X-Git-Url: http://git.chise.org/gitweb/?a=blobdiff_plain;f=lisp%2Frfc2047.el;h=10eff841e3b118d4801b3c56667e7882bfb5e198;hb=3738187cad20787b5b99c4061256e30e19ee721a;hp=7a86311396ec9b472e9a049fcc3b185b9b6e5837;hpb=3c19a9d1054e341f806d39714ddf1d70b03ef142;p=elisp%2Fgnus.git-

diff --git a/lisp/rfc2047.el b/lisp/rfc2047.el
index 7a86311..10eff84 100644
--- a/lisp/rfc2047.el
+++ b/lisp/rfc2047.el
@@ -1,5 +1,5 @@
 ;;; rfc2047.el --- Functions for encoding and decoding rfc2047 messages
-;; Copyright (C) 1998, 1999, 2000 Free Software Foundation, Inc.
+;; Copyright (C) 1998, 1999, 2000, 2001 Free Software Foundation, Inc.
 
 ;; Author: Lars Magne Ingebrigtsen <larsi@gnus.org>
 ;;	MORIOKA Tomohiko <morioka@jaist.ac.jp>
@@ -22,25 +22,31 @@
 
 ;;; 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))
 
 (require 'qp)
 (require 'mm-util)
 (require 'ietf-drums)
 (require 'mail-prsvr)
+(require 'base64)
+;; Fixme: Avoid this (for gnus-point-at-...) mm dependence on gnus.
+(require 'gnus-util)
+(autoload 'mm-body-7-or-8 "mm-bodies")
 
 (defvar rfc2047-header-encoding-alist
   '(("Newsgroups" . nil)
     ("Message-ID" . nil)
+    ("\\(Resent-\\)?\\(From\\|Cc\\|To\\|Bcc\\|Reply-To\\|Sender\\)" .
+     "-A-Za-z0-9!*+/=_")
     (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:
 
@@ -48,7 +54,8 @@ The values can be:
 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 charset;
 4) `default', in which case the field will be encoded as the rest
-   of the article.")
+   of the article.
+5) a string, like `mime', expect for using it as word-chars.")
 
 (defvar rfc2047-charset-encoding-alist
   '((us-ascii . nil)
@@ -61,9 +68,13 @@ The values can be:
     (iso-8859-7 . Q)
     (iso-8859-8 . Q)
     (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)
@@ -79,8 +90,12 @@ Valid encodings are nil, `Q' and `B'.")
   "Alist of RFC2047 encodings to encoding functions.")
 
 (defvar rfc2047-q-encoding-alist
-  '(("\\(From\\|Cc\\|To\\|Bcc\||Reply-To\\):" . "-A-Za-z0-9!*+/=_")
-    ("." . "^\000-\007\011\013\015-\037\200-\377=_?"))
+  '(("\\(Resent-\\)?\\(From\\|Cc\\|To\\|Bcc\\|Reply-To\\|Sender\\):"
+     . "-A-Za-z0-9!*+/" )
+    ;; = (\075), _ (\137), ? (\077) are used in the encoded word.
+    ;; Avoid using 8bit characters.
+    ;; Equivalent to "^\000-\007\011\013\015-\037\200-\377=_?"
+    ("." . "\010\012\014\040-\074\076\100-\136\140-\177"))
   "Alist of header regexps and valid Q characters.")
 
