Fix an errant string identity comparison in a test

Short strings can be interned as an optimization in the Python
interpreter, so comparing them by identity (“is”, memory address) can
succeed in practice, but it is neither correct nor reliable when
comparing by equality (“==”, value) is intended.

Author: musicinmybrain

tests/integration/validation/test_exceptions.py

@@ -465,7 +465,7 @@ def test_parameter_default_value_custom_format_invalid(self, validator_v30):
 
         @oas30_format_checker.checks("custom")
         def validate(to_validate) -> bool:
-            return to_validate is "valid"
+            return to_validate == "valid"
 
         spec = {
             "openapi": "3.0.0",