gh-99418: Make urllib.parse.urlparse enforce that a scheme must begin with an alphabetical ASCII character. (#99421)

Prevent urllib.parse.urlparse from accepting schemes that don't begin with an alphabetical ASCII character.

RFC 3986 defines a scheme like this: `scheme = ALPHA *( ALPHA / DIGIT / "+" / "-" / "." )`
RFC 2234 defines an ALPHA like this: `ALPHA = %x41-5A / %x61-7A`

The WHATWG URL spec defines a scheme like this:
`"A URL-scheme string must be one ASCII alpha, followed by zero or more of ASCII alphanumeric, U+002B (+), U+002D (-), and U+002E (.)."`

Author: kenballus

Lib/test/test_urlparse.py

@@ -668,6 +668,24 @@ def test_attributes_bad_port(self):
                         with self.assertRaises(ValueError):
                             p.port
 
+    def test_attributes_bad_scheme(self):
+        """Check handling of invalid schemes."""
+        for bytes in (False, True):
+            for parse in (urllib.parse.urlsplit, urllib.parse.urlparse):
+                for scheme in (".", "+", "-", "0", "http&", "६http"):
+                    with self.subTest(bytes=bytes, parse=parse, scheme=scheme):
+                        url = scheme + "://www.example.net"
+                        if bytes:
+                            if url.isascii():
+                                url = url.encode("ascii")
+                            else:
+                                continue
+                        p = parse(url)
+                        if bytes:
+                            self.assertEqual(p.scheme, b"")
+                        else:
+                            self.assertEqual(p.scheme, "")
+
     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

@@ -460,7 +460,7 @@ def urlsplit(url, scheme='', allow_fragments=True):
     allow_fragments = bool(allow_fragments)
     netloc = query = fragment = ''
     i = url.find(':')
-    if i > 0:
+    if i > 0 and url[0].isascii() and url[0].isalpha():
         for c in url[:i]:
             if c not in scheme_chars:
                 break

Misc/NEWS.d/next/Library/2022-11-12-15-45-51.gh-issue-99418.FxfAXS.rst

@@ -0,0 +1,2 @@
+Fix bug in :func:`urllib.parse.urlparse` that causes URL schemes that begin
+with a digit, a plus sign, or a minus sign to be parsed incorrectly.