code.delx.au - gnu-emacs/blob - lisp/language/china-util.el

   1 ;;; china-util.el --- utilities for Chinese  -*- coding: iso-2022-7bit -*-
   2
   3 ;; Copyright (C) 1995, 2001, 2002, 2003, 2004, 2005, 2006, 2007, 2008
   4 ;;   Free Software Foundation, Inc.
   5 ;; Copyright (C) 1995, 1996, 1997, 1998, 1999, 2000, 2001, 2002, 2003, 2004,
   6 ;;   2005, 2006, 2007, 2008
   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: mule, multilingual, Chinese
  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, or (at your option)
  20 ;; 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; see the file COPYING.  If not, write to the
  29 ;; Free Software Foundation, Inc., 51 Franklin Street, Fifth Floor,
  30 ;; Boston, MA 02110-1301, USA.
  31
  32 ;;; Commentary:
  33
  34 ;;; Code:
  35
  36 ;; Hz/ZW/EUC-TW encoding stuff
  37
  38 ;; HZ is an encoding method for Chinese character set GB2312 used
  39 ;; widely in Internet.  It is very similar to 7-bit environment of
  40 ;; ISO-2022.  The difference is that HZ uses the sequence "~{" and
  41 ;; "~}" for designating GB2312 and ASCII respectively, hence, it
  42 ;; doesn't uses ESC (0x1B) code.
  43
  44 ;; ZW is another encoding method for Chinese character set GB2312.  It
  45 ;; encodes Chinese characters line by line by starting each line with
  46 ;; the sequence "zW".  It also uses only 7-bit as HZ.
  47
  48 ;; EUC-TW is similar to EUC-KS or EUC-JP.  Its main character set is
  49 ;; plane 1 of CNS 11643; characters of planes 2 to 7 are accessed with
  50 ;; a single shift escape followed by three bytes: the first gives the
  51 ;; plane, the second and third the character code.  Note that characters
  52 ;; of plane 1 are (redundantly) accessible with a single shift escape
  53 ;; also.
  54
  55 ;; ISO-2022 escape sequence to designate GB2312.
  56 (defvar iso2022-gb-designation "\e$A")
  57 ;; HZ escape sequence to designate GB2312.
  58 (defvar hz-gb-designnation "~{")
  59 ;; ISO-2022 escape sequence to designate ASCII.
  60 (defvar iso2022-ascii-designation "\e(B")
  61 ;; HZ escape sequence to designate ASCII.
  62 (defvar hz-ascii-designnation "~}")
  63 ;; Regexp of ZW sequence to start GB2312.
  64 (defvar zw-start-gb "^zW")
  65 ;; Regexp for start of GB2312 in an encoding mixture of HZ and ZW.
  66 (defvar hz/zw-start-gb
  67   (concat hz-gb-designnation "\\|" zw-start-gb "\\|[^\0-\177]"))
  68
  69 (defvar decode-hz-line-continuation nil
  70   "Flag to tell if we should care line continuation convention of Hz.")
  71
  72 (defconst hz-set-msb-table
  73   (eval-when-compile
  74     (let ((chars nil)
  75           (i 0))
  76       (while (< i 33)
  77         (push i chars)
  78         (setq i (1+ i)))
  79       (while (< i 127)
  80         (push (decode-char 'eight-bit (+ i 128)) chars)
  81         (setq i (1+ i)))
  82       (apply 'string (nreverse chars)))))
  83
  84 ;;;###autoload
  85 (defun decode-hz-region (beg end)
  86   "Decode HZ/ZW encoded text in the current region.
  87 Return the length of resulting text."
  88   (interactive "r")
  89   (save-excursion
  90     (save-restriction
  91       (let (pos ch)
  92         (narrow-to-region beg end)
  93
  94         ;; We, at first, convert HZ/ZW to `euc-china',
  95         ;; then decode it.
