From: akr <akr>
Date: Fri, 20 Mar 1998 01:04:27 +0000 (+0000)
Subject: (eword-decode-string): Treat undecodable encoded-words.
X-Git-Tag: akr-199811302358~9
X-Git-Url: http://git.chise.org/gitweb/?a=commitdiff_plain;h=593139a8c169abdb0a0b19568bb3b8d59c1e0bc5;p=elisp%2Fsemi.git

(eword-decode-string): Treat undecodable encoded-words.
---

diff --git a/ChangeLog b/ChangeLog
index eac82f9..2d2d420 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,3 +1,8 @@
+1998-03-20  Tanaka Akira  <akr@jaist.ac.jp>
+
+	* eword-decode.el (eword-decode-string): Treat undecodable
+	encoded-words.
+
 1998-03-19  Tanaka Akira  <akr@jaist.ac.jp>
 
 	* eword-decode.el (eword-decode-region): Treat undecodable
diff --git a/eword-decode.el b/eword-decode.el
index 30cabd5..a7dd200 100644
--- a/eword-decode.el
+++ b/eword-decode.el
@@ -98,6 +98,30 @@
 ;;   (concat "\\(Q\\)\\?" eword-Q-encoded-text-regexp))
 
 
+;; @ encoded-word's neighbor
+;;
+
+(defvar eword-decode-before-ewords-regexp "^\\|[ \t]"
+  "Regexp that matches before encoded words.
+This value must not contain grouping construct.
+Default value is \"\\\\`\\\\|^\\\\|[ \\t]\".
+Another useful (but not RFC2047 compliant) value is \"\".")
+
+(defvar eword-decode-between-ewords-regexp "\\(\n?[ \t]\\)+"
+  "Regexp that matches between encoded words.
+This value must contain exactly one grouping construct.
+Default value is \"\\\\(\\n?[ \\t]\\\\)+\".
+Another useful (but not RFC2047 compliant) value is \"\\\\(\\n?[ \\t]\\\\)*\".")
+
+(defvar eword-decode-after-ewords-regexp "[ \t]\\|$"
+  "Regexp that matches after encoded words.
+Default value is \"[ \\t]\\\\|$\".
+Another useful (but not RFC2047 compliant) value is \"\".")
+
+; (setq eword-decode-before-ewords-regexp "")
+; (setq eword-decode-between-ewords-regexp "\\(\n?[ \t]\\)*")
+; (setq eword-decode-after-ewords-regexp "")
+
 ;;; @ for string
 ;;;
 
@@ -112,53 +136,43 @@ decode the charset included in it, it is not decoded.
 If MUST-UNFOLD is non-nil, it unfolds and eliminates line-breaks even
 if there are in decoded encoded-words (generated by bad manner MUA
 such as a version of Net$cape)."
-  (setq string (std11-unfold-string string))
-  (let ((dest "")(ew nil)
-	beg end)
-    (while (and (string-match eword-encoded-word-regexp string)
-		(setq beg (match-beginning 0)
-		      end (match-end 0))
-		)
-      (if (> beg 0)
-	  (if (not
-	       (and (eq ew t)
-		    (string-match "^[ \t]+$" (substring string 0 beg))
-		    ))
-	      (setq dest (concat dest (substring string 0 beg)))
-	    )
-	)
-      (setq dest
-	    (concat dest
-		    (eword-decode-encoded-word
-		     (substring string beg end) must-unfold)
-		    ))
-      (setq string (substring string end))
-      (setq ew t)
-      )
-    (concat dest string)
-    ))
+  (let ((src (std11-unfold-string string))
+  	(dst "")
+	b e ew dw)
+    (while
+	(string-match
+	 (concat "\\(" eword-decode-before-ewords-regexp "\\)"
+		 "\\(" eword-encoded-word-regexp "\\)"
+		 "\\(" eword-decode-after-ewords-regexp "\\)")
+	 src)
+      (setq b (match-beginning 2))
+      (setq e (match-end 2))
+      (setq ew (substring src b e))
+      (setq dw (eword-decode-encoded-word ew must-unfold))
+      (setq dst (concat dst (substring src 0 b) dw))
+      (setq src (substring src e))
+      (if (not (string= ew dw))
+	  (while
+	      (and
+	       (string-match
+		(concat "\\`"
+			"\\(" eword-decode-between-ewords-regexp "\\)"
+			"\\(" eword-encoded-word-regexp "\\)"
+			"\\(" eword-decode-after-ewords-regexp "\\)")
+		src)
+	       (progn
+		 (setq e (match-end 3))
+		 (setq ew (substring src (match-beginning 3) e))
+		 (setq dw (eword-decode-encoded-word ew must-unfold))
+		 (not (string= ew dw))))
+	    (setq dst (concat dst dw))
+	    (setq src (substring src e)))))
+    (concat dst src)))
 
