bpo-19094: Raise TypeError in urljoin(), urlparse(), and urlsplit() for inappropriate types

Co-authored-by: Vajrasky Kok <sky.kok@speaklikeaking.com>

Author: jacobtylerwalls

Lib/test/test_urlparse.py

@@ -847,6 +847,32 @@ def test_mixed_types_rejected(self):
         with self.assertRaisesRegex(TypeError, "Cannot mix str"):
             urllib.parse.urljoin(b"http://python.org", "http://python.org")
 
+    def test_forbidden_types(self):
+        with self.assertRaisesRegex(
+                TypeError,
+                "Expected string or bytes: got <class 'list'>"):
+            urllib.parse.urljoin('http://www.python.org', [])
+        with self.assertRaisesRegex(
+                TypeError,
+                "Expected string or bytes: got <class 'list'>"):
+            urllib.parse.urljoin([], b'docs')
+        with self.assertRaisesRegex(
+                TypeError,
+                "Expected string or bytes: got <class 'NoneType'>"):
+            urllib.parse.urlparse(b'www.python.org', None)
+        with self.assertRaisesRegex(
+                TypeError,
+                "Expected string or bytes: got <class 'dict'>"):
+            urllib.parse.urlparse({}, '')
+        with self.assertRaisesRegex(
+                TypeError,
+                "Expected string or bytes: got <class 'int'>"):
+            urllib.parse.urlsplit(0, 'http')
+        with self.assertRaisesRegex(
+                TypeError,
+                "Expected string or bytes: got <class 'NoneType'>"):
+            urllib.parse.urlsplit('http://www.python.org', None)
+
     def _check_result_type(self, str_type):
         num_args = len(str_type._fields)
         bytes_type = str_type._encoded_counterpart

Lib/urllib/parse.py

@@ -383,6 +383,10 @@ def urlparse(url, scheme='', allow_fragments=True):
 
     Note that % escapes are not expanded.
     """
+    if not isinstance(url, (str, bytes, bytearray)):
+        raise TypeError(f'Expected string or bytes: got {type(url)}')
+    if not isinstance(scheme, (str, bytes, bytearray)):
+        raise TypeError(f'Expected string or bytes: got {type(scheme)}')
     url, scheme, _coerce_result = _coerce_args(url, scheme)
     splitresult = urlsplit(url, scheme, allow_fragments)
     scheme, netloc, url, query, fragment = splitresult
@@ -451,6 +455,10 @@ def urlsplit(url, scheme='', allow_fragments=True):
 
     Note that % escapes are not expanded.
     """
+    if not isinstance(url, (str, bytes, bytearray)):
+        raise TypeError(f'Expected string or bytes: got {type(url)}')
+    if not isinstance(scheme, (str, bytes, bytearray)):
+        raise TypeError(f'Expected string or bytes: got {type(scheme)}')
 
     url, scheme, _coerce_result = _coerce_args(url, scheme)
 
@@ -514,6 +522,11 @@ def urlunsplit(components):
 def urljoin(base, url, allow_fragments=True):
     """Join a base URL and a possibly relative URL to form an absolute
     interpretation of the latter."""
+    if not isinstance(base, (str, bytes, bytearray)):
+        raise TypeError(f'Expected string or bytes: got {type(base)}')
+    if not isinstance(url, (str, bytes, bytearray)):
+        raise TypeError(f'Expected string or bytes: got {type(url)}')
+
     if not base:
         return url
     if not url:

Misc/ACKS

@@ -1871,6 +1871,7 @@ Wojtek Walczak
 Charles Waldman
 Richard Walker
 Larry Wall
+Jacob Walls
 Kevin Walzer
 Rodrigo Steinmuller Wanderley
 Dingyuan Wang

Misc/NEWS.d/next/Library/2021-06-11-20-00-16.bpo-19094.rMRoIL.rst

@@ -0,0 +1,3 @@
+Raised ``TypeError`` when providing anything but a string or bytes
+object to :func:`urllib.parse.urljoin`, :func:`urllib.parse.urlparse`,
+or :func:`urllib.parse.urlsplit`.