lisp/gnus/rfc1843.el

   1 ;;; rfc1843.el --- HZ (rfc1843) decoding
   2
   3 ;; Copyright (C) 1998, 1999, 2000, 2001, 2002, 2003, 2004,
   4 ;;   2005, 2006, 2007, 2008 Free Software Foundation, Inc.
   5
   6 ;; Author: Shenghuo Zhu <zsh@cs.rochester.edu>
   7 ;; Keywords: news HZ HZ+ mail i18n
   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
  13 ;; by the Free Software Foundation; either version 3, or (at your
  14 ;; option) any later version.
  15
  16 ;; GNU Emacs is distributed in the hope that it will be useful, but
  17 ;; WITHOUT ANY WARRANTY; without even the implied warranty of
  18 ;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  19 ;; 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., 51 Franklin Street, Fifth Floor,
  24 ;; Boston, MA 02110-1301, USA.
  25
  26 ;;; Commentary:
  27
  28 ;; Usage:
  29 ;; (require 'rfc1843)
  30 ;; (rfc1843-gnus-setup)
  31 ;;
  32 ;; Test:
  33 ;; (rfc1843-decode-string  "~{<:Ky2;S{#,NpJ)l6HK!#~}")
  34
  35 ;;; Code:
  36
  37 ;; For Emacs < 22.2.
  38 (eval-and-compile
  39   (unless (fboundp 'declare-function) (defmacro declare-function (&rest r))))
  40
  41 (eval-when-compile (require 'cl))
  42 (require 'mm-util)
  43
  44 (defvar gnus-decode-encoded-word-function)
  45 (defvar gnus-decode-header-function)
  46 (defvar gnus-newsgroup-name)
  47
  48 (defvar rfc1843-word-regexp
  49   "~\\({\\([\041-\167][\041-\176]\\| \\)+\\)\\(~}\\|$\\)")
  50
  51 (defvar rfc1843-word-regexp-strictly
  52   "~\\({\\([\041-\167][\041-\176]\\)+\\)\\(~}\\|$\\)")
  53
  54 (defvar rfc1843-hzp-word-regexp
  55   "~\\({\\([\041-\167][\041-\176]\\| \\)+\\|\
  56 \[<>]\\([\041-\175][\041-\176]\\| \\)+\\)\\(~}\\|$\\)")
  57
  58 (defvar rfc1843-hzp-word-regexp-strictly
  59   "~\\({\\([\041-\167][\041-\176]\\)+\\|\
  60 \[<>]\\([\041-\175][\041-\176]\\)+\\)\\(~}\\|$\\)")
  61
  62 (defcustom rfc1843-decode-loosely nil
  63   "Loosely check HZ encoding if non-nil.
  64 When it is set non-nil, only buffers or strings with strictly
  65 HZ-encoded are decoded."
  66   :type 'boolean
  67   :group 'mime)
  68
  69 (defcustom rfc1843-decode-hzp t
  70   "HZ+ decoding support if non-nil.
  71 HZ+ specification (also known as HZP) is to provide a standardized
  72 7-bit representation of mixed Big5, GB, and ASCII text for convenient
  73 e-mail transmission, news posting, etc.
  74 The document of HZ+ 0.78 specification can be found at
  75 ftp://ftp.math.psu.edu/pub/simpson/chinese/hzp/hzp.doc"
  76   :type 'boolean
  77   :group 'mime)
  78
  79 (defcustom rfc1843-newsgroups-regexp "chinese\\|hz"
  80   "Regexp of newsgroups in which might be HZ encoded."
  81   :type 'string
  82   :group 'mime)
  83
  84 (defun rfc1843-decode-region (from to)
  85   "Decode HZ in the region between FROM and TO."
