lisp/international/utf-7.el

   1 ;;; utf-7.el --- utf-7 coding system
   2
   3 ;; Copyright (C) 2003  Free Software Foundation, Inc.
   4
   5 ;; Author: Dave Love <fx@gnu.org>
   6 ;; Keywords: i18n, mail
   7
   8 ;; This file is free software; you can redistribute it and/or modify
   9 ;; it under the terms of the GNU General Public License as published by
  10 ;; the Free Software Foundation; either version 2, or (at your option)
  11 ;; any later version.
  12
  13 ;; This file is distributed in the hope that it will be useful,
  14 ;; but WITHOUT ANY WARRANTY; without even the implied warranty of
  15 ;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  16 ;; GNU General Public License for more details.
  17
  18 ;; You should have received a copy of the GNU General Public License
  19 ;; along with GNU Emacs; see the file COPYING.  If not, write to
  20 ;; the Free Software Foundation, Inc., 59 Temple Place - Suite 330,
  21 ;; Boston, MA 02111-1307, USA.
  22
  23 ;;; Commentary:
  24
  25 ;; Defines a coding system for UTF-7, defined in RFC 2152.  Non-ASCII
  26 ;; segments are encoded as base64-encoded big endian UTF-16.  Also
  27 ;; defines a variation required for IMAP (RFC 2060).
  28
  29 ;; The encoding and decoding was originally taken from Jon K Hellan's
  30 ;; implementation in Gnus, but has been substantially re-done.
  31
  32 ;; This probably needs more attention.  In particular, it's not
  33 ;; completely consistent with iconv's behaviour.  It's arguable
  34 ;; whether the IMAP version should be a coding system since it's
  35 ;; apparently only used for IMAP mailbox names, so it's commented out.
  36
  37 ;;; Code:
  38
  39 (make-coding-system
  40  'utf-7 0 ?U
  41  "UTF-7 encoding of Unicode (RFC 2152)"
  42  nil
  43  `((safe-chars . ,(coding-system-get 'utf-16be 'safe-chars))
  44    (mime-charset . utf-7)
  45    (pre-write-conversion . utf-7-pre-write-conversion)
  46    (post-read-conversion . utf-7-post-read-conversion)))
  47
  48 ;; (make-coding-system
  49 ;;  'utf-7-imap 0 ?u
  50 ;;  "UTF-7 encoding of Unicode, IMAP version (RFC 2060)"
  51 ;;  nil
  52 ;;  `((safe-chars . ,(coding-system-get 'utf-16be 'safe-chars))
  53 ;;    (pre-write-conversion . utf-7-imap-pre-write-conversion)
  54 ;;    (post-read-conversion . utf-7-imap-post-read-conversion)))
  55
  56 (defun utf-7-decode (len imap)
  57   "Decode LEN bytes of UTF-7 at point.
  58 IMAP non-nil means use the IMAP version."
  59   (save-excursion
  60     (save-restriction
  61       (narrow-to-region (point) (+ (point) len))
  62       (let ((not-esc (if imap "^&" "^+"))
  63             (skip-chars (if imap "A-Za-z0-9+," "A-Za-z0-9+/")))
  64         (while (not (eobp))
  65           (skip-chars-forward not-esc)
  66           (unless (eobp)
  67             (forward-char)
  68             (let ((p (point))
  69                   (run-length (skip-chars-forward skip-chars)))
  70               (if (eq ?- (char-after))
  71                   (delete-char 1))
  72               (unless (= run-length 0)  ; encoded lone esc-char
  73                 (let ((pl (mod (- run-length) 4)))
  74                   (insert-char ?= pl)
  75                   (if imap
  76                       (subst-char-in-region p (point) ?, ?/))
  77                   (base64-decode-region p (point)))
  78                 (decode-coding-region p (point) 'utf-16be)
  79                 (save-excursion
  80                   (goto-char p)
  81                   (delete-backward-char 1)))))))
  82       (- (point-max) (point-min)))))
  83
  84 (defun utf-7-post-read-conversion (len)
  85   (utf-7-decode len nil))
  86
  87 ;; (defun utf-7-imap-post-read-conversion (len)
  88 ;;   (utf-7-decode len t))
  89
  90 (defun utf-7-encode (from to imap)
  91   "Encode bytes between FROM and TO to UTF-7.
  92 ESC and SKIP-CHARS are adjusted for the normal and IMAP versions."
  93   (let* ((old-buf (current-buffer))
  94          (esc (if imap ?& ?+))
  95          ;; These are characters which can be encoded asis.
  96          (skip-chars (if imap
  97                          "\t\n\r\x20-\x25\x27-\x7e" ; rfc2060
  98                        ;; This includes the rfc2152 optional set.
  99                        ;; Perhaps it shouldn't (like iconv).
 100                        "\t\n\r -*,-[]-}"))
 101          (not-skip-chars (format "^%s%c" skip-chars esc)))
 102     (set-buffer (generate-new-buffer " *temp*"))
 103     (if (stringp from)
 104         (insert from)
 105       (insert-buffer-substring old-buf from to))
 106     (goto-char (point-min))
 107     (while (not (eobp))
 108       (skip-chars-forward skip-chars)
 109       (if (eq ?+ (char-after))
 110           (progn (forward-char)
 111                  (insert ?-))
 112         (unless (eobp)
 113           (insert esc)
 114           (let ((p (point)))
 115             (skip-chars-forward not-skip-chars)
 116             (save-restriction
 117               ;; encode-coding-region doesn't preserve point
 118               (narrow-to-region p (point))
 119               (encode-coding-region p (point-max) 'utf-16be)
 120               (base64-encode-region p (point-max))
 121               (if imap
 122                   (subst-char-in-region p (point-max) ?/ ?,))
 123               (goto-char p)
 124               ;; As I read the RFC, this isn't correct, but it's
 125               ;; consistent with iconv, at least regarding `='.
 126               (skip-chars-forward "^= \t\n")
 127               (delete-region (point) (point-max))))
 128           (unless (eobp)
 129             (insert ?-)))))
 130     nil))
 131
 132 (defun utf-7-pre-write-conversion (from to)
 133   (utf-7-encode from to nil))
 134
 135 ;; (defun utf-7-imap-pre-write-conversion (from to)
 136 ;;   (utf-7-encode from to t))
 137
 138 (provide 'utf-7)
 139 ;;; utf-7.el ends here