python · orsenthil · Apr 29, 2021 · Apr 25, 2021 · Apr 28, 2021 · Apr 28, 2021
diff --git a/Doc/library/urllib.parse.rst b/Doc/library/urllib.parse.rst
@@ -312,6 +312,10 @@ or on combining URL components into a URL string.
    ``#``, ``@``, or ``:`` will raise a :exc:`ValueError`. If the URL is
    decomposed before parsing, no error will be raised.
 
+   Following the specification in WHATWG which updates RFC 3986, ASCII newline
+   ``\n``, ``\r`` or ``\r\n`` and tab ``\t`` characters are stripped from
+   the url.
+
    .. versionchanged:: 3.6
       Out-of-range port numbers now raise :exc:`ValueError`, instead of
       returning :const:`None`.
@@ -320,6 +324,8 @@ or on combining URL components into a URL string.
       Characters that affect netloc parsing under NFKC normalization will
       now raise :exc:`ValueError`.
 
+   .. versionchanged:: 3.10
+      ASCII newline and tab characters are stripped from the url.
 
 .. function:: urlunsplit(parts)
 

diff --git a/Lib/test/test_urlparse.py b/Lib/test/test_urlparse.py
@@ -566,6 +566,21 @@ def test_urlsplit_attributes(self):
         self.assertEqual(p.port, 80)
         self.assertEqual(p.geturl(), url)
 
+        # Remove ASCII tabs and newlines from input
+
+        url = "http://www.python.org/java\nscript:\talert('msg\r\n')/#frag"
+        p = urllib.parse.urlsplit(url)
+        self.assertEqual(p.scheme, "http")
+        self.assertEqual(p.netloc, "www.python.org")
+        self.assertEqual(p.path, "/javascript:alert('msg')/")
+        self.assertEqual(p.query, "")
+        self.assertEqual(p.fragment, "frag")
+        self.assertEqual(p.username, None)
+        self.assertEqual(p.password, None)
+        self.assertEqual(p.hostname, "www.python.org")
+        self.assertEqual(p.port, None)
+        self.assertEqual(p.geturl(), "http://www.python.org/javascript:alert('msg')/#frag")
+
         # And check them all again, only with bytes this time
         url = b"HTTP://WWW.PYTHON.ORG/doc/#frag"
         p = urllib.parse.urlsplit(url)
@@ -606,6 +621,19 @@ def test_urlsplit_attributes(self):
         self.assertEqual(p.port, 80)
         self.assertEqual(p.geturl(), url)
 
+        url = b"http://www.python.org/java\nscript:\talert('msg\r\n')/#frag"
+        p = urllib.parse.urlsplit(url)
+        self.assertEqual(p.scheme, b"http")
+        self.assertEqual(p.netloc, b"www.python.org")
+        self.assertEqual(p.path, b"/javascript:alert('msg')/")
+        self.assertEqual(p.query, b"")
+        self.assertEqual(p.fragment, b"frag")
+        self.assertEqual(p.username, None)
+        self.assertEqual(p.password, None)
+        self.assertEqual(p.hostname, b"www.python.org")
+        self.assertEqual(p.port, None)
+        self.assertEqual(p.geturl(), b"http://www.python.org/javascript:alert('msg')/#frag")
+
         # Verify an illegal port raises ValueError
         url = b"HTTP://WWW.PYTHON.ORG:65536/doc/#frag"
         p = urllib.parse.urlsplit(url)

diff --git a/Lib/urllib/parse.py b/Lib/urllib/parse.py
@@ -469,6 +469,9 @@ def urlsplit(url, scheme='', allow_fragments=True):
         else:
             scheme, url = url[:i].lower(), url[i+1:]
 
+    _unsafe_chars_to_remove = ['\t', '\r', '\n']
+    url = url.translate({ord(c): None for c in _unsafe_chars_to_remove})
+
     if url[:2] == '//':
         netloc, url = _splitnetloc(url, 2)
         if (('[' in netloc and ']' not in netloc) or

diff --git a/Misc/NEWS.d/next/Security/2021-04-25-07-46-37.bpo-43882.Jpwx85.rst b/Misc/NEWS.d/next/Security/2021-04-25-07-46-37.bpo-43882.Jpwx85.rst
@@ -0,0 +1,6 @@
+Presence newline or tab characters in URL allowed attackers to write scripts
+in URL, hijack the the the web-server.
+
+Following the controlling specification for URLs defined by WHATWG
+urllib.parse strips ASCII newline and tabs from the url, preventing such
+attacks.