Add URL parsing tests from WHATWG (#3188)

Co-authored-by: Kar Petrosyan <92274156+karpetrosyan@users.noreply.github.com>

httpx/_urlparse.py

@@ -253,22 +253,27 @@ def urlparse(url: str = "", **kwargs: str | None) -> ParseResult:
         parsed_userinfo != "" or parsed_host != "" or parsed_port is not None
     )
     validate_path(path, has_scheme=has_scheme, has_authority=has_authority)
-    if has_authority:
+    if has_scheme or has_authority:
         path = normalize_path(path)
 
     # The GEN_DELIMS set is... : / ? # [ ] @
     # These do not need to be percent-quoted unless they serve as delimiters for the
     # specific component.
+    WHATWG_SAFE = '`{}%|^\\"'
 
     # For 'path' we need to drop ? and # from the GEN_DELIMS set.
-    parsed_path: str = quote(path, safe=SUB_DELIMS + ":/[]@")
+    parsed_path: str = quote(path, safe=SUB_DELIMS + WHATWG_SAFE + ":/[]@")
     # For 'query' we need to drop '#' from the GEN_DELIMS set.
     parsed_query: str | None = (
-        None if query is None else quote(query, safe=SUB_DELIMS + ":/?[]@")
+        None
+        if query is None
+        else quote(query, safe=SUB_DELIMS + WHATWG_SAFE + ":/?[]@")
     )
     # For 'fragment' we can include all of the GEN_DELIMS set.
     parsed_fragment: str | None = (
-        None if fragment is None else quote(fragment, safe=SUB_DELIMS + ":/?#[]@")
+        None
+        if fragment is None
+        else quote(fragment, safe=SUB_DELIMS + WHATWG_SAFE + ":/?#[]@")
     )
 
     # The parsed ASCII bytestrings are our canonical form.
@@ -321,7 +326,8 @@ def encode_host(host: str) -> str:
         # From https://datatracker.ietf.org/doc/html/rfc3986/#section-3.2.2
         #
         # reg-name    = *( unreserved / pct-encoded / sub-delims )
-        return quote(host.lower(), safe=SUB_DELIMS)
+        WHATWG_SAFE = '"`{}%|\\'
+        return quote(host.lower(), safe=SUB_DELIMS + WHATWG_SAFE)
 
     # IDNA hostnames
     try:
@@ -369,19 +375,17 @@ def validate_path(path: str, has_scheme: bool, has_authority: bool) -> None:
         # must either be empty or begin with a slash ("/") character."
         if path and not path.startswith("/"):
             raise InvalidURL("For absolute URLs, path must be empty or begin with '/'")
-    else:
+
+    if not has_scheme and not has_authority:
         # If a URI does not contain an authority component, then the path cannot begin
         # with two slash characters ("//").
         if path.startswith("//"):
-            raise InvalidURL(
-                "URLs with no authority component cannot have a path starting with '//'"
-            )
+            raise InvalidURL("Relative URLs cannot have a path starting with '//'")
+
         # In addition, a URI reference (Section 4.1) may be a relative-path reference,
         # in which case the first path segment cannot contain a colon (":") character.
-        if path.startswith(":") and not has_scheme:
-            raise InvalidURL(
-                "URLs with no scheme component cannot have a path starting with ':'"
-            )
+        if path.startswith(":"):
+            raise InvalidURL("Relative URLs cannot have a path starting with ':'")
 
 
 def normalize_path(path: str) -> str:

tests/models/test_url.py

@@ -230,8 +230,8 @@ def test_url_normalized_host():
 
 
 def test_url_percent_escape_host():
-    url = httpx.URL("https://exam%le.com/")
-    assert url.host == "exam%25le.com"
+    url = httpx.URL("https://exam le.com/")
+    assert url.host == "exam%20le.com"
 
 
 def test_url_ipv4_like_host():
@@ -415,17 +415,11 @@ def test_urlparse_with_invalid_path():
 
     with pytest.raises(httpx.InvalidURL) as exc:
         httpx.URL(path="//abc")
-    assert (
-        str(exc.value)
-        == "URLs with no authority component cannot have a path starting with '//'"
-    )
+    assert str(exc.value) == "Relative URLs cannot have a path starting with '//'"
 
     with pytest.raises(httpx.InvalidURL) as exc:
         httpx.URL(path=":abc")
-    assert (
-        str(exc.value)
-        == "URLs with no scheme component cannot have a path starting with ':'"
-    )
+    assert str(exc.value) == "Relative URLs cannot have a path starting with ':'"
 
 
 def test_url_with_relative_path():

tests/models/test_whatwg.py

@@ -0,0 +1,52 @@
+# The WHATWG have various tests that can be used to validate the URL parsing.
+#
+# https://url.spec.whatwg.org/
+
+import json
+
+import pytest
+
+from httpx._urlparse import urlparse
+
+# URL test cases from...
+# https://github.com/web-platform-tests/wpt/blob/master/url/resources/urltestdata.json
+with open("tests/models/whatwg.json", "r") as input:
+    test_cases = json.load(input)
+    test_cases = [
+        item
+        for item in test_cases
+        if not isinstance(item, str) and not item.get("failure")
+    ]
+
+
+@pytest.mark.parametrize("test_case", test_cases)
+def test_urlparse(test_case):
+    if test_case["href"] in ("a: foo.com", "lolscheme:x x#x%20x"):
+        # Skip these two test cases.
+        # WHATWG cases where are not using percent-encoding for the space character.
+        # Anyone know what's going on here?
+        return
+
+    p = urlparse(test_case["href"])
+
+    # Test cases include the protocol with the trailing ":"
+    protocol = p.scheme + ":"
+    # Include the square brackets for IPv6 addresses.
+    hostname = f"[{p.host}]" if ":" in p.host else p.host
+    # The test cases use a string representation of the port.
+    port = "" if p.port is None else str(p.port)
+    # I have nothing to say about this one.
+    path = p.path
+    # The 'search' and 'hash' components in the whatwg tests are semantic, not literal.
+    # Our parsing differentiates between no query/hash and empty-string query/hash.
+    search = "" if p.query in (None, "") else "?" + str(p.query)
+    hash = "" if p.fragment in (None, "") else "#" + str(p.fragment)
+
+    # URL hostnames are case-insensitive.
+    # We normalize these, unlike the WHATWG test cases.
+    assert protocol == test_case["protocol"]
+    assert hostname.lower() == test_case["hostname"].lower()
+    assert port == test_case["port"]
+    assert path == test_case["pathname"]
+    assert search == test_case["search"]
+    assert hash == test_case["hash"]