update tests

tests/test_atproto_client/models/tests/test_string_formats.py

@@ -14,18 +14,28 @@
 
 
 # TODO: 230 passed, 11 xfailed
-# not yet entirely sure on these remaining failures:
-# - AT-URI validation may be too strict
-# - DateTime validation needs to handle more ISO8601 formats
-# - Handle/NSID validation rules may need updating
+# These cases appear in both _valid.txt and _invalid.txt files.
+# Need investigation to determine if our validation is incorrect or if test data needs updating:
 SKIP_THESE_VALUES = [
-    'at://did:plc:asdf123',  # Appears valid - may need to update AT-URI validation rules
-    'at://did:plc:asdf123/com.atproto.feed.post',  # Appears valid - may need to update AT-URI validation rules
-    '1985-04-12T23:20:50.123Z',  # Valid ISO8601 - validator may be too strict
-    '1985-04-12T23:20:50.123-00:00',  # Valid ISO8601 with offset - validator may be too strict
-    '1985-04-12T23:20Z',  # Valid ISO8601 without seconds - validator may be too strict
-    'john.test',  # Appears valid - may need to update handle validation
-    'one.two.three',  # Appears valid - may need to update NSID validation
+    (
+        string_formats.AtUri,
+        'at://did:plc:asdf123',
+    ),  # Listed as both valid and invalid in AT-URI files under "enforces spec basics"
+    (
+        string_formats.AtUri,
+        'at://did:plc:asdf123/com.atproto.feed.post',
+    ),  # Same AT-URI pattern - appears in both valid/invalid files
+    (
+        string_formats.DateTime,
+        '1985-04-12T23:20:50.123Z',
+    ),  # Listed as "preferred" in valid but also appears in invalid under RFC-3339 section
+    (
+        string_formats.DateTime,
+        '1985-04-12T23:20:50.123-00:00',
+    ),  # Listed as "supported" in valid but marked invalid under timezone formats
+    (string_formats.DateTime, '1985-04-12T23:20Z'),  # Similar timezone format discrepancy between valid/invalid files
+    (string_formats.Handle, 'john.test'),  # Base pattern appears valid but numeric suffix versions are marked invalid
+    (string_formats.Nsid, 'one.two.three'),  # Same pattern - base form valid but numeric suffixes marked invalid
 ]
 
 
@@ -145,7 +155,7 @@ def test_string_format_validation(
     validator_type: type, field_name: str, error_keywords: List[str], invalid_value: str, valid_data: dict
 ) -> None:
     """Test validation for each string format type."""
-    if invalid_value in SKIP_THESE_VALUES:
+    if any(invalid_value == skip_value for _, skip_value in SKIP_THESE_VALUES):
         pytest.xfail(f'TODO: Fix validation for {invalid_value}')
 
     SomeTypeAdapter = TypeAdapter(validator_type)
@@ -222,3 +232,26 @@ class FooModel(BaseModel):
     assert isinstance(instance, FooModel)
     assert instance.handle == invalid_data['handle']
     assert instance.did == invalid_data['did']
+
+
+@pytest.mark.parametrize('validator_type,value', SKIP_THESE_VALUES)
+def test_skipped_validation_cases(validator_type: type, value: str) -> None:
+    """
+    Test each skipped case that we suspect is a discrepancy between valid/invalid test files
+    """
+    _TAdapter = TypeAdapter(validator_type)
+
+    # Should validate successfully with strict validation
+    validated = _TAdapter.validate_python(value, context={_OPT_IN_KEY: True})
+    assert validated == value
+
+    # Also verify it appears in the corresponding invalid test file
+    invalid_filename = {
+        string_formats.AtUri: 'aturi_syntax_invalid.txt',
+        string_formats.DateTime: 'datetime_syntax_invalid.txt',
+        string_formats.Handle: 'handle_syntax_invalid.txt',
+        string_formats.Nsid: 'nsid_syntax_invalid.txt',
+    }[validator_type]
+
+    invalid_cases = get_test_cases(invalid_filename)
+    assert value in invalid_cases, f'{value} not found in {invalid_filename} despite being marked as invalid'