Issue webcompat#767: don't add trailing slash; account for urls beginning with //

Author: daliacoss

tests/test_form.py

@@ -12,20 +12,22 @@ class TestForm(unittest.TestCase):
     def test_normalize_url(self):
 
         r = form.normalize_url('example.com')
-        self.assertEqual(r, 'http://example.com/')
+        self.assertEqual(r, 'http://example.com')
 
         r = form.normalize_url('http:/example.com')
-        self.assertEqual(r, 'http://example.com/')
+        self.assertEqual(r, 'http://example.com')
 
         r = form.normalize_url('https:/example.com')
-        self.assertEqual(r, 'https://example.com/')
+        self.assertEqual(r, 'https://example.com')
 
         r = form.normalize_url('http:example.com')
-        self.assertEqual(r, 'http://example.com/')
+        self.assertEqual(r, 'http://example.com')
 
         r = form.normalize_url('https:example.com')
-        self.assertEqual(r, 'https://example.com/')
+        self.assertEqual(r, 'https://example.com')
 
+        r = form.normalize_url('//example.com')
+        self.assertEqual(r, 'http://example.com')
 
     def test_domain_name(self):
 

webcompat/form.py

@@ -132,11 +132,10 @@ def normalize_url(url):
             url += '#' + parsed.fragment
     elif not parsed.scheme:
         # We assume that http is missing not https
-        url = 'http://%s' % (url)
-
-    # if url does not contain a path, ensure it has a trailing slash
-    if not urlparse.urlparse(url).path:
-        url += "/"
+        if url.startswith("//"):
+            url = "http://%s" % (url[2:])
+        else:
+            url = 'http://%s' % (url)
 
     return url