Fix my last change.

[gnus] / lisp / mm-bodies.el

diff --git a/lisp/mm-bodies.el b/lisp/mm-bodies.el
index ad7bbeb..5b8874c 100644 (file)
--- a/lisp/mm-bodies.el
+++ b/lisp/mm-bodies.el
@@ -1,5 +1,5 @@
 ;;; mm-bodies.el --- Functions for decoding MIME things
 ;;; mm-bodies.el --- Functions for decoding MIME things

-;; Copyright (C) 1998,99 Free Software Foundation, Inc.
+;; Copyright (C) 1998, 1999, 2000 Free Software Foundation, Inc.

 
 ;; Author: Lars Magne Ingebrigtsen <larsi@gnus.org>
 ;;     MORIOKA Tomohiko <morioka@jaist.ac.jp>
 
 ;; Author: Lars Magne Ingebrigtsen <larsi@gnus.org>
 ;;     MORIOKA Tomohiko <morioka@jaist.ac.jp>


@@ -36,11 +36,20 @@
 
 ;; 8bit treatment gets any char except: 0x32 - 0x7f, CR, LF, TAB, BEL,
 ;; BS, vertical TAB, form feed, and ^_
 
 ;; 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-7bit-chars "\x20-\x7f\r\n\t\x7\x8\xb\xc\x1f")

 
 

-(defvar mm-body-charset-encoding-alist nil
+(defcustom mm-body-charset-encoding-alist
+  '((iso-2022-jp . 7bit)
+    (iso-2022-jp-2 . 7bit))

   "Alist of MIME charsets to encodings.
   "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 ()
   "Encode a body.
 
 (defun mm-encode-body ()
   "Encode a body.


@@ -49,9 +58,9 @@ If there is more than one non-ASCII MULE charset, then list of found
 MULE charsets are returned.
 If successful, the MIME charset is returned.
 If no encoding was done, nil is returned."
 MULE charsets are returned.
 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
       ;; In the non-Mule case, we search for non-ASCII chars and

-      ;; return the value of `mm-default-charset' if any are found.
+      ;; return the value of `mail-parse-charset' if any are found.

       (save-excursion
        (goto-char (point-min))
        (if (re-search-forward "[^\x0-\x7f]" nil t)
       (save-excursion
        (goto-char (point-min))
        (if (re-search-forward "[^\x0-\x7f]" nil t)


@@ -80,7 +89,7 @@ If no encoding was done, nil is returned."
                      (not (mm-coding-system-equal
                            charset buffer-file-coding-system)))
              (while (not (eobp))
                      (not (mm-coding-system-equal
                            charset buffer-file-coding-system)))
              (while (not (eobp))

-               (if (eq (char-charset (char-after)) 'ascii)
+               (if (eq (mm-charset-after) 'ascii)

                    (when start
                      (save-restriction
                        (narrow-to-region start (point))
                    (when start
                      (save-restriction
                        (narrow-to-region start (point))


@@ -95,17 +104,23 @@ If no encoding was done, nil is returned."
                (setq start nil)))
            charset)))))))
 
                (setq start nil)))
            charset)))))))
 

-(defun mm-body-encoding (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)))
     (cond
   "Do Content-Transfer-Encoding and return the encoding of the current buffer."
   (let ((bits (mm-body-7-or-8)))
     (cond

-     ((eq bits '7bit)
+     ((and (not mm-use-ultra-safe-encoding) (eq bits '7bit))

       bits)
       bits)

-     ((eq charset mail-parse-charset)
+     ((and (not mm-use-ultra-safe-encoding)
+          (or (eq t (cdr message-posting-charset))
+              (memq charset (cdr message-posting-charset))
+              (eq charset mail-parse-charset)))

       bits)
      (t
       bits)
      (t

-      (let ((encoding (or (cdr (assq charset mm-body-charset-encoding-alist))
+      (let ((encoding (or encoding
+                         (cdr (assq charset mm-body-charset-encoding-alist))

