;;; mm-bodies.el --- Functions for decoding MIME things
-;; Copyright (C) 1998,99 Free Software Foundation, Inc.
+
+;; Copyright (C) 1998, 1999, 2000, 2001, 2003
+;; Free Software Foundation, Inc.
;; Author: Lars Magne Ingebrigtsen <larsi@gnus.org>
;; MORIOKA Tomohiko <morioka@jaist.ac.jp>
(eval-and-compile
(or (fboundp 'base64-decode-region)
- (require 'base64))
- (autoload 'binhex-decode-region "binhex"))
+ (require 'base64)))
+
+(eval-when-compile
+ (defvar mm-uu-decode-function)
+ (defvar mm-uu-binhex-decode-function))
(require 'mm-util)
(require 'rfc2047)
-(require 'qp)
-(require 'uudecode)
+(require 'mm-encode)
;; 8bit treatment gets any char except: 0x32 - 0x7f, CR, LF, TAB, BEL,
;; BS, vertical TAB, form feed, and ^_
-(defvar mm-8bit-char-regexp "[^\x20-\x7f\r\n\t\x7\x8\xb\xc\x1f]")
-
-(defvar mm-body-charset-encoding-alist
- '((us-ascii . 7bit)
- (iso-8859-1 . quoted-printable)
- (iso-8859-2 . quoted-printable)
- (iso-8859-3 . quoted-printable)
- (iso-8859-4 . quoted-printable)
- (iso-8859-5 . base64)
- (koi8-r . 8bit)
- (iso-8859-7 . quoted-printable)
- (iso-8859-8 . quoted-printable)
- (iso-8859-9 . quoted-printable)
- (iso-2022-jp . base64)
- (iso-2022-kr . base64)
- (gb2312 . base64)
- (cn-gb . base64)
- (cn-gb-2312 . base64)
- (euc-kr . 8bit)
- (iso-2022-jp-2 . base64)
- (iso-2022-int-1 . base64))
+(defvar mm-7bit-chars "\x20-\x7f\r\n\t\x7\x8\xb\xc\x1f")
+
+(defcustom mm-body-charset-encoding-alist
+ '((iso-2022-jp . 7bit)
+ (iso-2022-jp-2 . 7bit)
+ ;; We MUST encode UTF-16 because it can contain \0's which is
+ ;; known to break servers.
+ (utf-16 . base64)
+ (utf-16be . base64)
+ (utf-16le . base64))
"Alist of MIME charsets to encodings.
-Valid encodings are `7bit', `8bit', `quoted-printable' and `base64'.")
+Valid encodings are `7bit', `8bit', `quoted-printable' and `base64'."
+ :type '(repeat (cons (symbol :tag "charset")
+ (choice :tag "encoding"
+ (const 7bit)
+ (const 8bit)
+ (const quoted-printable)
+ (const base64))))
+ :group 'mime)
-(defun mm-encode-body ()
+(defun mm-encode-body (&optional charset)
"Encode a body.
Should be called narrowed to the body that is to be encoded.
If there is more than one non-ASCII MULE charset, then list of found
MULE charsets are returned.
+If CHARSET is non-nil, it is used.
If successful, the MIME charset is returned.
If no encoding was done, nil is returned."
- (if (not (featurep 'mule))
+ (if (not (mm-multibyte-p))
;; In the non-Mule case, we search for non-ASCII chars and
- ;; return the value of `mm-default-charset' if any are found.
- (save-excursion
- (goto-char (point-min))
- (if (re-search-forward "[^\x0-\x7f]" nil t)
- (or mail-parse-charset
- (mm-read-charset "Charset used in the article: "))
- ;; The logic in `mml-generate-mime-1' confirms that it's OK
- ;; to return nil here.
- nil))
+ ;; return the value of `mail-parse-charset' if any are found.
