Origin: upstream
Last-Update: 2014-05-14
Subject: Malformed URLs from user input incorrectly validated (CVE-2014-3730)

The validation for redirects did not correctly validate some malformed
URLs, which are accepted by some browsers. This allows a user to be
redirected to an unsafe URL unexpectedly.

Django relies on user input in some cases (e.g.
:func:`django.contrib.auth.views.login`, ``django.contrib.comments``,
and :doc:`i18n </topics/i18n/index>`) to redirect the user to an "on
success" URL. The security checks for these redirects (namely
``django.util.http.is_safe_url()``) did not correctly validate some
malformed URLs, such as `http:\\\\\\djangoproject.com`, which are
accepted by some browsers with more liberal URL parsing.

To remedy this, the validation in ``is_safe_url()`` has been tightened
to be able to handle and correctly validate these malformed URLs.

--- a/django/contrib/auth/tests/views.py
+++ b/django/contrib/auth/tests/views.py
@@ -307,8 +307,10 @@
 
         # Those URLs should not pass the security check
         for bad_url in ('http://example.com',
+                        'http:///example.com',
                         'https://example.com',
                         'ftp://exampel.com',
+                        '///example.com',
                         '//example.com',
                         'javascript:alert("XSS")'):
 
@@ -330,8 +332,8 @@
                          '/view/?param=https://example.com',
                          '/view?param=ftp://exampel.com',
                          'view/?param=//example.com',
-                         'https:///',
-                         'HTTPS:///',
+                         'https://testserver/',
+                         'HTTPS://testserver/',
                          '//testserver/',
                          '/url%20with%20spaces/'):  # see ticket #12534
             safe_url = '%(url)s?%(next)s=%(good_url)s' % {
@@ -467,8 +469,10 @@
 
         # Those URLs should not pass the security check
         for bad_url in ('http://example.com',
+                        'http:///example.com',
                         'https://example.com',
                         'ftp://exampel.com',
+                        '///example.com',
                         '//example.com',
                         'javascript:alert("XSS")'):
             nasty_url = '%(url)s?%(next)s=%(bad_url)s' % {
@@ -488,8 +492,8 @@
                          '/view/?param=https://example.com',
                          '/view?param=ftp://exampel.com',
                          'view/?param=//example.com',
-                         'https:///',
-                         'HTTPS:///',
+                         'https://testserver/',
+                         'HTTPS://testserver/',
                          '//testserver/',
                          '/url%20with%20spaces/'):  # see ticket #12534
             safe_url = '%(url)s?%(next)s=%(good_url)s' % {
--- a/django/utils/http.py
+++ b/django/utils/http.py
@@ -234,6 +234,18 @@
     """
     if not url:
         return False
+    # Chrome treats \ completely as /
+    url = url.replace('\\', '/')
+    # Chrome considers any URL with more than two slashes to be absolute, but
+    # urlaprse is not so flexible. Treat any url with three slashes as unsafe.
+    if url.startswith('///'):
+        return False
     url_info = urlparse.urlparse(url)
+    # Forbid URLs like http:///example.com - with a scheme, but without a hostname.
+    # In that URL, example.com is not the hostname but, a path component. However,
+    # Chrome will still consider example.com to be the hostname, so we must not
+    # allow this syntax.
+    if not url_info[1] and url_info[0]:
+        return False
     return (not url_info[1] or url_info[1] == host) and \
         (not url_info[0] or url_info[0] in ['http', 'https'])
--- a/tests/regressiontests/utils/http.py
+++ b/tests/regressiontests/utils/http.py
@@ -122,3 +122,33 @@
         for n, b36 in [(0, '0'), (1, '1'), (42, '16'), (818469960, 'django')]:
             self.assertEqual(http.int_to_base36(n), b36)
             self.assertEqual(http.base36_to_int(b36), n)
+
+    def test_is_safe_url(self):
+        for bad_url in ('http://example.com',
+                        'http:///example.com',
+                        'https://example.com',
+                        'ftp://exampel.com',
+                        r'\\example.com',
+                        r'\\\example.com',
+                        r'/\\/example.com',
+                        r'\\\example.com',
+                        r'\\example.com',
+                        r'\\//example.com',
+                        r'/\/example.com',
+                        r'\/example.com',
+                        r'/\example.com',
+                        'http:///example.com',
+                        'http:/\//example.com',
+                        'http:\/example.com',
+                        'http:/\example.com',
+                        'javascript:alert("XSS")'):
+            self.assertFalse(http.is_safe_url(bad_url, host='testserver'), "%s should be blocked" % bad_url)
+        for good_url in ('/view/?param=http://example.com',
+                     '/view/?param=https://example.com',
+                     '/view?param=ftp://exampel.com',
+                     'view/?param=//example.com',
+                     'https://testserver/',
+                     'HTTPS://testserver/',
+                     '//testserver/',
+                     '/url%20with%20spaces/'):
+            self.assertTrue(http.is_safe_url(good_url, host='testserver'), "%s should be allowed" % good_url)