1;;; rfc1843.el --- HZ (rfc1843) decoding 2 3;; Copyright (C) 1998, 1999, 2000, 2001, 2002, 2003, 2004, 4;; 2005, 2006, 2007 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 2, 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(eval-when-compile (require 'cl)) 38(require 'mm-util) 39 40(defvar gnus-decode-encoded-word-function) 41(defvar gnus-decode-header-function) 42(defvar gnus-newsgroup-name) 43 44(defvar rfc1843-word-regexp 45 "~\\({\\([\041-\167][\041-\176]\\| \\)+\\)\\(~}\\|$\\)") 46 47(defvar rfc1843-word-regexp-strictly 48 "~\\({\\([\041-\167][\041-\176]\\)+\\)\\(~}\\|$\\)") 49 50(defvar rfc1843-hzp-word-regexp 51 "~\\({\\([\041-\167][\041-\176]\\| \\)+\\|\ 52\[<>]\\([\041-\175][\041-\176]\\| \\)+\\)\\(~}\\|$\\)") 53 54(defvar rfc1843-hzp-word-regexp-strictly 55 "~\\({\\([\041-\167][\041-\176]\\)+\\|\ 56\[<>]\\([\041-\175][\041-\176]\\)+\\)\\(~}\\|$\\)") 57 58(defcustom rfc1843-decode-loosely nil 59 "Loosely check HZ encoding if non-nil. 60When it is set non-nil, only buffers or strings with strictly 61HZ-encoded are decoded." 62 :type 'boolean 63 :group 'mime) 64 65(defcustom rfc1843-decode-hzp t 66 "HZ+ decoding support if non-nil. 67HZ+ specification (also known as HZP) is to provide a standardized 687-bit representation of mixed Big5, GB, and ASCII text for convenient 69e-mail transmission, news posting, etc. 70The document of HZ+ 0.78 specification can be found at 71ftp://ftp.math.psu.edu/pub/simpson/chinese/hzp/hzp.doc" 72 :type 'boolean 73 :group 'mime) 74 75(defcustom rfc1843-newsgroups-regexp "chinese\\|hz" 76 "Regexp of newsgroups in which might be HZ encoded." 77 :type 'string 78 :group 'mime) 79 80(defun rfc1843-decode-region (from to) 81 "Decode HZ in the region between FROM and TO." 82 (interactive "r") 83 (let (str firstc) 84 (save-excursion 85 (goto-char from) 86 (if (or rfc1843-decode-loosely 87 (re-search-forward (if rfc1843-decode-hzp 88 rfc1843-hzp-word-regexp-strictly 89 rfc1843-word-regexp-strictly) to t)) 90 (save-restriction 91 (narrow-to-region from to) 92 (goto-char (point-min)) 93 (while (re-search-forward (if rfc1843-decode-hzp 94 rfc1843-hzp-word-regexp 95 rfc1843-word-regexp) (point-max) t) 96 ;;; Text with extents may cause XEmacs crash 97 (setq str (buffer-substring-no-properties 98 (match-beginning 1) 99 (match-end 1))) 100 (setq firstc (aref str 0)) 101 (insert (mm-decode-coding-string 102 (rfc1843-decode 103 (prog1 104 (substring str 1) 105 (delete-region (match-beginning 0) (match-end 0))) 106 firstc) 107 (if (eq firstc ?{) 'cn-gb-2312 'cn-big5)))) 108 (goto-char (point-min)) 109 (while (search-forward "~" (point-max) t) 110 (cond ((eq (char-after) ?\n) 111 (delete-char -1) 112 (delete-char 1)) 113 ((eq (char-after) ?~) 114 (delete-char 1))))))))) 115 116(defun rfc1843-decode-string (string) 117 "Decode HZ STRING and return the results." 118 (let ((m (mm-multibyte-p))) 119 (with-temp-buffer 120 (when m 121 (mm-enable-multibyte)) 122 (insert string) 123 (inline 124 (rfc1843-decode-region (point-min) (point-max))) 125 (buffer-string)))) 126 127(defun rfc1843-decode (word &optional firstc) 128 "Decode HZ WORD and return it." 129 (let ((i -1) (s (substring word 0)) v) 130 (if (or (not firstc) (eq firstc ?{)) 131 (while (< (incf i) (length s)) 132 (if (eq (setq v (aref s i)) ? ) nil 133 (aset s i (+ 128 v)))) 134 (while (< (incf i) (length s)) 135 (if (eq (setq v (aref s i)) ? ) nil 136 (setq v (+ (* 94 v) (aref s (1+ i)) -3135)) 137 (aset s i (+ (/ v 157) (if (eq firstc ?<) 201 161))) 138 (setq v (% v 157)) 139 (aset s (incf i) (+ v (if (< v 63) 64 98)))))) 140 s)) 141 142(defun rfc1843-decode-article-body () 143 "Decode HZ encoded text in the article body." 144 (if (string-match (concat "\\<\\(" rfc1843-newsgroups-regexp "\\)\\>") 145 (or gnus-newsgroup-name "")) 146 (save-excursion 147 (save-restriction 148 (message-narrow-to-head) 149 (let* ((inhibit-point-motion-hooks t) 150 (case-fold-search t) 151 (ct (message-fetch-field "Content-Type" t)) 152 (ctl (and ct (mail-header-parse-content-type ct)))) 153 (if (and ctl (not (string-match "/" (car ctl)))) 154 (setq ctl nil)) 155 (goto-char (point-max)) 156 (widen) 157 (forward-line 1) 158 (narrow-to-region (point) (point-max)) 159 (when (or (not ctl) 160 (equal (car ctl) "text/plain")) 161 (rfc1843-decode-region (point) (point-max)))))))) 162 163(defvar rfc1843-old-gnus-decode-header-function nil) 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;;; arch-tag: 5149c301-a6ca-4731-9c9d-ba616e2cb687 190;;; rfc1843.el ends here 191