lisp/gnus/rfc1843.el

   1 ;;; rfc1843.el --- HZ (rfc1843) decoding
   2
   3 ;; Copyright (C) 1998-2016 Free Software Foundation, Inc.
   4
   5 ;; Author: Shenghuo Zhu <zsh@cs.rochester.edu>
   6 ;; Keywords: news HZ HZ+ mail i18n
   7
   8 ;; This file is part of GNU Emacs.
   9
  10 ;; GNU Emacs is free software: you can redistribute it and/or modify
  11 ;; it under the terms of the GNU General Public License as published by
  12 ;; the Free Software Foundation, either version 3 of the License, or
  13 ;; (at your option) any later version.
  14
  15 ;; GNU Emacs is distributed in the hope that it will be useful,
  16 ;; but WITHOUT ANY WARRANTY; without even the implied warranty of
  17 ;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  18 ;; GNU General Public License for more details.
  19
  20 ;; You should have received a copy of the GNU General Public License
  21 ;; along with GNU Emacs.  If not, see <http://www.gnu.org/licenses/>.
  22
  23 ;;; Commentary:
  24
  25 ;; Usage:
  26 ;; (require 'rfc1843)
  27 ;; (rfc1843-gnus-setup)
  28 ;;
  29 ;; Test:
  30 ;; (rfc1843-decode-string  "~{<:Ky2;S{#,NpJ)l6HK!#~}")
  31
  32 ;;; Code:
  33
  34 (eval-when-compile (require 'cl))
  35 (require 'mm-util)
  36
  37 (defvar gnus-decode-encoded-word-function)
  38 (defvar gnus-decode-header-function)
  39 (defvar gnus-newsgroup-name)
  40
  41 (defvar rfc1843-word-regexp
  42   "~\\({\\([\041-\167][\041-\176]\\| \\)+\\)\\(~}\\|$\\)")
  43
  44 (defvar rfc1843-word-regexp-strictly
  45   "~\\({\\([\041-\167][\041-\176]\\)+\\)\\(~}\\|$\\)")
  46
  47 (defvar rfc1843-hzp-word-regexp
  48   "~\\({\\([\041-\167][\041-\176]\\| \\)+\\|\
  49 [<>]\\([\041-\175][\041-\176]\\| \\)+\\)\\(~}\\|$\\)")
  50
  51 (defvar rfc1843-hzp-word-regexp-strictly
  52   "~\\({\\([\041-\167][\041-\176]\\)+\\|\
  53 [<>]\\([\041-\175][\041-\176]\\)+\\)\\(~}\\|$\\)")
  54
  55 (defcustom rfc1843-decode-loosely nil
  56   "Loosely check HZ encoding if non-nil.
  57 When it is set non-nil, only buffers or strings with strictly
  58 HZ-encoded are decoded."
  59   :type 'boolean
  60   :group 'mime)
  61
  62 (defcustom rfc1843-decode-hzp t
  63   "HZ+ decoding support if non-nil.
  64 HZ+ specification (also known as HZP) is to provide a standardized
  65 7-bit representation of mixed Big5, GB, and ASCII text for convenient
  66 e-mail transmission, news posting, etc.
  67 The document of HZ+ 0.78 specification can be found at
  68 ftp://ftp.math.psu.edu/pub/simpson/chinese/hzp/hzp.doc"
  69   :type 'boolean
  70   :group 'mime)
  71
  72 (defcustom rfc1843-newsgroups-regexp "chinese\\|hz"
  73   "Regexp of newsgroups in which might be HZ encoded."
  74   :type 'string
  75   :group 'mime)
  76
  77 (defun rfc1843-decode-region (from to)
  78   "Decode HZ in the region between FROM and TO."
