-;;; rfc2047.el --- Functions for encoding and decoding rfc2047 messages
-;; Copyright (C) 1998-2000 Free Software Foundation, Inc.
+;;; rfc2047.el --- functions for encoding and decoding rfc2047 messages
+;; Copyright (C) 1998, 1999, 2000, 2002, 2003 Free Software Foundation, Inc.
;; Author: Lars Magne Ingebrigtsen <larsi@gnus.org>
;; MORIOKA Tomohiko <morioka@jaist.ac.jp>
;;; Commentary:
+;; RFC 2047 is "MIME (Multipurpose Internet Mail Extensions) Part
+;; Three: Message Header Extensions for Non-ASCII Text".
+
;;; Code:
-(eval-and-compile
- (eval
- '(unless (fboundp 'base64-decode-string)
- (require 'base64))))
+(eval-when-compile
+ (require 'cl)
+ (defvar message-posting-charset))
(require 'qp)
(require 'mm-util)
-(require 'ietf-drums)
+;; Fixme: Avoid this (used for mail-parse-charset) mm dependence on gnus.
(require 'mail-prsvr)
+(require 'base64)
+(autoload 'mm-body-7-or-8 "mm-bodies")
(defvar rfc2047-header-encoding-alist
'(("Newsgroups" . nil)
+ ("Followup-To" . nil)
("Message-ID" . nil)
+ ("\\(Resent-\\)?\\(From\\|Cc\\|To\\|Bcc\\|Reply-To\\|Sender\
+\\|Mail-Followup-To\\|Mail-Copies-To\\|Approved\\)" . address-mime)
(t . mime))
"*Header/encoding method alist.
The list is traversed sequentially. The keys can either be
-header regexps or `t'.
+header regexps or t.
The values can be:
1) nil, in which case no encoding is done;
2) `mime', in which case the header will be encoded according to RFC2047;
-3) a charset, in which case it will be encoded as that charse;
-4) `default', in which case the field will be encoded as the rest
+3) `address-mime', like `mime', but takes account of the rules for address
+ fields (where quoted strings and comments must be treated separately);
+4) a charset, in which case it will be encoded as that charset;
+5) `default', in which case the field will be encoded as the rest
of the article.")
(defvar rfc2047-charset-encoding-alist
(iso-8859-4 . Q)
(iso-8859-5 . B)
(koi8-r . B)
- (iso-8859-7 . Q)
- (iso-8859-8 . Q)
+ (iso-8859-7 . B)
+ (iso-8859-8 . B)
(iso-8859-9 . Q)
+ (iso-8859-14 . Q)
+ (iso-8859-15 . Q)
(iso-2022-jp . B)
(iso-2022-kr . B)
(gb2312 . B)
+ (big5 . B)
+ (cn-big5 . B)
(cn-gb . B)
(cn-gb-2312 . B)
(euc-kr . B)
(iso-2022-jp-2 . B)
- (iso-2022-int-1 . B))
+ (iso-2022-int-1 . B)
+ (viscii . Q))
"Alist of MIME charsets to RFC2047 encodings.
-Valid encodings are nil, `Q' and `B'.")
+Valid encodings are nil, `Q' and `B'. These indicate binary (no) encoding,
+quoted-printable and base64 respectively.")
(defvar rfc2047-encoding-function-alist
'((Q . rfc2047-q-encode-region)
(nil . ignore))
"Alist of RFC2047 encodings to encoding functions.")
-(defvar rfc2047-q-encoding-alist
- '(("\\(From\\|Cc\\|To\\|Bcc\||Reply-To\\):" . "-A-Za-z0-9!*+/=_")
- ("." . "^\000-\007\013\015-\037\200-\377=_?"))
- "Alist of header regexps and valid Q characters.")
-
;;;
;;; Functions for encoding RFC2047 messages
;;;
+(defun rfc2047-qp-or-base64 ()
+ "Return the type with which to encode the buffer.
+This is either `base64' or `quoted-printable'."
+ (save-excursion
+ (let ((limit (min (point-max) (+ 2000 (point-min))))
+ (n8bit 0))
+ (goto-char (point-min))
+ (skip-chars-forward "\x20-\x7f\r\n\t" limit)
+ (while (< (point) limit)
+ (incf n8bit)
+ (forward-char 1)
+ (skip-chars-forward "\x20-\x7f\r\n\t" limit))
+ (if (or (< (* 6 n8bit) (- limit (point-min)))
+ ;; Don't base64, say, a short line with a single
+ ;; non-ASCII char when splitting parts by charset.
+ (= n8bit 1))
+ 'quoted-printable
+ 'base64))))
+
(defun rfc2047-narrow-to-field ()
"Narrow the buffer to the header on the current line."
(beginning-of-line)
(progn
(forward-line 1)
(if (re-search-forward "^[^ \n\t]" nil t)
- (progn
- (beginning-of-line)
- (point))
+ (point-at-bol)
(point-max))))
(goto-char (point-min)))
+(defun rfc2047-field-value ()
+ "Return the value of the field at point."
+ (save-excursion
+ (save-restriction
+ (rfc2047-narrow-to-field)
+ (re-search-forward ":[ \t\n]*" nil t)
+ (buffer-substring (point) (point-max)))))
+
+(defvar rfc2047-encoding-type 'address-mime
+ "The type of encoding done by `rfc2047-encode-region'.
+This should be dynamically bound around calls to
+`rfc2047-encode-region' to either `mime' or `address-mime'. See
+`rfc2047-header-encoding-alist', for definitions.")
+
(defun rfc2047-encode-message-header ()
"Encode the message header according to `rfc2047-header-encoding-alist'.
Should be called narrowed to the head of the message."
