Synch with Oort Gnus.

[elisp/gnus.git-] / lisp / nnweb.el

diff --git a/lisp/nnweb.el b/lisp/nnweb.el
index 9a41a66..760ef75 100644 (file)
--- a/lisp/nnweb.el
+++ b/lisp/nnweb.el
@@ -63,9 +63,12 @@ and `altavista'.")
 (defvar nnweb-type-definition
   '(
     (dejanews ;; bought by google.com
-     (article . nnweb-google-wash-article)
-     (id . "http://groups.google.com/groups?as_umsgid=%s")
-     (reference . nnweb-google-reference)
+     ;;(article . nnweb-google-wash-article)
+     ;;(id . "http://groups.google.com/groups?as_umsgid=%s")
+     (article . ignore)
+     (id . "http://groups.google.com/groups?selm=%s&output=gplain")
+     ;;(reference . nnweb-google-reference)
+     (reference . identity)
      (map . nnweb-google-create-mapping)
      (search . nnweb-google-search)
      (address . "http://groups.google.com/groups")
@@ -189,16 +192,15 @@ and `altavista'.")
                       (when (string-match "^<\\(.*\\)>$" article)
                         (setq art (match-string 1 article)))
                       (when (and fetch art)
-                        (setq url (format fetch article))
+                        (setq url (format fetch art))
                         (mm-with-unibyte-current-buffer
                           (nnweb-fetch-url url))
                         (if (nnweb-definition 'reference t)
                             (setq article
-                                  (funcall (nnweb-definition 
+                                  (funcall (nnweb-definition
                                             'reference) article)))))))
        (unless nnheader-callback-function
-         (funcall (nnweb-definition 'article))
-         (nnweb-decode-entities))
+         (funcall (nnweb-definition 'article)))
        (nnheader-report 'nnweb "Fetched article %s" article)
        (cons group (and (numberp article) article))))))
 
@@ -574,6 +576,7 @@ and `altavista'.")
        (while (search-forward "," nil t)
          (replace-match " " t t)))
       (widen)
+      (nnweb-decode-entities)
       (set-marker body nil))))
 
 (defun nnweb-reference-search (search)
@@ -678,7 +681,8 @@ and `altavista'.")
       (while (re-search-forward "<A.*\\?id@\\([^\"]+\\)\">[0-9]+</A>" nil t)
        (replace-match "&lt;\\1&gt; " t)))
     (widen)
-    (nnweb-remove-markup)))
+    (nnweb-remove-markup)
+    (nnweb-decode-entities)))
 
 (defun nnweb-altavista-search (search &optional part)
   (url-insert-file-contents
@@ -733,18 +737,18 @@ and `altavista'.")
        (case-fold-search t)
        (active (cadr (assoc nnweb-group nnweb-group-alist)))
        Subject Score Date Newsgroups From
-       map url)
+       map url mid)
     (unless active
-      (push (list nnweb-group (setq active (cons 1 0)) 
+      (push (list nnweb-group (setq active (cons 1 0))
                  nnweb-type nnweb-search)
            nnweb-group-alist))
     ;; Go through all the article hits on this page.
     (goto-char (point-min))
     (while (re-search-forward
-            "a href=/groups\\(\\?[^ \">]*selm=[^ \">]*\\)" nil t)
-      (setq url
-           (concat (nnweb-definition 'address)
-                   (match-string 1)))
+           "a href=/groups\\(\\?[^ \">]*selm=\\([^ &\">]+\\)\\)" nil t)
+      (setq mid (match-string 2)
+           url (format 
+                "http://groups.google.com/groups?selm=%s&output=gplain" mid))
       (narrow-to-region (search-forward ">" nil t)
                        (search-forward "</a>" nil t))
       (nnweb-remove-markup)
@@ -765,7 +769,7 @@ and `altavista'.")
        (goto-char (point-max))
        (widen)
        (skip-chars-forward "- \t"))
-      (when (looking-at 
+      (when (looking-at
             "\\([0-9]+[/ ][A-Za-z]+[/ ][0-9]+\\)[ \t]*by[ \t]*\\([^<]*\\) - <a")
        (setq From (match-string 2)
              Date (match-string 1)))
@@ -777,9 +781,9 @@ and `altavista'.")
          (incf (cdr active))
          (make-full-mail-header
           (cdr active) (if Newsgroups
-                           (concat  "(" Newsgroups ") " Subject) 
+                           (concat  "(" Newsgroups ") " Subject)
                          Subject)
-          From Date Message-ID
+          From Date (or Message-ID mid)
           nil 0 0 url))
         map)
        (nnweb-set-hashtb (cadar map) (car map))))
@@ -787,7 +791,7 @@ and `altavista'.")
 
 (defun nnweb-google-reference (id)
   (let ((map (nnweb-google-parse-1 id)) header)
-    (setq nnweb-articles 
+    (setq nnweb-articles
          (nconc nnweb-articles map))
     (when (setq header (cadar map))
       (mm-with-unibyte-current-buffer
@@ -800,15 +804,15 @@ and `altavista'.")
     (set-buffer nnweb-buffer)
     (erase-buffer)
     (when (funcall (nnweb-definition 'search) nnweb-search)
-       (let ((more t))
-         (while more
-           (setq nnweb-articles
-                 (nconc nnweb-articles (nnweb-google-parse-1)))
-           ;; FIXME: There is more.
-           (setq more nil))
-         ;; Return the articles in the right order.
+      (let ((more t))
+       (while more
          (setq nnweb-articles
-               (sort nnweb-articles 'car-less-than-car))))))
+               (nconc nnweb-articles (nnweb-google-parse-1)))
+         ;; FIXME: There is more.
+         (setq more nil))
+       ;; Return the articles in the right order.
+       (setq nnweb-articles
+             (sort nnweb-articles 'car-less-than-car))))))
 
 (defun nnweb-google-search (search)
   (nnweb-insert
@@ -827,7 +831,7 @@ and `altavista'.")
 
 (defun nnweb-google-identity (url)
   "Return an unique identifier based on URL."
-  (if (string-match "seld=\\([0-9]+\\)" url)
+  (if (string-match "selm=\\([^ &>]+\\)" url)
       (match-string 1 url)
     url))
 
@@ -877,13 +881,13 @@ and `altavista'.")
   (goto-char (point-min))
   (while (re-search-forward "&\\(#[0-9]+\\|[a-z]+\\);" nil t)
     (let ((elem (if (eq (aref (match-string 1) 0) ?\#)
-                       (let ((c
-                              (string-to-number (substring
-                                                 (match-string 1) 1))))
-                         (if (mm-char-or-char-int-p c) c 32))
-                     (or (cdr (assq (intern (match-string 1))
-                                    w3-html-entities))
-                         ?#))))
+                   (let ((c
+                          (string-to-number (substring
+                                             (match-string 1) 1))))
+                     (if (mm-char-or-char-int-p c) c 32))
+                 (or (cdr (assq (intern (match-string 1))
+                                w3-html-entities))
+                     ?#))))
       (unless (stringp elem)
        (setq elem (char-to-string elem)))
       (replace-match elem t t))))