From f1509306d266a091c6c12ec3b78b8eaaa9d0aff9 Mon Sep 17 00:00:00 2001
From: "guido@google.com" <guido@google.com>
Date: Mon, 28 Mar 2011 13:47:01 -0700
Subject: Add tests for the urllib[2] vulnerability. Change to raise
 exceptions.

---
 Lib/test/test_urllib.py  | 14 ++++++++++++++
 Lib/test/test_urllib2.py | 21 +++++++++++++++++++++
 Lib/urllib.py            | 10 ++++++----
 Lib/urllib2.py           |  5 ++++-
 4 files changed, 45 insertions(+), 5 deletions(-)

diff --git a/Lib/test/test_urllib.py b/Lib/test/test_urllib.py
index 294ed5e..72db1ef 100644
--- a/Lib/test/test_urllib.py
+++ b/Lib/test/test_urllib.py
@@ -122,6 +122,20 @@ class urlopen_HttpTests(unittest.TestCase):
         finally:
             self.unfakehttp()
 
+    def test_invalid_redirect(self):
+        # urlopen() should raise IOError for many error codes.
+        self.fakehttp("""HTTP/1.1 302 Found
+Date: Wed, 02 Jan 2008 03:03:54 GMT
+Server: Apache/1.3.33 (Debian GNU/Linux) mod_ssl/2.8.22 OpenSSL/0.9.7e
+Location: file:README
+Connection: close
+Content-Type: text/html; charset=iso-8859-1
+""")
+        try:
+            self.assertRaises(IOError, urllib.urlopen, "http://python.org/")
+        finally:
+            self.unfakehttp()
+
     def test_empty_socket(self):
         """urlopen() raises IOError if the underlying socket does not send any
         data. (#1680230) """
diff --git a/Lib/test/test_urllib2.py b/Lib/test/test_urllib2.py
index 96a7db9..8e9f1b2 100644
--- a/Lib/test/test_urllib2.py
+++ b/Lib/test/test_urllib2.py
@@ -857,6 +857,27 @@ class HandlerTests(unittest.TestCase):
             self.assertEqual(count,
                              urllib2.HTTPRedirectHandler.max_redirections)
 
+    def test_invalid_redirect(self):
+        from_url = "http://example.com/a.html"
+        valid_schemes = ['http', 'https', 'ftp']
+        invalid_schemes = ['file', 'imap', 'ldap']
+        schemeless_url = "example.com/b.html"
+        h = urllib2.HTTPRedirectHandler()
+        o = h.parent = MockOpener()
+        req = Request(from_url)
+
+        for scheme in invalid_schemes:
+            invalid_url = scheme + '://' + schemeless_url
+            self.assertRaises(urllib2.HTTPError, h.http_error_302,
+                              req, MockFile(), 302, "Security Loophole",
+                              MockHeaders({"location": invalid_url}))
+
+        for scheme in valid_schemes:
+            valid_url = scheme + '://' + schemeless_url
+            h.http_error_302(req, MockFile(), 302, "That's fine",
+                MockHeaders({"location": valid_url}))
+            self.assertEqual(o.req.get_full_url(), valid_url)
+
     def test_cookie_redirect(self):
         # cookies shouldn't leak into redirected requests
         from cookielib import CookieJar
diff --git a/Lib/urllib.py b/Lib/urllib.py
index b835f52..97597f4 100644
--- a/Lib/urllib.py
+++ b/Lib/urllib.py
@@ -638,7 +638,8 @@ class FancyURLopener(URLopener):
             newurl = headers['uri']
         else:
             return
-
+        void = fp.read()
+        fp.close()
         # In case the server sent a relative URL, join with original:
         newurl = basejoin(self.type + ":" + url, newurl)
 
@@ -648,10 +649,11 @@ class FancyURLopener(URLopener):
         if not (newurl_lower.startswith('http://') or
                 newurl_lower.startswith('https://') or
                 newurl_lower.startswith('ftp://')):
-            return
+            raise IOError('redirect error', errcode,
+                          errmsg + " - Redirection to url '%s' is not allowed" %
+                          newurl,
+                          headers)
 
-        void = fp.read()
-        fp.close()
         return self.open(newurl)
 
     def http_error_301(self, url, fp, errcode, errmsg, headers, data=None):
diff --git a/Lib/urllib2.py b/Lib/urllib2.py
index 0bb69a0..a537d36 100644
--- a/Lib/urllib2.py
+++ b/Lib/urllib2.py
@@ -561,7 +561,10 @@ class HTTPRedirectHandler(BaseHandler):
         if not (newurl_lower.startswith('http://') or
                 newurl_lower.startswith('https://') or
                 newurl_lower.startswith('ftp://')):
-            return
+            raise HTTPError(newurl, code,
+                            msg + " - Redirection to url '%s' is not allowed" %
+                            newurl,
+                            headers, fp)
 
         # XXX Probably want to forget about the state of the current
         # request, although that might interact poorly with other
-- 
cgit v0.12