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 Electrotechnical Laboratory, JAPAN.
   4 ;; Licensed to the Free Software Foundation.
   5 ;; Copyright (C) 1995, 2001 Free Software Foundation, Inc.
   6
   7 ;; Keywords: mule, multilingual, Chinese
   8
   9 ;; This file is part of GNU Emacs.
  10
  11 ;; GNU Emacs is free software; you can redistribute it and/or modify
  12 ;; it under the terms of the GNU General Public License as published by
  13 ;; the Free Software Foundation; either version 2, or (at your option)
  14 ;; any later version.
  15
  16 ;; GNU Emacs is distributed in the hope that it will be useful,
  17 ;; but WITHOUT ANY WARRANTY; without even the implied warranty of
  18 ;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  19 ;; GNU General Public License for more details.
  20
  21 ;; You should have received a copy of the GNU General Public License
  22 ;; along with GNU Emacs; see the file COPYING.  If not, write to the
  23 ;; Free Software Foundation, Inc., 59 Temple Place - Suite 330,
  24 ;; Boston, MA 02111-1307, USA.
  25
  26 ;;; Commentary:
  27
  28 ;;; Code:
  29
  30 ;; Hz/ZW/EUC-TW encoding stuff
  31
  32 ;; HZ is an encoding method for Chinese character set GB2312 used
  33 ;; widely in Internet.  It is very similar to 7-bit environment of
  34 ;; ISO-2022.  The difference is that HZ uses the sequence "~{" and
  35 ;; "~}" for designating GB2312 and ASCII respectively, hence, it
  36 ;; doesn't uses ESC (0x1B) code.
  37
  38 ;; ZW is another encoding method for Chinese character set GB2312.  It
  39 ;; encodes Chinese characters line by line by starting each line with
  40 ;; the sequence "zW".  It also uses only 7-bit as HZ.
  41
  42 ;; EUC-TW is similar to EUC-KS or EUC-JP.  Its main character set is
  43 ;; plane 1 of CNS 11643; characters of planes 2 to 7 are accessed with
  44 ;; a single shift escape followed by three bytes: the first gives the
  45 ;; plane, the second and third the character code.  Note that characters
  46 ;; of plane 1 are (redundantly) accessible with a single shift escape
  47 ;; also.
  48
  49 ;; ISO-2022 escape sequence to designate GB2312.
  50 (defvar iso2022-gb-designation "\e$A")
  51 ;; HZ escape sequence to designate GB2312.
  52 (defvar hz-gb-designnation "~{")
  53 ;; ISO-2022 escape sequence to designate ASCII.
  54 (defvar iso2022-ascii-designation "\e(B")
  55 ;; HZ escape sequence to designate ASCII.
  56 (defvar hz-ascii-designnation "~}")
  57 ;; Regexp of ZW sequence to start GB2312.
  58 (defvar zw-start-gb "^zW")
  59 ;; Regexp for start of GB2312 in an encoding mixture of HZ and ZW.
  60 (defvar hz/zw-start-gb
  61   (concat hz-gb-designnation "\\|" zw-start-gb "\\|[^\0-\177]"))
  62
  63 (defvar decode-hz-line-continuation nil
  64   "Flag to tell if we should care line continuation convention of Hz.")
  65
  66 (defconst hz-set-msb-table
  67   (let ((str (make-string 127 0))
  68         (i 0))
  69     (while (< i 33)
  70       (aset str i i)
  71       (setq i (1+ i)))
  72     (while (< i 127)
  73       (aset str i (+ i 128))
  74       (setq i (1+ i)))
  75     str))
  76
  77 ;;;###autoload
  78 (defun decode-hz-region (beg end)
  79   "Decode HZ/ZW encoded text in the current region.
  80 Return the length of resulting text."
  81   (interactive "r")
  82   (save-excursion
  83     (save-restriction
  84       (let (pos ch)
  85         (narrow-to-region beg end)
  86
  87         ;; We, at first, convert HZ/ZW to `euc-china',
  88         ;; then decode it.