  86   (interactive "r")
  87   (let (str firstc)
  88     (save-excursion
  89       (goto-char from)
  90       (if (or rfc1843-decode-loosely
  91               (re-search-forward (if rfc1843-decode-hzp
  92                                      rfc1843-hzp-word-regexp-strictly
  93                                    rfc1843-word-regexp-strictly) to t))
  94           (save-restriction
  95             (narrow-to-region from to)
  96             (goto-char (point-min))
  97             (while (re-search-forward (if rfc1843-decode-hzp
  98                                           rfc1843-hzp-word-regexp
  99                                         rfc1843-word-regexp) (point-max) t)
 100               ;;; Text with extents may cause XEmacs crash
 101               (setq str (buffer-substring-no-properties
 102                          (match-beginning 1)
 103                          (match-end 1)))
 104               (setq firstc (aref str 0))
 105               (insert (mm-decode-coding-string
 106                        (rfc1843-decode
 107                         (prog1
 108                             (substring str 1)
 109                           (delete-region (match-beginning 0) (match-end 0)))
 110                         firstc)
 111                        (if (eq firstc ?{) 'cn-gb-2312 'cn-big5))))
 112             (goto-char (point-min))
 113             (while (search-forward "~" (point-max) t)
 114               (cond ((eq (char-after) ?\n)
 115                      (delete-char -1)
 116                      (delete-char 1))
 117                     ((eq (char-after) ?~)
 118                      (delete-char 1)))))))))
 119
 120 (defun rfc1843-decode-string (string)
 121   "Decode HZ STRING and return the results."
 122   (let ((m (mm-multibyte-p)))
 123     (with-temp-buffer
 124       (when m
 125         (mm-enable-multibyte))
 126       (insert string)
 127       (inline
 128         (rfc1843-decode-region (point-min) (point-max)))
 129       (buffer-string))))
 130
 131 (defun rfc1843-decode (word &optional firstc)
 132   "Decode HZ WORD and return it."
 133   (let ((i -1) (s (substring word 0)) v)
 134     (if (or (not firstc) (eq firstc ?{))
 135         (while (< (incf i) (length s))
 136           (if (eq (setq v (aref s i)) ? ) nil
 137             (aset s i (+ 128 v))))
 138       (while (< (incf i) (length s))
 139         (if (eq (setq v (aref s i)) ? ) nil
 140           (setq v (+ (* 94 v) (aref s (1+ i)) -3135))
 141           (aset s i (+ (/ v 157) (if (eq firstc ?<) 201 161)))
 142           (setq v (% v 157))
 143           (aset s (incf i) (+ v (if (< v 63) 64 98))))))
 144     s))
 145
 146 (autoload 'mail-header-parse-content-type "mail-parse")
 147 (autoload 'message-narrow-to-head "message")
 148 (declare-function message-fetch-field "message" (header &optional not-all))
 149
 150 (defun rfc1843-decode-article-body ()
 151   "Decode HZ encoded text in the article body."
 152   (if (string-match (concat "\\<\\(" rfc1843-newsgroups-regexp "\\)\\>")
 153                     (or gnus-newsgroup-name ""))
 154       (save-excursion
 155         (save-restriction
 156           (message-narrow-to-head)
 157           (let* ((inhibit-point-motion-hooks t)
 158                  (case-fold-search t)
 159                  (ct (message-fetch-field "Content-Type" t))
 160                  (ctl (and ct (mail-header-parse-content-type ct))))
 161             (if (and ctl (not (string-match "/" (car ctl))))
 162                 (setq ctl nil))
 163             (goto-char (point-max))
 164             (widen)
 165             (forward-line 1)
 166             (narrow-to-region (point) (point-max))
 167             (when (or (not ctl)
 168                       (equal (car ctl) "text/plain"))
 169               (rfc1843-decode-region (point) (point-max))))))))
 170
 171 (defvar rfc1843-old-gnus-decode-header-function  nil)
 172 (defvar gnus-decode-header-methods)
 173 (defvar gnus-decode-encoded-word-methods)
 174
 175 (defun rfc1843-gnus-setup ()
 176   "Setup HZ decoding for Gnus."
 177   (require 'gnus-art)
 178   (require 'gnus-sum)
 179   (add-hook 'gnus-article-decode-hook 'rfc1843-decode-article-body t)
 180   (setq gnus-decode-encoded-word-function
 181         'gnus-multi-decode-encoded-word-string
 182         gnus-decode-header-function
 183         'gnus-multi-decode-header
 184         gnus-decode-encoded-word-methods
 185         (nconc gnus-decode-encoded-word-methods
 186                (list
 187                 (cons (concat "\\<\\(" rfc1843-newsgroups-regexp "\\)\\>")
 188                       'rfc1843-decode-string)))
 189         gnus-decode-header-methods
 190         (nconc gnus-decode-header-methods
 191                (list
 192                 (cons (concat "\\<\\(" rfc1843-newsgroups-regexp "\\)\\>")
 193                       'rfc1843-decode-region)))))
 194
 195 (provide 'rfc1843)
 196
 197 ;; arch-tag: 5149c301-a6ca-4731-9c9d-ba616e2cb687
 198 ;;; rfc1843.el ends here