+ (or charset
+ (save-excursion
+ (goto-char (point-min))
+ (if (re-search-forward "[^\x0-\x7f]" nil t)
+ (or mail-parse-charset
+ (message-options-get 'mm-encody-body-charset)
+ (message-options-set
+ 'mm-encody-body-charset
+ (mm-read-charset "Charset used in the article: ")))
+ ;; The logic in `mml-generate-mime-1' confirms that it's OK
+ ;; to return nil here.
+ nil)))
(save-excursion
- (goto-char (point-min))
- (let ((charsets (mm-find-mime-charset-region (point-min) (point-max)))
+ (if charset
+ (progn
+ (mm-encode-coding-region (point-min) (point-max) charset)
charset)
- (cond
- ;; No encoding.
- ((null charsets)
- nil)
- ;; Too many charsets.
- ((> (length charsets) 1)
- charsets)
- ;; We encode.
- (t
- (let ((charset (car charsets))
- start)
- (when (or t
- ;; We always decode.
- (not (mm-coding-system-equal
- charset buffer-file-coding-system)))
- (while (not (eobp))
- (if (eq (char-charset (char-after)) 'ascii)
- (when start
- (save-restriction
- (narrow-to-region start (point))
- (mm-encode-coding-region start (point) charset)
- (goto-char (point-max)))
- (setq start nil))
- (unless start
- (setq start (point))))
- (forward-char 1))
- (when start
- (mm-encode-coding-region start (point) charset)
- (setq start nil)))
- charset)))))))
-
-(defun mm-body-encoding (charset)
+ (goto-char (point-min))
+ (let ((charsets (mm-find-mime-charset-region (point-min) (point-max)
+ mm-hack-charsets)))
+ (cond
+ ;; No encoding.
+ ((null charsets)
+ nil)
+ ;; Too many charsets.
+ ((> (length charsets) 1)
+ charsets)
+ ;; We encode.
+ (t
+ (prog1
+ (setq charset (car charsets))
+ (mm-encode-coding-region (point-min) (point-max)
+ (mm-charset-to-coding-system charset))))
+ ))))))
+
+(defun mm-long-lines-p (length)
+ "Say whether any of the lines in the buffer is longer than LINES."
+ (save-excursion
+ (goto-char (point-min))
+ (end-of-line)
+ (while (and (not (eobp))
+ (not (> (current-column) length)))
+ (forward-line 1)
+ (end-of-line))
+ (and (> (current-column) length)
+ (current-column))))
+
+(defvar message-posting-charset)
+
+(defun mm-body-encoding (charset &optional encoding)
"Do Content-Transfer-Encoding and return the encoding of the current buffer."
- (let ((bits (mm-body-7-or-8)))
+ (when (stringp encoding)
+ (setq encoding (intern (downcase encoding))))
+ (let ((bits (mm-body-7-or-8))
+ (longp (mm-long-lines-p 1000)))
+ (require 'message)
(cond
- ((eq bits '7bit)
+ ((and (not mm-use-ultra-safe-encoding)
+ (not longp)
+ (eq bits '7bit))
bits)
- ((eq charset mail-parse-charset)
+ ((and (not mm-use-ultra-safe-encoding)
+ (not longp)
+ (not (cdr (assq charset mm-body-charset-encoding-alist)))
+ (or (eq t (cdr message-posting-charset))
+ (memq charset (cdr message-posting-charset))
+ (eq charset mail-parse-charset)))
bits)
(t
- (let ((encoding (or (cdr (assq charset mm-body-charset-encoding-alist))
- 'quoted-printable)))
+ (let ((encoding (or encoding
+ (cdr (assq charset mm-body-charset-encoding-alist))
+ (mm-qp-or-base64))))
+ (when mm-use-ultra-safe-encoding
+ (setq encoding (mm-safer-encoding encoding)))
(mm-encode-content-transfer-encoding encoding "text/plain")
encoding)))))
((not (featurep 'mule))
(if (save-excursion
(goto-char (point-min))
- (re-search-forward mm-8bit-char-regexp nil t))
- '8bit
- '7bit))
+ (skip-chars-forward mm-7bit-chars)
+ (eobp))
+ '7bit
+ '8bit))
(t
;; Mule version
(if (and (null (delq 'ascii
;;!!!Emacs 20.3. Sometimes.