 ;;;
@@ -101,6 +116,14 @@ Valid encodings are nil, `Q' and `B'.")
        (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)))))
+
 (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."
@@ -112,15 +135,26 @@ Should be called narrowed to the head of the message."
 	(save-restriction
 	  (rfc2047-narrow-to-field)
 	  (if (not (rfc2047-encodable-p))
-	      (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.
-		       ;; Is message-posting-charset a coding system?
-		       (mm-encode-coding-region 
-			(point-min) (point-max) 
-			(car message-posting-charset)))
+	      (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.
+		    ;; Is message-posting-charset a coding system?
+		    (mm-encode-coding-region
+		     (point-min) (point-max)
+		     (car message-posting-charset))
+		  nil)
+		;; 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)
@@ -131,23 +165,57 @@ Should be called narrowed to the head of the message."
 		(setq alist nil
 		      method (cdr elem))))
 	    (cond
+	     ((stringp method)
+	      (rfc2047-encode-region (point-min) (point-max) method))
 	     ((eq method 'mime)
-	      (rfc2047-encode-region (point-min) (point-max))
-	      (rfc2047-fold-region (point-min) (point-max)))
+	      (rfc2047-encode-region (point-min) (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-min) (point-max) 
+		  (mm-encode-coding-region (point-min) (point-max)
 					   mail-parse-charset)))
+	     ;; We get this when CC'ing messsages to newsgroups with
+	     ;; 8-bit names.  The group name mail copy just get
+	     ;; 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.
+	     ((null method)
+	      (when (delq 'ascii 
+			  (mm-find-charset-region (point-min) (point-max)))
+		(rfc2047-encode-region (point-min) (point-max))))
+;;;	     ((null method)
+;;;	      (and (delq 'ascii
+;;;			 (mm-find-charset-region (point-min)
+;;;						 (point-max)))
+;;;		   (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 (featurep 'mule)
+	      (if (and (featurep 'mule)
+		       (if (boundp 'default-enable-multibyte-characters)
+			   default-enable-multibyte-characters))
 		  (mm-encode-coding-region (point-min) (point-max) method)))
 	     ;; Hm.
 	     (t)))
 	  (goto-char (point-max)))))))
 
-(defun rfc2047-encodable-p (&optional header)
-  "Say whether the current (narrowed) buffer contains characters that need encoding in headers."
+;; 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 ()
+  "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
@@ -159,95 +227,94 @@ Should be called narrowed to the head of the message."
 	(setq found t)))
     found))
 
-(defun rfc2047-dissect-region (b e)
+(defun rfc2047-dissect-region (b e &optional word-chars)
   "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)
+  (unless word-chars
+    ;; Anything except most CTLs, WSP
+    (setq word-chars "\010\012\014\041-\177"))
+  (let (mail-parse-mule-charset
+	words point current
+	result word)
     (save-restriction
       (narrow-to-region b e)
       (goto-char (point-min))
-      (skip-chars-forward all-specials)
-      (setq b (point))
+      (skip-chars-forward "\000-\177")
       (while (not (eobp))
-	(cond
-	 ((not state)
-	  (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)
-	    (unless b
-		(setq b (point)))
-	    (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)
-	    (if (not current)
-		(setq b nil)
-	      (push (list b (point) current) words)
