From: Senthil Kumaran <orsenthil@gmail.com>
Date: Mon, 22 Nov 2010 05:04:33 +0000 (+0000)
Subject: Merged revisions 86676 via svnmerge from
X-Git-Tag: v2.7.1~41
X-Git-Url: https://granicus.if.org/sourcecode?a=commitdiff_plain;h=0b7cac11bfd3061ac78c25e5ae118faf16eb9924;p=python

Merged revisions 86676 via svnmerge from
svn+ssh://pythondev@svn.python.org/python/branches/py3k

........
  r86676 | senthil.kumaran | 2010-11-22 12:48:26 +0800 (Mon, 22 Nov 2010) | 4 lines

  Fix Issue4493 - urllib2 adds '/' to the path component of url, when it does not
  starts with one. This behavior is exhibited by browser and other clients.
........
---

diff --git a/Lib/test/test_urllib2.py b/Lib/test/test_urllib2.py
index 46db6b4257..b74320d2ac 100644
--- a/Lib/test/test_urllib2.py
+++ b/Lib/test/test_urllib2.py
@@ -838,6 +838,25 @@ class HandlerTests(unittest.TestCase):
             p_ds_req = h.do_request_(ds_req)
             self.assertEqual(p_ds_req.unredirected_hdrs["Host"],"example.com")
 
+    def test_fixpath_in_weirdurls(self):
+        # Issue4493: urllib2 to supply '/' when to urls where path does not
+        # start with'/'
+
+        h = urllib2.AbstractHTTPHandler()
+        o = h.parent = MockOpener()
+
+        weird_url = 'http://www.python.org?getspam'
+        req = Request(weird_url)
+        newreq = h.do_request_(req)
+        self.assertEqual(newreq.get_host(),'www.python.org')
+        self.assertEqual(newreq.get_selector(),'/?getspam')
+
+        url_without_path = 'http://www.python.org'
+        req = Request(url_without_path)
+        newreq = h.do_request_(req)
+        self.assertEqual(newreq.get_host(),'www.python.org')
+        self.assertEqual(newreq.get_selector(),'')
+
     def test_errors(self):
         h = urllib2.HTTPErrorProcessor()
         o = h.parent = MockOpener()
diff --git a/Lib/urllib.py b/Lib/urllib.py
index d85dedbb64..1553f9dcb2 100644
--- a/Lib/urllib.py
+++ b/Lib/urllib.py
@@ -1052,7 +1052,12 @@ def splithost(url):
         _hostprog = re.compile('^//([^/?]*)(.*)$')
 
     match = _hostprog.match(url)
-    if match: return match.group(1, 2)
+    if match:
+        host_port = match.group(1)
+        path = match.group(2)
+        if path and not path.startswith('/'):
+            path = '/' + path
+        return host_port, path
     return None, url
 
 _userprog = None
diff --git a/Misc/NEWS b/Misc/NEWS
index 315b55694b..cb9e489716 100644
--- a/Misc/NEWS
+++ b/Misc/NEWS
@@ -13,6 +13,9 @@ Core and Builtins
 Library
 -------
 
+- Issue #4493: urllib2 adds '/' in front of path components which does not
+  start with '/. Common behavior exhibited by browsers and other clients.
+
 - Issue #6378: idle.bat now runs with the appropriate Python version rather than
   the system default. Patch by Sridhar Ratnakumar.