(save-excursion
(goto-char (point-min))
- (skip-chars-forward "\0-\177")
+ (skip-chars-forward mm-7bit-chars)
(eobp)))
'7bit
'8bit))))
;;;
(defun mm-decode-content-transfer-encoding (encoding &optional type)
+ "Decodes buffer encoded with ENCODING, returning success status.
+If TYPE is `text/plain' CRLF->LF translation may occur."
(prog1
(condition-case error
(cond
((eq encoding 'quoted-printable)
- (quoted-printable-decode-region (point-min) (point-max)))
+ (quoted-printable-decode-region (point-min) (point-max))
+ t)
((eq encoding 'base64)
- (base64-decode-region (point-min)
- ;; Some mailers insert whitespace
- ;; junk at the end which
- ;; base64-decode-region dislikes.
- (save-excursion
- (goto-char (point-max))
- (skip-chars-backward "\n\t ")
- (delete-region (point) (point-max))
- (point))))
+ (base64-decode-region
+ (point-min)
+ ;; Some mailers insert whitespace
+ ;; junk at the end which
+ ;; base64-decode-region dislikes.
+ ;; Also remove possible junk which could
+ ;; have been added by mailing list software.
+ (save-excursion
+ (goto-char (point-min))
+ (while (re-search-forward "^[\t ]*\r?\n" nil t)
+ (delete-region (match-beginning 0) (match-end 0)))
+ (goto-char (point-max))
+ (when (re-search-backward "^[A-Za-z0-9+/]+=*[\t ]*$" nil t)
+ (forward-line))
+ (point))))
((memq encoding '(7bit 8bit binary))
- )
+ ;; Do nothing.
+ t)
((null encoding)
- )
+ ;; Do nothing.
+ t)
((memq encoding '(x-uuencode x-uue))
- (funcall mm-uu-decode-function (point-min) (point-max)))
+ (require 'mm-uu)
+ (funcall mm-uu-decode-function (point-min) (point-max))
+ t)
((eq encoding 'x-binhex)
- (funcall mm-uu-binhex-decode-function (point-min) (point-max)))
+ (require 'mm-uu)
+ (funcall mm-uu-binhex-decode-function (point-min) (point-max))
+ t)
+ ((eq encoding 'x-yenc)
+ (require 'mm-uu)
+ (funcall mm-uu-yenc-decode-function (point-min) (point-max))
+ )
((functionp encoding)
- (funcall encoding (point-min) (point-max)))
+ (funcall encoding (point-min) (point-max))
+ t)
(t
(message "Unknown encoding %s; defaulting to 8bit" encoding)))
(error
(message "Error while decoding: %s" error)
nil))
(when (and
- (memq encoding '(base64 x-uuencode x-uue x-binhex))
+ (memq encoding '(base64 x-uuencode x-uue x-binhex x-yenc))
(equal type "text/plain"))
(goto-char (point-min))
(while (search-forward "\r\n" nil t)
(replace-match "\n" t t)))))
-(defun mm-decode-body (charset &optional encoding type)
+(defun mm-decode-body (charset &optional encoding type force)
"Decode the current article that has been encoded with ENCODING.
-The characters in CHARSET should then be decoded."
- (if (stringp charset)
- (setq charset (intern (downcase charset))))
- (if (or (not charset) (memq charset mail-parse-ignored-charsets))
- (setq charset mail-parse-charset))
- (save-excursion
- (when encoding
- (mm-decode-content-transfer-encoding encoding type))
- (when (featurep 'mule)
- (let (mule-charset)
- (when (and charset
- (setq mule-charset (mm-charset-to-coding-system charset))
- ;; buffer-file-coding-system
- ;;Article buffer is nil coding system
- ;;in XEmacs
- enable-multibyte-characters
- (or (not (eq mule-charset 'ascii))
- (setq mule-charset mail-parse-charset)))
- (mm-decode-coding-region (point-min) (point-max) mule-charset))))))
+The characters in CHARSET should then be decoded. If FORCE is non-nil
+use the supplied charset unconditionally."
