;;; rfc2047.el --- functions for encoding and decoding rfc2047 messages
-;; Copyright (C) 1998, 1999, 2000, 2002, 2003, 2004, 2005
-;; Free Software Foundation, Inc.
+;; Copyright (C) 1998, 1999, 2000, 2001, 2002, 2003, 2004,
+;; 2005, 2006 Free Software Foundation, Inc.
;; Author: Lars Magne Ingebrigtsen <larsi@gnus.org>
;; MORIOKA Tomohiko <morioka@jaist.ac.jp>
(re-search-forward ":[ \t\n]*" nil t)
(buffer-substring-no-properties (point) (point-max)))))
+(defun rfc2047-quote-special-characters-in-quoted-strings (&optional
+ encodable-regexp)
+ "Quote special characters with `\\'s in quoted strings.
+Quoting will not be done in a quoted string if it contains characters
+matching ENCODABLE-REGEXP."
+ (goto-char (point-min))
+ (let ((tspecials (concat "[" ietf-drums-tspecials "]"))
+ beg end)
+ (with-syntax-table (standard-syntax-table)
+ (while (search-forward "\"" nil t)
+ (setq beg (match-beginning 0))
+ (unless (eq (char-before beg) ?\\)
+ (goto-char beg)
+ (setq beg (1+ beg))
+ (condition-case nil
+ (progn
+ (forward-sexp)
+ (setq end (1- (point)))
+ (goto-char beg)
+ (if (and encodable-regexp
+ (re-search-forward encodable-regexp end t))
+ (goto-char (1+ end))
+ (save-restriction
+ (narrow-to-region beg end)
+ (while (re-search-forward tspecials nil 'move)
+ (if (eq (char-before) ?\\)
+ (if (looking-at tspecials) ;; Already quoted.
+ (forward-char)
+ (insert "\\"))
+ (goto-char (match-beginning 0))
+ (insert "\\")
+ (forward-char))))
+ (forward-char)))
+ (error
+ (goto-char beg))))))))
+
(defvar rfc2047-encoding-type 'address-mime
"The type of encoding done by `rfc2047-encode-region'.
This should be dynamically bound around calls to
(while (not (eobp))
(save-restriction
(rfc2047-narrow-to-field)
+ (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))))
(if (not (rfc2047-encodable-p))
- (prog1
+ (prog2
+ (when (eq method 'address-mime)
+ (rfc2047-quote-special-characters-in-quoted-strings))
(if (and (eq (mm-body-7-or-8) '8bit)
(mm-multibyte-p)
(mm-coding-system-p
(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))))
(re-search-forward "^[^:]+: *" nil t)
(cond
((eq method 'address-mime)
(rfc2047-encode start (point))
(goto-char end))))
;; `address-mime' case -- take care of quoted words, comments.
+ (rfc2047-quote-special-characters-in-quoted-strings encodable-regexp)
(with-syntax-table rfc2047-syntax-table
(goto-char (point-min))
(condition-case err ; in case of unbalanced quotes
(defvar rfc2047-quote-decoded-words-containing-tspecials nil
"If non-nil, quote decoded words containing special characters.")
+(defvar rfc2047-allow-incomplete-encoded-text t
+ "*Non-nil means allow incomplete encoded-text in successive encoded-words.
+Dividing of encoded-text in the place other than character boundaries
+violates RFC2047 section 5, while we have a capability to decode it.
+If it is non-nil, the decoder will decode B- or Q-encoding in each
+encoded-word, concatenate them, and decode it by charset. Otherwise,
+the decoder will fully decode each encoded-word before concatenating
+them.")
+
+(defun rfc2047-strip-backslashes-in-quoted-strings ()
+ "Strip backslashes in quoted strings. `\\\"' remains."
+ (goto-char (point-min))
+ (let (beg)
+ (with-syntax-table (standard-syntax-table)
+ (while (search-forward "\"" nil t)
+ (unless (eq (char-before) ?\\)
+ (setq beg (match-end 0))
+ (goto-char (match-beginning 0))
+ (condition-case nil
+ (progn
+ (forward-sexp)
+ (save-restriction
+ (narrow-to-region beg (1- (point)))
+ (goto-char beg)
+ (while (search-forward "\\" nil 'move)
+ (unless (memq (char-after) '(?\"))
+ (delete-backward-char 1))
+ (forward-char)))
+ (forward-char))
+ (error
+ (goto-char beg))))))))
+
+(defun rfc2047-charset-to-coding-system (charset)
+ "Return coding-system corresponding to MIME CHARSET.
+If your Emacs implementation can't decode CHARSET, return nil."