                          (mm-qp-or-base64))))
                          (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)))))
 
        (mm-encode-content-transfer-encoding encoding "text/plain")
        encoding)))))
 


@@ -115,9 +130,10 @@ If no encoding was done, nil is returned."
    ((not (featurep 'mule))
     (if (save-excursion
          (goto-char (point-min))
    ((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
    (t
     ;; Mule version
     (if (and (null (delq 'ascii


@@ -127,7 +143,7 @@ If no encoding was done, nil is returned."
             ;;!!!Emacs 20.3.  Sometimes.
             (save-excursion
               (goto-char (point-min))
             ;;!!!Emacs 20.3.  Sometimes.
             (save-excursion
               (goto-char (point-min))

-              (skip-chars-forward "\0-\177")
+              (skip-chars-forward mm-7bit-chars)

               (eobp)))
        '7bit
       '8bit))))
               (eobp)))
        '7bit
       '8bit))))


@@ -143,18 +159,27 @@ If no encoding was done, nil is returned."
           ((eq encoding 'quoted-printable)
            (quoted-printable-decode-region (point-min) (point-max)))
           ((eq encoding 'base64)
           ((eq encoding 'quoted-printable)
            (quoted-printable-decode-region (point-min) (point-max)))
           ((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)
+                (delete-region (point) (point-max)))
+              (point-max))))

           ((memq encoding '(7bit 8bit binary))
           ((memq encoding '(7bit 8bit binary))

+           ;; Do nothing.

            )
           ((null encoding)
            )
           ((null encoding)

+           ;; Do nothing.

            )
           ((memq encoding '(x-uuencode x-uue))
            (funcall mm-uu-decode-function (point-min) (point-max)))
            )
           ((memq encoding '(x-uuencode x-uue))
            (funcall mm-uu-decode-function (point-min) (point-max)))


@@ -178,16 +203,23 @@ If no encoding was done, nil is returned."
   "Decode the current article that has been encoded with ENCODING.
 The characters in CHARSET should then be decoded."
   (if (stringp charset)
   "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 (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))
   (save-excursion
     (when encoding
       (mm-decode-content-transfer-encoding encoding type))
     (when (featurep 'mule)
       (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))
+      (let ((mule-charset (mm-charset-to-coding-system charset)))
+       (if (and (not mule-charset)
+                (listp mail-parse-ignored-charsets)
+                (memq 'gnus-unknown mail-parse-ignored-charsets))
+           (setq mule-charset 
+                 (mm-charset-to-coding-system mail-parse-charset)))
+       (when (and charset mule-charset

                   ;; buffer-file-coding-system
                   ;;Article buffer is nil coding system
                   ;;in XEmacs
                   ;; buffer-file-coding-system
                   ;;Article buffer is nil coding system
                   ;;in XEmacs


@@ -199,15 +231,22 @@ The characters in CHARSET should then be decoded."
 
 (defun mm-decode-string (string charset)
   "Decode STRING with CHARSET."
 
 (defun mm-decode-string (string charset)
   "Decode STRING with CHARSET."

-  (if (stringp charset)
+  (when (stringp charset)

     (setq charset (intern (downcase 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)
   (or
    (when (featurep 'mule)

-     (let (mule-charset)
-       (when (and charset
-                 (setq mule-charset (mm-charset-to-coding-system charset))
+     (let ((mule-charset (mm-charset-to-coding-system charset)))
+       (if (and (not mule-charset)
+               (listp mail-parse-ignored-charsets)
+               (memq 'gnus-unknown mail-parse-ignored-charsets))
+          (setq mule-charset 
+                (mm-charset-to-coding-system mail-parse-charset)))
+       (when (and charset mule-charset

                  (mm-multibyte-p)
                  (or (not (eq mule-charset 'ascii))
                      (setq mule-charset mail-parse-charset)))
                  (mm-multibyte-p)
                  (or (not (eq mule-charset 'ascii))
                      (setq mule-charset mail-parse-charset)))