(interactive "*")
(save-excursion
(goto-char (point-min))
- (let ((alist rfc2047-header-encoding-alist)
- elem method)
+ (let (alist elem method)
(while (not (eobp))
(save-restriction
(rfc2047-narrow-to-field)
- (when (rfc2047-encodable-p)
+ (if (not (rfc2047-encodable-p))
+ (prog1
+ (if (and (eq (mm-body-7-or-8) '8bit)
+ (mm-multibyte-p)
+ (mm-coding-system-p
+ (car message-posting-charset)))
+ ;; 8 bit must be decoded.
+ (mm-encode-coding-region
+ (point-min) (point-max)
+ (mm-charset-to-coding-system
+ (car message-posting-charset))))
+ ;; No encoding necessary, but folding is nice
+ (rfc2047-fold-region
+ (save-excursion
+ (goto-char (point-min))
+ (skip-chars-forward "^:")
+ (when (looking-at ": ")
+ (forward-char 2))
+ (point))
+ (point-max)))
;; We found something that may perhaps be encoded.
+ (setq method nil
+ alist rfc2047-header-encoding-alist)
(while (setq elem (pop alist))
(when (or (and (stringp (car elem))
(looking-at (car elem)))
(eq (car elem) t))
(setq alist nil
method (cdr elem))))
+ (goto-char (point-min))
+ (re-search-forward "^[^:]+: *" nil t)
(cond
+ ((eq method 'address-mime)
+ (rfc2047-encode-region (point) (point-max)))
((eq method 'mime)
- (rfc2047-encode-region (point-min) (point-max))
- (rfc2047-fold-region (point-min) (point-max)))
+ (let ((rfc2047-encoding-type 'mime))
+ (rfc2047-encode-region (point) (point-max))))
+ ((eq method 'default)
+ (if (and (featurep 'mule)
+ (if (boundp 'default-enable-multibyte-characters)
+ default-enable-multibyte-characters)
+ mail-parse-charset)
+ (mm-encode-coding-region (point) (point-max)
+ mail-parse-charset)))
+ ;; We get this when CC'ing messsages to newsgroups with
+ ;; 8-bit names. The group name mail copy just got
+ ;; unconditionally encoded. Previously, it would ask
+ ;; whether to encode, which was quite confusing for the
+ ;; user. If the new behaviour is wrong, tell me. I have
+ ;; left the old code commented out below.
+ ;; -- Per Abrahamsen <abraham@dina.kvl.dk> Date: 2001-10-07.
+ ;; Modified by Dave Love, with the commented-out code changed
+ ;; in accordance with changes elsewhere.
+ ((null method)
+ (rfc2047-encode-region (point) (point-max)))
+;;; ((null method)
+;;; (if (or (message-options-get
+;;; 'rfc2047-encode-message-header-encode-any)
+;;; (message-options-set
+;;; 'rfc2047-encode-message-header-encode-any
+;;; (y-or-n-p
+;;; "Some texts are not encoded. Encode anyway?")))
+;;; (rfc2047-encode-region (point-min) (point-max))
+;;; (error "Cannot send unencoded text")))
+ ((mm-coding-system-p method)
+ (if (and (featurep 'mule)
+ (if (boundp 'default-enable-multibyte-characters)
+ default-enable-multibyte-characters))
+ (mm-encode-coding-region (point) (point-max) method)))
;; Hm.
(t)))
- (goto-char (point-max)))))
- (when mail-parse-charset
- (encode-coding-region
- (point-min) (point-max) mail-parse-charset))))
+ (goto-char (point-max)))))))
+
+;; Fixme: This, and the require below may not be the Right Thing, but
+;; should be safe just before release. -- fx 2001-02-08
+(eval-when-compile (defvar message-posting-charset))
-(defun rfc2047-encodable-p (&optional header)
- "Say whether the current (narrowed) buffer contains characters that need encoding in headers."
+(defun rfc2047-encodable-p ()
+ "Return non-nil if any characters in current buffer need encoding in headers.
+The buffer may be narrowed."
+ (require 'message) ; for message-posting-charset
(let ((charsets
- (mapcar
- 'mm-mime-charset
- (mm-find-charset-region (point-min) (point-max))))
- (cs (list 'us-ascii (car message-posting-charset)))
- found)
- (while charsets
- (unless (memq (pop charsets) cs)
- (setq found t)))
- found))
-
-(defun rfc2047-dissect-region (b e)
- "Dissect the region between B and E into words."
- (let ((all-specials (concat ietf-drums-tspecials " \t\n\r"))
- (special-list (mapcar 'identity ietf-drums-tspecials))
- (blank-list '(? ?\t ?\n ?\r))
- words current cs state mail-parse-mule-charset)
- (save-restriction
- (narrow-to-region b e)
- (goto-char (point-min))
- (skip-chars-forward all-specials)
- (setq b (point))
- (while (not (eobp))
- (cond
- ((not state)
- (if (memq (char-after) blank-list)
- (setq state 'blank)
- (setq state 'word)
- (if (not (eq (setq cs (mm-charset-after)) 'ascii))
- (setq current cs)))
- (setq b (point)))
- ((eq state 'blank)
- (cond
- ((memq (char-after) special-list)
- (setq state nil))
- ((memq (char-after) blank-list))
- (t
- (setq state 'word)
- (if (not (eq (setq cs (mm-charset-after)) 'ascii))
- (setq current cs)))))
- ((eq state 'word)
- (cond
- ((memq (char-after) special-list)
- (setq state nil)
- (push (list b (point) current) words)
- (setq current nil))
- ((memq (char-after) blank-list)
- (setq state 'blank)
- (push (list b (point) current) words)
- (setq current nil)
- (setq b (point)))
- ((or (eq (setq cs (mm-charset-after)) 'ascii)
- (if current
- (eq current cs)
- (setq current cs))))
-