+ (when (stringp charset)
+ (setq charset (intern (downcase charset))))
+ (when (or (not charset)
+ (eq 'gnus-all mail-parse-ignored-charsets)
+ (memq 'gnus-all mail-parse-ignored-charsets)
+ (memq charset mail-parse-ignored-charsets))
+ (setq charset mail-parse-charset))
+ (let ((cs (mm-charset-to-coding-system charset)))
+ (cond ((eq cs 'ascii)
+ (setq cs (or (mm-charset-to-coding-system mail-parse-charset)
+ 'raw-text)))
+ ((mm-coding-system-p cs))
+ ((and charset
+ (listp mail-parse-ignored-charsets)
+ (memq 'gnus-unknown mail-parse-ignored-charsets))
+ (setq cs (mm-charset-to-coding-system mail-parse-charset))))
+ (if (eq cs 'ascii)
+ 'raw-text
+ cs)))
+
+(defun rfc2047-decode-encoded-words (words)
+ "Decode successive encoded-words in WORDS and return a decoded string.
+Each element of WORDS looks like (CHARSET ENCODING ENCODED-TEXT
+ENCODED-WORD)."
+ (let (word charset cs encoding text rest)
+ (while words
+ (setq word (pop words))
+ (if (and (setq cs (rfc2047-charset-to-coding-system
+ (setq charset (car word))))
+ (condition-case code
+ (cond ((char-equal ?B (nth 1 word))
+ (setq text (base64-decode-string
+ (rfc2047-pad-base64 (nth 2 word)))))
+ ((char-equal ?Q (nth 1 word))
+ (setq text (quoted-printable-decode-string
+ (mm-subst-char-in-string
+ ?_ ? (nth 2 word) t)))))
+ (error
+ (message "%s" (error-message-string code))
+ nil)))
+ (if (and rfc2047-allow-incomplete-encoded-text
+ (eq cs (caar rest)))
+ ;; Concatenate text of which the charset is the same.
+ (setcdr (car rest) (concat (cdar rest) text))
+ (push (cons cs text) rest))
+ ;; Don't decode encoded-word.
+ (push (cons nil (nth 3 word)) rest)))
+ (while rest
+ (setq words (concat
+ (or (and (setq cs (caar rest))
+ (condition-case code
+ (mm-decode-coding-string (cdar rest) cs)
+ (error
+ (message "%s" (error-message-string code))
+ nil)))
+ (concat (when (cdr rest) " ")
+ (cdar rest)
+ (when (and words
+ (not (eq (string-to-char words) ? )))
+ " ")))
+ words)
+ rest (cdr rest)))
+ words))
+
;; Fixme: This should decode in place, not cons intermediate strings.
;; Also check whether it needs to worry about delimiting fields like
;; encoding.
;; and worthwhile (is it more correct or not?), e.g. something like
;; `=?iso-8859-1?q?foo?=@'.
-(defun rfc2047-decode-region (start end)
- "Decode MIME-encoded words in region between START and END."
+(defun rfc2047-decode-region (start end &optional address-mime)
+ "Decode MIME-encoded words in region between START and END.
+If ADDRESS-MIME is non-nil, strip backslashes which precede characters
+other than `\"' and `\\' in quoted strings."
(interactive "r")
(let ((case-fold-search t)
- b e)
+ (eword-regexp (eval-when-compile
+ ;; Ignore whitespace between encoded-words.
+ (concat "[\n\t ]*\\(" rfc2047-encoded-word-regexp
+ "\\)")))
+ b e match words)
(save-excursion
(save-restriction
(narrow-to-region start end)
- (goto-char (point-min))
- ;; Remove whitespace between encoded words.
- (while (re-search-forward
- (eval-when-compile
- (concat "\\(" rfc2047-encoded-word-regexp "\\)"
- "\\(\n?[ \t]\\)+"
- "\\(" rfc2047-encoded-word-regexp "\\)"))
- nil t)
- (delete-region (goto-char (match-end 1)) (match-beginning 7)))
- ;; Decode the encoded words.
- (setq b (goto-char (point-min)))
- (while (re-search-forward rfc2047-encoded-word-regexp nil t)
- (setq e (match-beginning 0))
- (insert (rfc2047-parse-and-decode
- (prog1
- (match-string 0)
- (delete-region e (match-end 0)))))
- (while (looking-at rfc2047-encoded-word-regexp)
- (insert (rfc2047-parse-and-decode
- (prog1
- (match-string 0)
- (delete-region (point) (match-end 0))))))
+ (when address-mime
+ (rfc2047-strip-backslashes-in-quoted-strings))
+ (goto-char (setq b start))
+ ;; Look for the encoded-words.