 
 ;;; @ for region
 ;;;
 
-(defvar eword-decode-before-ewords-regexp "\\=\\|^\\|[ \t]"
-  "Regexp that matches before encoded words.
-This value must not contain grouping construct.
-Default value is \"\\\\\\==\\\\|^\\\\|[ \\t]\".
-Another useful (but not RFC2047 compliant) value is \"\".")
-
-(defvar eword-decode-between-ewords-regexp "\\(\n?[ \t]\\)+"
-  "Regexp that matches between encoded words.
-This value must contain exactly one grouping construct.
-Default value is \"\\\\(\\n?[ \\t]\\\\)+\".
-Another useful (but not RFC2047 compliant) value is \"\\\\(\\n?[ \\t]\\\\)*\".")
-
-(defvar eword-decode-after-ewords-regexp "[ \t]\\|$"
-  "Regexp that matches after encoded words.
-Default value is \"[ \\t]\\\\|$\".
-Another useful (but not RFC2047 compliant) value is \"\".")
-
 (defun eword-decode-region (start end &optional unfolding must-unfold)
   "Decode MIME encoded-words in region between START and END.
 
@@ -177,13 +191,15 @@ such as a version of Net$cape)."
       (goto-char (point-min))
       (let (b e ew dw)
 	(while 
-	    (re-search-forward
-	      (concat "\\(" eword-decode-before-ewords-regexp "\\)"
-		      "\\(" eword-encoded-word-regexp "\\)"
-		      "\\(" eword-decode-after-ewords-regexp "\\)") nil t)
+	    (progn
+	      (narrow-to-region (point) (point-max))
+	      (re-search-forward
+		(concat "\\(" eword-decode-before-ewords-regexp "\\)"
+			"\\(" eword-encoded-word-regexp "\\)"
+			"\\(" eword-decode-after-ewords-regexp "\\)") nil t))
 	  (setq b (match-beginning 2))
 	  (setq e (match-end 2))
-	  (setq ew (buffer-substring (match-beginning 2) (match-end 2)))
+	  (setq ew (buffer-substring b e))
 	  (setq dw (eword-decode-encoded-word ew must-unfold))
 	  (if (not (string= ew dw))
 	      (progn
@@ -193,13 +209,13 @@ such as a version of Net$cape)."
 		(while
 		    (and
 		     (looking-at
-		       (concat eword-decode-between-ewords-regexp
+		       (concat "\\(" eword-decode-between-ewords-regexp "\\)"
 			       "\\(" eword-encoded-word-regexp "\\)"
 			       "\\(" eword-decode-after-ewords-regexp "\\)"))
 		     (progn
 		       (setq b (match-beginning 0))
-		       (setq e (match-end 2))
-		       (setq ew (buffer-substring (match-beginning 2) e))
+		       (setq e (match-end 3))
+		       (setq ew (buffer-substring (match-beginning 3) e))
 		       (setq dw (eword-decode-encoded-word ew must-unfold))
 		       (not (string= ew dw))))
 		  (goto-char e)