-	      (setq b (point))
-	      (setq current nil)))
-	   ((or (eq (setq cs (mm-charset-after)) 'ascii)
-		(if current
-		    (eq current cs)
-		  (setq current cs))))
-	   (t
-	    (push (list b (point) current) words)
-	    (setq current cs)
-	    (setq b (point))))))
-	(if state
-	    (forward-char)
-	  (skip-chars-forward all-specials)))
-      (if (eq state 'word)
-	  (push (list b (point) current) words)))
-    words))
-
-(defun rfc2047-encode-region (b e)
-  "Encode all encodable words in REGION."
-  (let ((words (rfc2047-dissect-region b e))
-	beg end current word)
-    (while (setq word (pop words))
-      (if (equal (nth 2 word) current)
-	  (setq beg (nth 0 word))
-	(when current
-	  (if (and (eq beg (nth 1 word)) (nth 2 word))
+	(setq point (point))
+	(skip-chars-backward word-chars b)
+	(unless (eq b (point))
+	  (push (cons (buffer-substring b (point)) nil) words))
+	(setq b (point))
+	(goto-char point)
+	(setq current (mm-charset-after))
+	(forward-char 1)
+	(skip-chars-forward word-chars)
+	(while (and (not (eobp))
+		    (eq (mm-charset-after) current))
+	  (forward-char 1)
+	  (skip-chars-forward word-chars))
+	(unless (eq b (point))
+	  (push (cons (buffer-substring b (point)) current) words))
+	(setq b (point))
+	(skip-chars-forward "\000-\177"))
+      (unless (eq b (point))
+	(push (cons (buffer-substring b (point)) nil) words)))
+    ;; merge adjacent words
+    (setq word (pop words))
+    (while word
+      (if (and (cdr word)
+	       (caar words)
+	       (not (cdar words))
+	       (not (string-match "[^ \t]" (caar words))))
+	  (if (eq (cdr (nth 1 words)) (cdr word))
 	      (progn
-		;; There might be a bug in Emacs Mule.
-		;; A space must be inserted before encoding.
-		(goto-char beg)
-		(insert " ")
-		(rfc2047-encode (1+ beg) (1+ end) current))
-	    (rfc2047-encode beg end current)))
-	(setq current (nth 2 word)
-	      beg (nth 0 word)
-	      end (nth 1 word))))
-    (when current
-      (rfc2047-encode beg end current))))
-
-(defun rfc2047-encode-string (string)
+		(setq word (cons (concat
+				  (car (nth 1 words)) (caar words)
+				  (car word))
+				 (cdr word)))
+		(pop words)
+		(pop words))
+	    (push (cons (concat (caar words) (car word)) (cdr word))
+		  result)
+	    (pop words)
+	    (setq word (pop words)))
+	(push word result)
+	(setq word (pop words))))
+    result))
+
+(defun rfc2047-encode-region (b e &optional word-chars)
+  "Encode all encodable words in region B to E."
+  (let ((words (rfc2047-dissect-region b e word-chars)) word)
+    (save-restriction
+      (narrow-to-region b e)
+      (delete-region (point-min) (point-max))
+      (while (setq word (pop words))
+	(if (not (cdr word))
+	    (insert (car word))
+	  (rfc2047-fold-region (gnus-point-at-bol) (point))
+	  (goto-char (point-max))
+	  (if (> (- (point) (save-restriction
+			      (widen)
+			      (gnus-point-at-bol))) 76)
+	      (insert "\n "))
+	  ;; Insert blank between encoded words
+	  (if (eq (char-before) ?=) (insert " "))
+	  (rfc2047-encode (point)
+			  (progn (insert (car word)) (point))
+			  (cdr word))))
+      (rfc2047-fold-region (point-min) (point-max)))))
+
+(defun rfc2047-encode-string (string &optional word-chars)
   "Encode words in STRING."
   (with-temp-buffer
     (insert string)
-    (rfc2047-encode-region (point-min) (point-max))
+    (rfc2047-encode-region (point-min) (point-max) word-chars)
     (buffer-string)))
 
