242          (#x2b65 . [#x02E9 #x02E5])
 243          (#x2b66 . [#x02E5 #x02E9])))
 244       table)
 245   (dolist (elt map)
 246     (setcar elt (decode-char 'japanese-jisx0213-1 (car elt))))
 247   (setq table (make-translation-table-from-alist map))
 248   (define-translation-table 'jisx0213-to-unicode table)
 249   (define-translation-table 'unicode-to-jisx0213
 250     (char-table-extra-slot table 0)))
 251
 252 (defun compose-gstring-for-variation-glyph (gstring)
 253   "Compose glyph-string GSTRING for graphic display.
 254 GSTRING must have two glyphs; the first is a glyph for a han character,
 255 and the second is a glyph for a variation selector."
 256   (let* ((font (lgstring-font gstring))
 257          (han (lgstring-char gstring 0))
 258          (vs (lgstring-char gstring 1))
 259          (glyphs (font-variation-glyphs font han))
 260          (g0 (lgstring-glyph gstring 0))
 261          (g1 (lgstring-glyph gstring 1)))
 262     (catch 'tag
 263       (dolist (elt glyphs)
 264         (if (= (car elt) vs)
 265             (progn
 266               (lglyph-set-code g0 (cdr elt))
 267               (lglyph-set-from-to g0 (lglyph-from g0) (lglyph-to g1))
 268               (lgstring-set-glyph gstring 1 nil)
 269               (throw 'tag gstring)))))))
 270
 271 (let ((elt '([".." 1 compose-gstring-for-variation-glyph])))
 272   (set-char-table-range composition-function-table '(#xFE00 . #xFE0F) elt)
 273   (set-char-table-range composition-function-table '(#xE0100 . #xE01EF) elt))
 274
 275 (provide 'japanese)
 276
 277 ;; arch-tag: 450f5537-9d53-4d5e-b731-4cf116d8cbc9
 278 ;;; japanese.el ends here