  96
  97         ;; "~\n" -> "\n", "~~" -> "~"
  98         (goto-char (point-min))
  99         (while (search-forward "~" nil t)
 100           (setq ch (following-char))
 101           (if (or (= ch ?\n) (= ch ?~)) (delete-char -1)))
 102
 103         ;; "^zW...\n" -> Chinese GB2312
 104         ;; "~{...~}"  -> Chinese GB2312
 105         (goto-char (point-min))
 106         (setq beg nil)
 107         (while (re-search-forward hz/zw-start-gb nil t)
 108           (setq pos (match-beginning 0)
 109                 ch (char-after pos))
 110           ;; Record the first position to start conversion.
 111           (or beg (setq beg pos))
 112           (end-of-line)
 113           (setq end (point))
 114           (if (>= ch 128)               ; 8bit GB2312
 115               nil
 116             (goto-char pos)
 117             (delete-char 2)
 118             (setq end (- end 2))
 119             (if (= ch ?z)                       ; ZW -> euc-china
 120                 (progn
 121                   (translate-region (point) end hz-set-msb-table)
 122                   (goto-char end))
 123               (if (search-forward hz-ascii-designnation
 124                                   (if decode-hz-line-continuation nil end)
 125                                   t)
 126                   (delete-char -2))
 127               (setq end (point))
 128               (translate-region pos (point) hz-set-msb-table))))
 129         (if beg
 130             (decode-coding-region beg end 'euc-china)))
 131       (- (point-max) (point-min)))))
 132
 133 ;;;###autoload
 134 (defun decode-hz-buffer ()
 135   "Decode HZ/ZW encoded text in the current buffer."
 136   (interactive)
 137   (decode-hz-region (point-min) (point-max)))
 138
 139 ;;;###autoload
 140 (defun encode-hz-region (beg end)
 141   "Encode the text in the current region to HZ.
 142 Return the length of resulting text."
 143   (interactive "r")
 144   (save-excursion
 145     (save-restriction
 146       (narrow-to-region beg end)
 147
 148       ;; "~" -> "~~"
 149       (goto-char (point-min))
 150       (while (search-forward "~" nil t) (insert ?~))
 151
 152       ;; Chinese GB2312 -> "~{...~}"
 153       (goto-char (point-min))
 154       (if (re-search-forward "\\cc" nil t)
 155           (let (pos)
 156             (goto-char (setq pos (match-beginning 0)))
 157             (encode-coding-region pos (point-max) 'iso-2022-7bit)
 158             (goto-char pos)
 159             (while (search-forward iso2022-gb-designation nil t)
 160               (delete-char -3)
 161               (insert hz-gb-designnation))
 162             (goto-char pos)
 163             (while (search-forward iso2022-ascii-designation nil t)
 164               (delete-char -3)
 165               (insert hz-ascii-designnation))))
 166       (- (point-max) (point-min)))))
 167
 168 ;;;###autoload
 169 (defun encode-hz-buffer ()
 170   "Encode the text in the current buffer to HZ."
 171   (interactive)
 172   (encode-hz-region (point-min) (point-max)))
 173
 174 ;;;###autoload
 175 (defun post-read-decode-hz (len)
 176   (let ((pos (point))
 177         (buffer-modified-p (buffer-modified-p))
 178         last-coding-system-used)
 179     (prog1
 180         (decode-hz-region pos (+ pos len))
 181       (set-buffer-modified-p buffer-modified-p))))
 182
 183 ;;;###autoload
 184 (defun pre-write-encode-hz (from to)
 185   (let ((buf (current-buffer)))
 186     (set-buffer (generate-new-buffer " *temp*"))
 187     (if (stringp from)
 188         (insert from)
 189       (insert-buffer-substring buf from to))
 190     (let (last-coding-system-used)
 191       (encode-hz-region 1 (point-max)))
 192     nil))
 193 ;;
 194 (provide 'china-util)
 195
 196 ;;; arch-tag: 5a47b084-b9ac-420e-8191-70c5b3a14836
 197 ;;; china-util.el ends here