  79   (interactive "r")
  80   (let (str firstc)
  81     (save-excursion
  82       (goto-char from)
  83       (if (or rfc1843-decode-loosely
  84               (re-search-forward (if rfc1843-decode-hzp
  85                                      rfc1843-hzp-word-regexp-strictly
  86                                    rfc1843-word-regexp-strictly) to t))
  87           (save-restriction
  88             (narrow-to-region from to)
  89             (goto-char (point-min))
  90             (while (re-search-forward (if rfc1843-decode-hzp
  91                                           rfc1843-hzp-word-regexp
  92                                         rfc1843-word-regexp) (point-max) t)
  93               ;;; Text with extents may cause XEmacs crash
  94               (setq str (buffer-substring-no-properties
  95                          (match-beginning 1)
  96                          (match-end 1)))
  97               (setq firstc (aref str 0))
  98               (insert (mm-decode-coding-string
  99                        (rfc1843-decode
 100                         (prog1
 101                             (substring str 1)
 102                           (delete-region (match-beginning 0) (match-end 0)))
 103                         firstc)
 104                        (if (eq firstc ?{) 'cn-gb-2312 'cn-big5))))
 105             (goto-char (point-min))
 106             (while (search-forward "~" (point-max) t)
 107               (cond ((eq (char-after) ?\n)
 108                      (delete-char -1)
 109                      (delete-char 1))
 110                     ((eq (char-after) ?~)
 111                      (delete-char 1)))))))))
 112
 113 (defun rfc1843-decode-string (string)
 114   "Decode HZ STRING and return the results."
 115   (let ((m (mm-multibyte-p)))
 116     (with-temp-buffer
 117       (when m
 118         (mm-enable-multibyte))
 119       (insert string)
 120       (inline
 121         (rfc1843-decode-region (point-min) (point-max)))
 122       (buffer-string))))
 123
 124 (defun rfc1843-decode (word &optional firstc)
 125   "Decode HZ WORD and return it."
 126   (let ((i -1) (s (substring word 0)) v)
 127     (if (or (not firstc) (eq firstc ?{))
 128         (while (< (incf i) (length s))
 129           (if (eq (setq v (aref s i)) ? ) nil
 130             (aset s i (+ 128 v))))
 131       (while (< (incf i) (length s))
 132         (if (eq (setq v (aref s i)) ? ) nil
 133           (setq v (+ (* 94 v) (aref s (1+ i)) -3135))
 134           (aset s i (+ (/ v 157) (if (eq firstc ?<) 201 161)))
 135           (setq v (% v 157))
 136           (aset s (incf i) (+ v (if (< v 63) 64 98))))))
 137     s))
 138
 139 (autoload 'mail-header-parse-content-type "mail-parse")
 140 (autoload 'message-narrow-to-head "message")
 141 (declare-function message-fetch-field "message" (header &optional not-all))
 142
 143 (defun rfc1843-decode-article-body ()
 144   "Decode HZ encoded text in the article body."
 145   (if (string-match (concat "\\<\\(" rfc1843-newsgroups-regexp "\\)\\>")
 146                     (or gnus-newsgroup-name ""))
 147       (save-excursion
 148         (save-restriction
 149           (message-narrow-to-head)
 150           (let* ((inhibit-point-motion-hooks t)
 151                  (case-fold-search t)
 152                  (ct (message-fetch-field "Content-Type" t))
 153                  (ctl (and ct (mail-header-parse-content-type ct))))
 154             (if (and ctl (not (string-match "/" (car ctl))))
 155                 (setq ctl nil))
 156             (goto-char (point-max))
 157             (widen)
 158             (forward-line 1)
 159             (narrow-to-region (point) (point-max))
 160             (when (or (not ctl)
 161                       (equal (car ctl) "text/plain"))
 162               (rfc1843-decode-region (point) (point-max))))))))
 163
 164 (defvar gnus-decode-header-methods)
 165 (defvar gnus-decode-encoded-word-methods)
 166
 167 (defun rfc1843-gnus-setup ()
 168   "Setup HZ decoding for Gnus."
 169   (require 'gnus-art)
 170   (require 'gnus-sum)
 171   (add-hook 'gnus-article-decode-hook 'rfc1843-decode-article-body t)
 172   (setq gnus-decode-encoded-word-function
 173         'gnus-multi-decode-encoded-word-string
 174         gnus-decode-header-function
 175         'gnus-multi-decode-header
 176         gnus-decode-encoded-word-methods
 177         (nconc gnus-decode-encoded-word-methods
 178                (list
 179                 (cons (concat "\\<\\(" rfc1843-newsgroups-regexp "\\)\\>")
 180                       'rfc1843-decode-string)))
 181         gnus-decode-header-methods
 182         (nconc gnus-decode-header-methods
 183                (list
 184                 (cons (concat "\\<\\(" rfc1843-newsgroups-regexp "\\)\\>")
 185                       'rfc1843-decode-region)))))
 186
 187 (provide 'rfc1843)
 188
 189 ;;; rfc1843.el ends here