+ (while (setq match (re-search-forward eword-regexp nil t))
+ (setq e (match-beginning 1)
+ end (match-end 0)
+ words nil)
+ (while match
+ (push (list (match-string 2) ;; charset
+ (char-after (match-beginning 4)) ;; encoding
+ (match-string 5) ;; encoded-text
+ (match-string 1)) ;; encoded-word
+ words)
+ ;; Look for the subsequent encoded-words.
+ (when (setq match (looking-at eword-regexp))
+ (goto-char (setq end (match-end 0)))))
+ ;; Replace the encoded-words with the decoded one.
+ (delete-region e end)
+ (insert (rfc2047-decode-encoded-words (nreverse words)))
(save-restriction
(narrow-to-region e (point))
(goto-char e)
(not (eq mail-parse-charset 'gnus-decoded)))
(mm-decode-coding-region b (point-max) mail-parse-charset))))))
-(defun rfc2047-decode-string (string)
- "Decode the quoted-printable-encoded STRING and return the results."
+(defun rfc2047-decode-address-region (start end)
+ "Decode MIME-encoded words in region between START and END.
+Backslashes which precede characters other than `\"' and `\\' in quoted
+strings are stripped."
+ (rfc2047-decode-region start end t))
+
+(defun rfc2047-decode-string (string &optional address-mime)
+ "Decode MIME-encoded STRING and return the result.
+If ADDRESS-MIME is non-nil, strip backslashes which precede characters
+other than `\"' and `\\' in quoted strings."
(let ((m (mm-multibyte-p)))
(if (string-match "=\\?" string)
(with-temp-buffer
(mm-enable-multibyte))
(insert string)
(inline
- (rfc2047-decode-region (point-min) (point-max)))
+ (rfc2047-decode-region (point-min) (point-max) address-mime))
(buffer-string))
+ (when address-mime
+ (setq string
+ (with-temp-buffer
+ (when (mm-multibyte-string-p string)
+ (mm-enable-multibyte))
+ (insert string)
+ (rfc2047-strip-backslashes-in-quoted-strings)
+ (buffer-string))))
;; Fixme: As above, `m' here is inappropriate.
(if (and m
mail-parse-charset
(mm-decode-coding-string string mail-parse-charset))
(mm-string-as-multibyte string)))))
-(defun rfc2047-parse-and-decode (word)
- "Decode WORD and return it if it is an encoded word.
-Return WORD if it is not not an encoded word or if the charset isn't
-decodable."
- (if (not (string-match rfc2047-encoded-word-regexp word))
- word
- (or
- (condition-case nil
- (rfc2047-decode
- (match-string 1 word)
- (string-to-char (match-string 3 word))
- (match-string 4 word))
- (error word))
- word))) ; un-decodable
+(defun rfc2047-decode-address-string (string)
+ "Decode MIME-encoded STRING and return the result.
+Backslashes which precede characters other than `\"' and `\\' in quoted
+strings are stripped."
+ (rfc2047-decode-string string t))
(defun rfc2047-pad-base64 (string)
"Pad STRING to quartets."
(2 (concat string "=="))
(3 (concat string "=")))))
-(defun rfc2047-decode (charset encoding string)
- "Decode STRING from the given MIME CHARSET in the given ENCODING.
-Valid ENCODINGs are the characters \"B\" and \"Q\".
-If your Emacs implementation can't decode CHARSET, return nil."
- (if (stringp charset)
- (setq charset (intern (downcase charset))))
- (if (or (not charset)
- (eq 'gnus-all mail-parse-ignored-charsets)
- (memq 'gnus-all mail-parse-ignored-charsets)
- (memq charset mail-parse-ignored-charsets))
- (setq charset mail-parse-charset))
- (let ((cs (mm-charset-to-coding-system charset)))
- (if (and (not cs) charset
- (listp mail-parse-ignored-charsets)
- (memq 'gnus-unknown mail-parse-ignored-charsets))
- (setq cs (mm-charset-to-coding-system mail-parse-charset)))
- (when cs
- (when (eq cs 'ascii)
- (setq cs (or mail-parse-charset 'raw-text)))
- (mm-decode-coding-string
- (cond
- ((char-equal ?B encoding)
- (base64-decode-string
- (rfc2047-pad-base64 string)))
- ((char-equal ?Q encoding)
- (quoted-printable-decode-string
- (mm-subst-char-in-string ?_ ? string t)))
- (t (error "Invalid encoding: %c" encoding)))
- cs))))
-
(provide 'rfc2047)
;;; arch-tag: a07fe3d4-22b5-4c4a-bd89-b1f82d5d36f6