 (defun rfc2047-encode (b e charset)
-  "Encode the word in the region with CHARSET."
+  "Encode the word in the region B to E with CHARSET."
   (let* ((mime-charset (mm-mime-charset charset))
+	 (cs (mm-charset-to-coding-system mime-charset))
 	 (encoding (or (cdr (assq mime-charset
 				  rfc2047-charset-encoding-alist))
 		       'B))
@@ -265,8 +332,8 @@ Should be called narrowed to the head of the message."
 	  (unless (eobp)
 	    (insert "\n"))))
       (if (and (mm-multibyte-p)
-	       (mm-coding-system-p mime-charset))
-	  (mm-encode-coding-region (point-min) (point-max) mime-charset))
+	       (mm-coding-system-p cs))
+	  (mm-encode-coding-region (point-min) (point-max) cs))
       (funcall (cdr (assq encoding rfc2047-encoding-function-alist))
 	       (point-min) (point-max))
       (goto-char (point-min))
@@ -279,30 +346,111 @@ Should be called narrowed to the head of the message."
 	(insert "?=")
 	(forward-line 1)))))
 
+(defun rfc2047-fold-field ()
+  "Fold the current line."
+  (save-excursion
+    (save-restriction
+      (rfc2047-narrow-to-field)
+      (rfc2047-fold-region (point-min) (point-max)))))
+
 (defun rfc2047-fold-region (b e)
-  "Fold the long lines in the region."
+  "Fold long lines in region B to E."
   (save-restriction
     (narrow-to-region b e)
     (goto-char (point-min))
-    (let ((break nil))
+    (let ((break nil)
+	  (qword-break nil)
+	  (first t)
+	  (bol (save-restriction
+		 (widen)
+		 (gnus-point-at-bol))))
       (while (not (eobp))
+	(when (and (or break qword-break)
+		   (> (- (point) bol) 76))
+	  (goto-char (or break qword-break))
+	  (setq break nil
+		qword-break nil)
+	  (if (looking-at "[ \t]")
+	      (insert "\n")
+	    (insert "\n "))
+	  (setq bol (1- (point)))
+	  ;; Don't break before the first non-LWSP characters.
+	  (skip-chars-forward " \t")
+	  (unless (eobp)
+	    (forward-char 1)))
 	(cond
+	 ((eq (char-after) ?\n)
+	  (forward-char 1)
+	  (setq bol (point)
+		break nil
+		qword-break nil)
+	  (skip-chars-forward " \t")
+	  (unless (or (eobp) (eq (char-after) ?\n))
+	    (forward-char 1)))
+	 ((eq (char-after) ?\r)
+	  (forward-char 1))
 	 ((memq (char-after) '(?  ?\t))
-	  (setq break (point)))
-	 ((and (not break)
-	       (looking-at "=\\?"))
-	  (setq break (point)))
-	 ((and break
-	       (looking-at "\\?=")
-	       (> (- (point) (save-excursion (beginning-of-line) (point))) 76))
-	  (goto-char break)
-	  (setq break nil)
-	  (insert "\n ")))
+	  (skip-chars-forward " \t")
+	  (if first
+	      ;; Don't break just after the header name.
+	      (setq first nil)
+	    (setq break (1- (point)))))
+	 ((not break)
+	  (if (not (looking-at "=\\?[^=]"))
+	      (if (eq (char-after) ?=)
+		  (forward-char 1)
+		(skip-chars-forward "^ \t\n\r="))
+	    (setq qword-break (point))
+	    (skip-chars-forward "^ \t\n\r")))
+	 (t
+	  (skip-chars-forward "^ \t\n\r"))))
+      (when (and (or break qword-break)
+		 (> (- (point) bol) 76))
+	(goto-char (or break qword-break))
+	(setq break nil
+	      qword-break nil)
+	  (if (looking-at "[ \t]")
+	      (insert "\n")
+	    (insert "\n "))
+	(setq bol (1- (point)))
+	;; Don't break before the first non-LWSP characters.
+	(skip-chars-forward " \t")
 	(unless (eobp)
 	  (forward-char 1))))))
 
+(defun rfc2047-unfold-field ()
+  "Fold the current line."
+  (save-excursion
+    (save-restriction
+      (rfc2047-narrow-to-field)
+      (rfc2047-unfold-region (point-min) (point-max)))))
+
+(defun rfc2047-unfold-region (b e)
+  "Unfold lines in region B to E."
+  (save-restriction
+    (narrow-to-region b e)
+    (goto-char (point-min))
+    (let ((bol (save-restriction
+		 (widen)
+		 (gnus-point-at-bol)))
+	  (eol (gnus-point-at-eol))
+	  leading)
+      (forward-line 1)
+      (while (not (eobp))
+	(looking-at "[ \t]*")
+	(setq leading (- (match-end 0) (match-beginning 0)))
+	(if (< (- (gnus-point-at-eol) bol leading) 76)
+	    (progn
+	      (goto-char eol)
+	      (delete-region eol (progn
+				   (skip-chars-forward " \t\n\r")
+				   (1- (point)))))
+	  (setq bol (gnus-point-at-bol)))
+	(setq eol (gnus-point-at-eol))
+	(forward-line 1)))))
+
 (defun rfc2047-b-encode-region (b e)
-  "Encode the header contained in REGION with the B encoding."
+  "Base64-encode the header contained in region B to E."
   (save-restriction
     (narrow-to-region (goto-char b) e)
     (while (not (eobp))
@@ -312,30 +460,41 @@ Should be called narrowed to the head of the message."
       (forward-line))))
 
