Synch with Oort Gnus.

[elisp/gnus.git-] / lisp / mm-bodies.el

diff --git a/lisp/mm-bodies.el b/lisp/mm-bodies.el
index ad7bbeb..dacb792 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>

@@ -26,86 +26,128 @@
 
 (eval-and-compile
   (or (fboundp  'base64-decode-region)
 
 (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 '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 ^_
 
 ;; 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 ()
+(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.
   "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 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.
-      (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
     (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)
            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)))
+       (goto-char (point-min))
+       (let ((charsets (mm-find-mime-charset-region (point-min) (point-max)))
+             start)
+         (cond
+          ;; No encoding.
+          ((null charsets)
+           nil)
+          ;; Too many charsets.
+          ((> (length charsets) 1)
+           charsets)
+          ;; We encode.
+          (t
+           (setq charset (car charsets))
+           (while (not (eobp))
+             (if (eq (mm-charset-after) 'ascii)
+                 (when start
+                   (save-restriction
+                     (narrow-to-region start (point))
+                     (mm-encode-coding-region
+                      start (point) (mm-charset-to-coding-system charset))
+                     (goto-char (point-max)))
+                   (setq start nil))
+               (unless start
+                 (setq start (point))))
+             (forward-char 1))
+           (when start
+             (mm-encode-coding-region start (point)
+                                      (mm-charset-to-coding-system charset))
+             (setq start nil))

            charset)))))))
 
            charset)))))))
 

-(defun mm-body-encoding (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."
   "Do Content-Transfer-Encoding and return the encoding of the current buffer."

-  (let ((bits (mm-body-7-or-8)))
+  (let ((bits (mm-body-7-or-8))
+       (longp (mm-long-lines-p 1000)))
+    (require 'message)

     (cond
     (cond

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

       bits)
       bits)

-     ((eq charset mail-parse-charset)
+     ((and (not mm-use-ultra-safe-encoding)
+          (not longp)
+          (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 +157,10 @@ 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 +170,7 @@ 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))))

@@ -137,31 +180,47 @@ If no encoding was done, nil is returned."
 ;;;
 
 (defun mm-decode-content-transfer-encoding (encoding &optional type)
 ;;;
 
 (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)
   (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)
           ((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))
           ((memq encoding '(7bit 8bit binary))

-           )
+           ;; Do nothing.
+           t)

           ((null encoding)
           ((null encoding)

-           )
+           ;; Do nothing.
+           t)

           ((memq encoding '(x-uuencode x-uue))
           ((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)
           ((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)

           ((functionp encoding)
           ((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
           (t
            (message "Unknown encoding %s; defaulting to 8bit" encoding)))
        (error

@@ -178,42 +237,56 @@ 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 ((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)
                   ;; buffer-file-coding-system
                   ;;Article buffer is nil coding system
                   ;;in XEmacs
                   (mm-multibyte-p)

-                  (or (not (eq mule-charset 'ascii))
-                      (setq mule-charset mail-parse-charset))
-                  (not (eq mule-charset 'gnus-decoded)))
-         (mm-decode-coding-region (point-min) (point-max) mule-charset))))))
+                  (or (not (eq coding-system 'ascii))
+                      (setq coding-system mail-parse-charset))
+                  (not (eq coding-system 'gnus-decoded)))
+         (mm-decode-coding-region (point-min) (point-max) coding-system))))))

 
 (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 ((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)
                  (mm-multibyte-p)

-                 (or (not (eq mule-charset 'ascii))
-                     (setq mule-charset mail-parse-charset)))
-        (mm-decode-coding-string string mule-charset))))
+                 (or (not (eq coding-system 'ascii))
+                     (setq coding-system mail-parse-charset)))
+        (mm-decode-coding-string string coding-system))))

    string))
 
 (provide 'mm-bodies)
 
    string))
 
 (provide 'mm-bodies)
 

-;; mm-bodies.el ends here
+;;; mm-bodies.el ends here