+ (let ((charset-supplied charset))
+ (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
+ charset-supplied nil))
+ (save-excursion
+ (when encoding
+ (mm-decode-content-transfer-encoding encoding type))
+ (when (featurep 'mule)
+ (let ((coding-system (mm-charset-to-coding-system charset)))
+ (if (and (not coding-system)
+ (listp mail-parse-ignored-charsets)
+ (memq 'gnus-unknown mail-parse-ignored-charsets))
+ (setq coding-system
+ (mm-charset-to-coding-system mail-parse-charset)))
+ (when (and charset coding-system
+ ;; buffer-file-coding-system
+ ;;Article buffer is nil coding system
+ ;;in XEmacs
+ (mm-multibyte-p)
+ (or (not (eq coding-system 'ascii))
+ (setq coding-system mail-parse-charset))
+ (not (eq coding-system 'gnus-decoded)))
+ (if (or force
+ ;; If a charset was supplied and `code-pages'
+ ;; hasn't been loaded, then use the supplied
+ ;; charset unconditionally.
+ (and (not (featurep 'code-pages))
+ charset-supplied))
+ (mm-decode-coding-region (point-min) (point-max)
+ coding-system)
+ ;; Otherwise allow Emacs to auto-detect the charset.
+ ;; Messages in windows-125x are frequently incorrectly
+ ;; advertised as iso-8859-x -- if `code-pages' has been
+ ;; loaded and configured, we're able to deal with them
+ ;; properly.
+ (mm-decode-coding-region-safely (point-min) (point-max)
+ coding-system)))
+ (setq buffer-file-coding-system
+ (if (boundp 'last-coding-system-used)
+ (symbol-value 'last-coding-system-used)
+ coding-system)))))))
+
+(defun mm-decode-coding-region-safely (start end coding-system)
+ "Decode region between START and END with CODING-SYSTEM.
+If CODING-SYSTEM is not a valid coding system for the text, let Emacs
+decide which coding system to use."
+ (let* ((orig (buffer-substring start end))
+ charsets)
+ (save-restriction
+ (narrow-to-region start end)
+ (mm-decode-coding-region (point-min) (point-max) coding-system)
+ (setq charsets (find-charset-region (point-min) (point-max)))
+ (when (or (memq 'eight-bit-control charsets)
+ (memq 'eight-bit-graphic charsets))
+ (delete-region (point-min) (point-max))
+ (insert orig)
+ (mm-decode-coding-region (point-min) (point-max) 'undecided)))))
(defun mm-decode-string (string charset)
"Decode STRING with CHARSET."
- (if (stringp charset)
+ (when (stringp charset)
(setq charset (intern (downcase charset))))
- (if (or (not charset) (memq charset mail-parse-ignored-charsets))
- (setq charset mail-parse-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))
(or
(when (featurep 'mule)
- (let (mule-charset)
- (when (and charset
- (setq mule-charset (mm-charset-to-coding-system charset))
- enable-multibyte-characters
- (or (not (eq mule-charset 'ascii))
- (setq mule-charset mail-parse-charset)))
- (mm-decode-coding-string string mule-charset))))
+ (let ((coding-system (mm-charset-to-coding-system charset)))
+ (if (and (not coding-system)
+ (listp mail-parse-ignored-charsets)
+ (memq 'gnus-unknown mail-parse-ignored-charsets))
+ (setq coding-system
+ (mm-charset-to-coding-system mail-parse-charset)))
+ (when (and charset coding-system
+ (mm-multibyte-p)
+ (or (not (eq coding-system 'ascii))
+ (setq coding-system mail-parse-charset)))
+ (mm-decode-coding-string string coding-system))))
string))
(provide 'mm-bodies)
-;; mm-bodies.el ends here
+;;; mm-bodies.el ends here