  89
  90         ;; "~\n" -> "\n", "~~" -> "~"
  91         (goto-char (point-min))
  92         (while (search-forward "~" nil t)
  93           (setq ch (following-char))
  94           (if (or (= ch ?\n) (= ch ?~)) (delete-char -1)))
  95
  96         ;; "^zW...\n" -> Chinese GB2312
  97         ;; "~{...~}"  -> Chinese GB2312
  98         (goto-char (point-min))
  99         (setq beg nil)
 100         (while (re-search-forward hz/zw-start-gb nil t)
 101           (setq pos (match-beginning 0)
 102                 ch (char-after pos))
 103           ;; Record the first position to start conversion.
 104           (or beg (setq beg pos))
 105           (end-of-line)
 106           (setq end (point))
 107           (if (>= ch 128)               ; 8bit GB2312
 108               nil
 109             (goto-char pos)
 110             (delete-char 2)
 111             (setq end (- end 2))
 112             (if (= ch ?z)                       ; ZW -> euc-china
 113                 (progn
 114                   (translate-region (point) end hz-set-msb-table)
 115                   (goto-char end))
 116               (if (search-forward hz-ascii-designnation
 117                                   (if decode-hz-line-continuation nil end)
 118                                   t)
 119                   (delete-char -2))
 120               (setq end (point))
 121               (translate-region pos (point) hz-set-msb-table))))
 122         (if beg
 123             (decode-coding-region beg end 'euc-china)))
 124       (- (point-max) (point-min)))))
 125
 126 ;;;###autoload
 127 (defun decode-hz-buffer ()
 128   "Decode HZ/ZW encoded text in the current buffer."
 129   (interactive)
 130   (decode-hz-region (point-min) (point-max)))
 131
 132 ;;;###autoload
 133 (defun encode-hz-region (beg end)
 134   "Encode the text in the current region to HZ.
 135 Return the length of resulting text."
 136   (interactive "r")
 137   (save-excursion
 138     (save-restriction
 139       (narrow-to-region beg end)
 140
 141       ;; "~" -> "~~"
 142       (goto-char (point-min))
 143       (while (search-forward "~" nil t) (insert ?~))
 144
 145       ;; Chinese GB2312 -> "~{...~}"
 146       (goto-char (point-min))
 147       (if (re-search-forward "\\cc" nil t)
 148           (let (pos)
 149             (goto-char (setq pos (match-beginning 0)))
 150             (encode-coding-region pos (point-max) 'iso-2022-7bit)
 151             (goto-char pos)
 152             (while (search-forward iso2022-gb-designation nil t)
 153               (delete-char -3)
 154               (insert hz-gb-designnation))
 155             (goto-char pos)
 156             (while (search-forward iso2022-ascii-designation nil t)
 157               (delete-char -3)
 158               (insert hz-ascii-designnation))))
 159       (- (point-max) (point-min)))))
 160
 161 ;;;###autoload
 162 (defun encode-hz-buffer ()
 163   "Encode the text in the current buffer to HZ."
 164   (interactive)
 165   (encode-hz-region (point-min) (point-max)))
 166
 167 ;;;###autoload
 168 (defun post-read-decode-hz (len)
 169   (let ((pos (point))
 170         (buffer-modified-p (buffer-modified-p))
 171         last-coding-system-used)
 172     (prog1
 173         (decode-hz-region pos (+ pos len))
 174       (set-buffer-modified-p buffer-modified-p))))
 175
 176 ;;;###autoload
 177 (defun pre-write-encode-hz (from to)
 178   (let ((buf (current-buffer)))
 179     (set-buffer (generate-new-buffer " *temp*"))
 180     (if (stringp from)
 181         (insert from)
 182       (insert-buffer-substring buf from to))
 183     (let (last-coding-system-used)
 184       (encode-hz-region 1 (point-max)))
 185     nil))
 186 ;;
 187 (provide 'china-util)
 188
 189 ;;; china-util.el ends here