Improved detection of whether a URL is internal or not for the purpose

of broken link e-mails (referred links from images.google were being
misclassified).


git-svn-id: http://code.djangoproject.com/svn/django/trunk@3870 bcc190cf-cafb-0310-a4f2-bffc1f526a37

django/middleware/common.py

@@ -2,6 +2,7 @@ from django.conf import settings
 from django import http
 from django.core.mail import mail_managers
 import md5
+import re
 
 class CommonMiddleware(object):
     """
@@ -61,7 +62,7 @@ class CommonMiddleware(object):
                 # send a note to the managers.
                 domain = http.get_host(request)
                 referer = request.META.get('HTTP_REFERER', None)
-                is_internal = referer and (domain in referer)
+                is_internal = _is_internal_request(domain, referer)
                 path = request.get_full_path()
                 if referer and not _is_ignorable_404(path) and (is_internal or '?' not in referer):
                     ua = request.META.get('HTTP_USER_AGENT', '<none>')
@@ -88,3 +89,8 @@ def _is_ignorable_404(uri):
         if uri.endswith(end):
             return True
     return False
+
+def _is_internal_request(domain, referer):
+    "Return true if the referring URL is the same domain as the current request"
+    # Different subdomains are treated as different domains.
+    return referer is not None and re.match("^https?://%s/" % re.escape(domain), referer)