--- Doc/library/urlparse.rst | 14 ++ Doc/whatsnew/2.7.rst | 7 + Lib/test/test_urlparse.py | 49 ++++++++++ Lib/urlparse.py | 12 ++ Misc/NEWS.d/next/Security/2021-04-25-07-46-37.bpo-43882.Jpwx85.rst | 6 + 5 files changed, 88 insertions(+) --- a/Doc/library/urlparse.rst +++ b/Doc/library/urlparse.rst @@ -248,6 +248,10 @@ The :mod:`urlparse` module defines the f decomposed before parsing, or is not a Unicode string, no error will be raised. + Following the `WHATWG spec`_ that updates RFC 3986, ASCII + newline ``\n``, ``\r`` and tab ``\t`` characters are stripped + from the URL. + .. versionadded:: 2.2 .. versionchanged:: 2.5 @@ -257,6 +261,10 @@ The :mod:`urlparse` module defines the f Characters that affect netloc parsing under NFKC normalization will now raise :exc:`ValueError`. + .. versionchanged:: 3.6.14 + ASCII newline and tab characters are stripped from the URL. + +.. _WHATWG spec: https://url.spec.whatwg.org/#concept-basic-url-parser .. function:: urlunsplit(parts) @@ -308,6 +316,11 @@ The :mod:`urlparse` module defines the f .. seealso:: + `WHATWG`_ - URL Living standard + Working Group for the URL Standard that defines URLs, + domains, IP addresses, the application/x-www-form-urlencoded format, + and their API. + :rfc:`3986` - Uniform Resource Identifiers This is the current standard (STD66). Any changes to urlparse module should conform to this. Certain deviations could be observed, which are @@ -332,6 +345,7 @@ The :mod:`urlparse` module defines the f :rfc:`1738` - Uniform Resource Locators (URL) This specifies the formal syntax and semantics of absolute URLs. +.. _WHATWG: https://url.spec.whatwg.org/ .. _urlparse-result-object: --- a/Doc/whatsnew/2.7.rst +++ b/Doc/whatsnew/2.7.rst @@ -175,6 +175,13 @@ channel. We reuse the ftp server IP add requiring the old behavior, set a ``trust_server_pasv_ipv4_address`` attribute on your FTP instance to ``True``. (See :issue:`43285`) +The presence of newline or tab characters in parts of a URL allows for some +forms of attacks. Following the WHATWG specification that updates RFC 3986, +ASCII newline ``\n``, ``\r`` and tab ``\t`` characters are stripped from the +URL by the parser :func:`urlparse` preventing such attacks. The removal +characters are controlled by a new module level variable +``urlparse._UNSAFE_URL_BYTES_TO_REMOVE``. (See :issue:`43882`) + Python 3.1 Features ======================= --- a/Lib/test/test_urlparse.py +++ b/Lib/test/test_urlparse.py @@ -492,6 +492,55 @@ class UrlParseTestCase(unittest.TestCase p = urlparse.urlsplit(url) self.assertEqual(p.port, None) + def test_urlsplit_remove_unsafe_bytes(self): + # Remove ASCII tabs and newlines from input, for http common case scenario. + url = "h\nttp://www.python\n.org\t/java\nscript:\talert('msg\r\n')/?query\n=\tsomething#frag\nment" + p = urlparse.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, "query=something") + self.assertEqual(p.fragment, "fragment") + 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')/?query=something#fragment") + + # Remove ASCII tabs and newlines from input as bytes, for http common case scenario. + url = b"h\nttp://www.python\n.org\t/java\nscript:\talert('msg\r\n')/?query\n=\tsomething#frag\nment" + p = urlparse.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"query=something") + self.assertEqual(p.fragment, b"fragment") + 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')/?query=something#fragment") + + # any scheme + url = "x-new-scheme\t://www.python\n.org\t/java\nscript:\talert('msg\r\n')/?query\n=\tsomething#frag\nment" + p = urlparse.urlsplit(url) + self.assertEqual(p.geturl(), "x-new-scheme://www.python.org/javascript:alert('msg')/?query=something#fragment") + + # Remove ASCII tabs and newlines from input as bytes, any scheme. + url = b"x-new-scheme\t://www.python\n.org\t/java\nscript:\talert('msg\r\n')/?query\n=\tsomething#frag\nment" + p = urlparse.urlsplit(url) + self.assertEqual(p.geturl(), b"x-new-scheme://www.python.org/javascript:alert('msg')/?query=something#fragment") + + # Unsafe bytes is not returned from urlparse cache. + # scheme is stored after parsing, sending an scheme with unsafe bytes *will not* return an unsafe scheme + url = "https://www.python\n.org\t/java\nscript:\talert('msg\r\n')/?query\n=\tsomething#frag\nment" + scheme = "htt\nps" + for _ in range(2): + p = urlparse.urlsplit(url, scheme=scheme) + self.assertEqual(p.scheme, "https") + self.assertEqual(p.geturl(), "https://www.python.org/javascript:alert('msg')/?query=something#fragment") + + def test_issue14072(self): p1 = urlparse.urlsplit('tel:+31-641044153') self.assertEqual(p1.scheme, 'tel') --- a/Lib/urlparse.py +++ b/Lib/urlparse.py @@ -62,6 +62,9 @@ scheme_chars = ('abcdefghijklmnopqrstuvw '0123456789' '+-.') +# Unsafe bytes to be removed per WHATWG spec +_UNSAFE_URL_BYTES_TO_REMOVE = ['\t', '\r', '\n'] + MAX_CACHE_SIZE = 20 _parse_cache = {} @@ -184,12 +187,21 @@ def _checknetloc(netloc): "under NFKC normalization" % netloc) + +def _remove_unsafe_bytes_from_url(url): + for b in _UNSAFE_URL_BYTES_TO_REMOVE: + url = url.replace(b, "") + return url + + def urlsplit(url, scheme='', allow_fragments=True): """Parse a URL into 5 components: :///?# Return a 5-tuple: (scheme, netloc, path, query, fragment). Note that we don't break the components up in smaller bits (e.g. netloc is a single string) and we don't expand % escapes.""" + url = _remove_unsafe_bytes_from_url(url) + scheme = _remove_unsafe_bytes_from_url(scheme) allow_fragments = bool(allow_fragments) key = url, scheme, allow_fragments, type(url), type(scheme) cached = _parse_cache.get(key, None) --- /dev/null +++ b/Misc/NEWS.d/next/Security/2021-04-25-07-46-37.bpo-43882.Jpwx85.rst @@ -0,0 +1,6 @@ +The presence of newline or tab characters in parts of a URL could allow +some forms of attacks. + +Following the controlling specification for URLs defined by WHATWG +:func:`urlparse` now removes ASCII newlines and tabs from URLs, +preventing such attacks.