gh-102153: fix CVE-2023-24329

Lib/test/test_urlparse.py

@@ -686,6 +686,13 @@ def test_attributes_bad_scheme(self):
                         else:
                             self.assertEqual(p.scheme, "")
 
+    def test_attributes_bad_scheme_CVE_2023_24329(self):
+        """Check handling of invalid schemes that starts with blank characters."""
+        for parse in (urllib.parse.urlsplit, urllib.parse.urlparse):
+            url = " https://www.example.net"
+            p = parse(url)
+            self.assertEqual(p.scheme, "https")
+
     def test_attributes_without_netloc(self):
         # This example is straight from RFC 3261.  It looks like it
         # should allow the username, hostname, and port to be filled

Lib/urllib/parse.py

@@ -451,6 +451,7 @@ def urlsplit(url, scheme='', allow_fragments=True):
     Note that % escapes are not expanded.
     """
 
+    url = url.lstrip()
     url, scheme, _coerce_result = _coerce_args(url, scheme)
 
     for b in _UNSAFE_URL_BYTES_TO_REMOVE:

Misc/NEWS.d/next/Security/2023-03-06-22-48-08.gh-issue-102153.eiaVrE.rst

@@ -0,0 +1 @@
+fix CVE-2023-24329