 (defun rfc2047-q-encode-region (b e)
-  "Encode the header contained in REGION with the Q encoding."
+  "Quoted-printable-encode the header in region B to E."
   (save-excursion
     (save-restriction
       (narrow-to-region (goto-char b) e)
-      (let ((alist rfc2047-q-encoding-alist))
+      (let ((alist rfc2047-q-encoding-alist)
+	    (bol (save-restriction
+		   (widen)
+		   (gnus-point-at-bol))))
 	(while alist
 	  (when (looking-at (caar alist))
-	    (quoted-printable-encode-region b e nil (cdar alist))
+	    (mm-with-unibyte-current-buffer-mule4
+	      (quoted-printable-encode-region
+	       (point-min) (point-max) nil (cdar alist)))
 	    (subst-char-in-region (point-min) (point-max) ?  ?_)
 	    (setq alist nil))
 	  (pop alist))
-	(goto-char (point-min))
-	(while (not (eobp))
-	  (goto-char (min (point-max) (+ 64 (point))))
-	  (search-backward "=" (- (point) 2) t)
-	  (unless (eobp)
-	    (insert "\n")))))))
+	;; The size of QP encapsulation is about 20, so set limit to
+	;; 56=76-20.
+	(unless (< (- (point-max) (point-min)) 56)
+	  ;; Don't break if it could fit in one line.
+	  ;; Let rfc2047-encode-region break it later.
+	  (goto-char (1+ (point-min)))
+	  (while (and (not (bobp)) (not (eobp)))
+	    (goto-char (min (point-max) (+ 56 bol)))
+	    (search-backward "=" (- (point) 2) t)
+	    (unless (or (bobp) (eobp))
+	      (insert "\n")
+	      (setq bol (point)))))))))
 
 ;;;
 ;;; Functions for decoding RFC2047 messages
 ;;;
 
 (defvar rfc2047-encoded-word-regexp
-  "=\\?\\([^][\000-\040()<>@,\;:\\\"/?.=]+\\)\\?\\(B\\|Q\\)\\?\\([!->@-~ +]+\\)\\?=")
+  "=\\?\\([^][\000-\040()<>@,\;:\\\"/?.=]+\\)\\?\\(B\\|Q\\)\\?\\([!->@-~ +]*\\)\\?=")
 
 (defun rfc2047-decode-region (start end)
   "Decode MIME-encoded words in region between START and END."
@@ -370,7 +529,8 @@ Should be called narrowed to the head of the message."
 		   mail-parse-charset
 		   (not (eq mail-parse-charset 'us-ascii))
 		   (not (eq mail-parse-charset 'gnus-decoded)))
-	  (mm-decode-coding-region b (point-max) mail-parse-charset))))))
+	  (mm-decode-coding-region b (point-max) mail-parse-charset))
+	(rfc2047-unfold-region (point-min) (point-max))))))
 
 (defun rfc2047-decode-string (string)
   "Decode the quoted-printable-encoded STRING and return the results."
@@ -397,19 +557,30 @@ Return WORD if not."
        (error word))
      word)))
 
+(defun rfc2047-pad-base64 (string)
+  "Pad STRING to quartets."
+  ;; Be more liberal to accept buggy base64 strings. If
+  ;; base64-decode-string accepts buggy strings, this function could
+  ;; be aliased to identity.
+  (case (mod (length string) 4)
+    (0 string)
+    (1 string) ;; Error, don't pad it.
+    (2 (concat string "=="))
+    (3 (concat string "="))))
+
 (defun rfc2047-decode (charset encoding string)
-  "Decode STRING that uses CHARSET with ENCODING.
+  "Decode STRING from the given MIME CHARSET in the given ENCODING.
 Valid ENCODINGs are \"B\" and \"Q\".
-If your Emacs implementation can't decode CHARSET, it returns nil."
+If your Emacs implementation can't decode CHARSET, return nil."
   (if (stringp charset)
       (setq charset (intern (downcase charset))))
-  (if (or (not 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 
+    (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)))
@@ -417,15 +588,18 @@ If your Emacs implementation can't decode CHARSET, it returns nil."
       (when (and (eq cs 'ascii)
 		 mail-parse-charset)
 	(setq cs mail-parse-charset))
-      (mm-decode-coding-string
-       (cond
-	((equal "B" encoding)
-	 (base64-decode-string string))
-	((equal "Q" encoding)
-	 (quoted-printable-decode-string
-	  (mm-replace-chars-in-string string ?_ ? )))
-	(t (error "Invalid encoding: %s" encoding)))
-       cs))))
+      (mm-with-unibyte-current-buffer-mule4
+	;; In Emacs Mule 4, decoding UTF-8 should be in unibyte mode.
+	(mm-decode-coding-string
+	 (cond
+	  ((equal "B" encoding)
+	   (base64-decode-string
+	    (rfc2047-pad-base64 string)))
+	  ((equal "Q" encoding)
+	   (quoted-printable-decode-string
+	    (mm-replace-chars-in-string string ?_ ? )))
+	  (t (error "Invalid encoding: %s" encoding)))
+	 cs)))))
 
 (provide 'rfc2047)