Importing Gnus v5.8.4.

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

diff --git a/lisp/mm-bodies.el b/lisp/mm-bodies.el
index 0e6640b..2ee6bf2 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 Free Software Foundation, Inc.
+;; Copyright (C) 1998,99 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>

@@ -25,10 +25,22 @@
 ;;; Code:
 
 (eval-and-compile
 ;;; Code:
 
 (eval-and-compile

-  (if (not (fboundp 'base64-encode-string))
-      (require 'base64)))
+  (or (fboundp  'base64-decode-region)
+      (require 'base64))
+  (autoload 'binhex-decode-region "binhex"))
+

 (require 'mm-util)
 (require 'mm-util)

+(require 'rfc2047)

 (require 'qp)
 (require 'qp)

+(require 'uudecode)
+
+;; 8bit treatment gets any char except: 0x32 - 0x7f, CR, LF, TAB, BEL,
+;; BS, vertical TAB, form feed, and ^_
+(defvar mm-7bit-chars "\x20-\x7f\r\n\t\x7\x8\xb\xc\x1f")
+
+(defvar mm-body-charset-encoding-alist nil
+  "Alist of MIME charsets to encodings.
+Valid encodings are `7bit', `8bit', `quoted-printable' and `base64'.")

 
 (defun mm-encode-body ()
   "Encode a body.
 
 (defun mm-encode-body ()
   "Encode a body.

@@ -37,90 +49,195 @@ 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."

-  (save-excursion
-    (goto-char (point-min))
-    (let ((charsets
-          (delq 'ascii (find-charset-region (point-min) (point-max))))
-         charset)
-      (cond
-       ;; No encoding.
-       ((null charsets)
-       nil)
-       ;; Too many charsets.
-       ((> (length charsets) 1)
-       charsets)
-       ;; We encode.
-       (t
-       (let ((mime-charset 
-              (mm-mime-charset (car charsets) (point-min) (point-max)))
-             start)
-         (when (or t
-                   ;; We always decode.
-                   (not (mm-coding-system-equal
-                         mime-charset buffer-file-coding-system)))
-           (while (not (eobp))
-             (if (eq (char-charset (following-char)) 'ascii)
-                 (when start
-                   (mm-encode-coding-region start (point) mime-charset)
-                   (setq start nil))
-               (unless start
-                 (setq start (point))))
-             (forward-char 1))
-           (when start
-             (mm-encode-coding-region start (point) mime-charset)
-             (setq start nil)))
-         mime-charset))))))
-
-(defun mm-body-encoding ()
-  "Return the encoding of the current buffer."
-  (if (and  
-       (null (delq 'ascii (find-charset-region (point-min) (point-max))))
-       ;;;!!!The following is necessary because the function
-       ;;;!!!above seems to return the wrong result under Emacs 20.3.
-       ;;;!!!Sometimes.
-       (save-excursion
-        (goto-char (point-min))
-        (skip-chars-forward "\0-\177")
-        (eobp)))
-      '7bit
-    '8bit))
+  (if (not (featurep 'mule))
+      ;; 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))
+    (save-excursion
+      (goto-char (point-min))
+      (let ((charsets (mm-find-mime-charset-region (point-min) (point-max)))
+           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 (mm-charset-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 &optional encoding)
+  "Do Content-Transfer-Encoding and return the encoding of the current buffer."
+  (let ((bits (mm-body-7-or-8)))
+    (cond
+     ((eq bits '7bit)
+      bits)
+     ((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
+      (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)))))
+
+(defun mm-body-7-or-8 ()
+  "Say whether the body is 7bit or 8bit."
+  (cond
+   ((not (featurep 'mule))
+    (if (save-excursion
+         (goto-char (point-min))
+         (skip-chars-forward mm-7bit-chars)
+         (eobp))
+       '7bit
+      '8bit))
+   (t
+    ;; Mule version
+    (if (and (null (delq 'ascii
+                        (mm-find-charset-region (point-min) (point-max))))
+            ;;!!!The following is necessary because the function
+            ;;!!!above seems to return the wrong result under
+            ;;!!!Emacs 20.3.  Sometimes.
+            (save-excursion
+              (goto-char (point-min))
+              (skip-chars-forward mm-7bit-chars)
+              (eobp)))
+       '7bit
+      '8bit))))

 
 ;;;
 ;;; Functions for decoding
 ;;;
 
 
 ;;;
 ;;; Functions for decoding
 ;;;
 

-(defun mm-decode-content-transfer-encoding (encoding)
-  (cond
-   ((eq encoding 'quoted-printable)
-    (quoted-printable-decode-region (point-min) (point-max)))
-   ((eq encoding 'base64)
-    (condition-case ()
-       (base64-decode-region (point-min) (point-max))
-      (error nil)))
-   ((memq encoding '(7bit 8bit binary))
-    )
-   ((null encoding)
-    )
-   (t
-    (error "Can't decode encoding %s" encoding))))
+(defun mm-decode-content-transfer-encoding (encoding &optional type)
+  (prog1
+      (condition-case error
+         (cond
+          ((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))))
+          ((memq encoding '(7bit 8bit binary))
+           ;; Do nothing.
+           )
+          ((null encoding)
+           ;; Do nothing.
+           )
+          ((memq encoding '(x-uuencode x-uue))
+           (funcall mm-uu-decode-function (point-min) (point-max)))
+          ((eq encoding 'x-binhex)
+           (funcall mm-uu-binhex-decode-function (point-min) (point-max)))
+          ((functionp encoding)
+           (funcall encoding (point-min) (point-max)))
+          (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))
+          (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 encoding)
+(defun mm-decode-body (charset &optional encoding type)

   "Decode the current article that has been encoded with ENCODING.
 The characters in CHARSET should then be decoded."
   "Decode the current article that has been encoded with ENCODING.
 The characters in CHARSET should then be decoded."

-  (setq charset (or charset rfc2047-default-charset))
+  (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))

   (save-excursion
     (when encoding
   (save-excursion
     (when encoding

-      (mm-decode-content-transfer-encoding encoding))
+      (mm-decode-content-transfer-encoding encoding type))

     (when (featurep 'mule)
     (when (featurep 'mule)

-      (let (mule-charset)
-       (when (and charset
-                  (setq mule-charset (mm-charset-to-coding-system charset))
-                  buffer-file-coding-system
-                  ;;(not (mm-coding-system-equal
-                  ;;    buffer-file-coding-system mule-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
+                  (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))))))
 
          (mm-decode-coding-region (point-min) (point-max) mule-charset))))))
 

+(defun mm-decode-string (string charset)
+  "Decode STRING with CHARSET."
+  (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))
+  (or
+   (when (featurep 'mule)
+     (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-decode-coding-string string mule-charset))))
+   string))
+

 (provide 'mm-bodies)
 
 ;; mm-bodies.el ends here
 (provide 'mm-bodies)
 
 ;; mm-bodies.el ends here