oneOf and discriminator value mismatch doesnot fail

[stevehu]

No description provided.

[Copilot]

Pull request overview

This PR fixes a bug (issue 1225) where oneOf validation with a discriminator mapping incorrectly succeeded when the discriminator-indicated schema failed validation but a different, non-discriminated schema happened to pass. The fix adds a new assertion in OneOfValidator to detect this mismatch and report a discriminator error.

Changes:

• Added a new condition in OneOfValidator.validate() to fail validation when the discriminator-mapped schema has errors but exactly one other schema passes, treating this as a discriminator mismatch.
• Updated the existing test discriminatorOneOfOneMatchWrongDiscriminatorShouldFail (renamed from ...ShouldSucceed) in OpenApi31Test to verify the fix using the petstore schema.
• Added a new test oneOfDiscriminatorEnabledWithDiscriminatorMismatch in DiscriminatorValidatorTest to verify the fix with an inline schema using explicit discriminator mappings.

Reviewed changes

Copilot reviewed 3 out of 3 changed files in this pull request and generated no comments.

File	Description
src/main/java/com/networknt/schema/keyword/OneOfValidator.java	Core fix: adds discriminator mismatch detection when the mapped schema fails but another schema passes in oneOf
src/test/java/com/networknt/schema/oas/OpenApi31Test.java	Updates existing test to expect failure and verify discriminator error keyword
src/test/java/com/networknt/schema/DiscriminatorValidatorTest.java	Adds new test for discriminator value mismatch with explicit mapping

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[stevehu]

After the fix, it breaks an existing test case. I am not sure if this fix is correct or not. Please pay more attention before merging this PR. Thanks.

[justin-tay]

The spec at https://spec.openapis.org/oas/v3.1.1.html#discriminator-object states:

Note that discriminator MUST NOT change the validation outcome of the schema.

So this change likely goes against the spec. Removing or adding the discriminator in a schema shouldn't change the validation outcome.

[stevehu]

Based on the comment from @justin-tay, we should drop this PR as the current behaviour is correct.