Always encode forward slashes as %2F in query parameters

httpx/_urlparse.py

@@ -260,8 +260,10 @@ def urlparse(url: str = "", **kwargs: typing.Optional[str]) -> ParseResult:
     # For 'path' we need to drop ? and # from the GEN_DELIMS set.
     parsed_path: str = quote(path, safe=SUB_DELIMS + ":/[]@")
     # For 'query' we need to drop '#' from the GEN_DELIMS set.
+    # We also exclude '/' because it is more robust to replace it with a percent
+    # encoding despite it not being a requirement of the spec.
     parsed_query: typing.Optional[str] = (
-        None if query is None else quote(query, safe=SUB_DELIMS + ":/?[]@")
+        None if query is None else quote(query, safe=SUB_DELIMS + ":?[]@")
     )
     # For 'fragment' we can include all of the GEN_DELIMS set.
     parsed_fragment: typing.Optional[str] = (
@@ -452,11 +454,11 @@ def urlencode(items: typing.List[typing.Tuple[str, str]]) -> str:
     #
     # https://github.com/python/cpython/blob/b2f7b2ef0b5421e01efb8c7bee2ef95d3bab77eb/Lib/urllib/parse.py#L926
     #
-    # Note that we use '%20' encoding for spaces, and treat '/' as a safe
-    # character. This means our query params have the same escaping as other
-    # characters in the URL path. This is slightly different to `requests`,
-    # but is the behaviour that browsers use.
+    # Note that we use '%20' encoding for spaces. and '%2F  for '/'.
+    # This is slightly different than `requests`, but is the behaviour that browsers use.
     #
-    # See https://github.com/encode/httpx/issues/2536 and
-    # https://docs.python.org/3/library/urllib.parse.html#urllib.parse.urlencode
-    return "&".join([quote(k) + "=" + quote(v) for k, v in items])
+    # See
+    # - https://github.com/encode/httpx/issues/2536
+    # - https://github.com/encode/httpx/issues/2721
+    # - https://docs.python.org/3/library/urllib.parse.html#urllib.parse.urlencode
+    return "&".join([quote(k, safe="") + "=" + quote(v, safe="") for k, v in items])

tests/models/test_url.py

@@ -360,10 +360,10 @@ def test_url_query_encoding():
     and https://github.com/encode/httpx/discussions/2460
     """
     url = httpx.URL("https://www.example.com/?a=b c&d=e/f")
-    assert url.raw_path == b"/?a=b%20c&d=e/f"
+    assert url.raw_path == b"/?a=b%20c&d=e%2Ff"
 
     url = httpx.URL("https://www.example.com/", params={"a": "b c", "d": "e/f"})
-    assert url.raw_path == b"/?a=b%20c&d=e/f"
+    assert url.raw_path == b"/?a=b%20c&d=e%2Ff"
 
 
 def test_url_with_url_encoded_path():

tests/test_urlparse.py

@@ -141,7 +141,7 @@ def test_param_does_not_require_encoding():
 
 def test_param_with_existing_escape_requires_encoding():
     url = httpx.URL("http://webservice", params={"u": "http://example.com?q=foo%2Fa"})
-    assert str(url) == "http://webservice?u=http%3A//example.com%3Fq%3Dfoo%252Fa"
+    assert str(url) == "http://webservice?u=http%3A%2F%2Fexample.com%3Fq%3Dfoo%252Fa"
 
 
 # Tests for invalid URLs
@@ -264,9 +264,9 @@ def test_path_percent_encoding():
 def test_query_percent_encoding():
     # Test percent encoding for SUB_DELIMS ALPHA NUM and allowable GEN_DELIMS
     url = httpx.URL("https://example.com/?!$&'()*+,;= abc ABC 123 :/[]@" + "?")
-    assert url.raw_path == b"/?!$&'()*+,;=%20abc%20ABC%20123%20:/[]@?"
+    assert url.raw_path == b"/?!$&'()*+,;=%20abc%20ABC%20123%20:%2F[]@?"
     assert url.path == "/"
-    assert url.query == b"!$&'()*+,;=%20abc%20ABC%20123%20:/[]@?"
+    assert url.query == b"!$&'()*+,;=%20abc%20ABC%20123%20:%2F[]@?"
     assert url.fragment == ""