Transform 3.0 schema with allOf + nullable into oneOf

[alexrjones]

Fixes #229

Per the linked issue, if a schema has nullable: true and an allOf component, when transforming the nullability property for JSON Schema, the transform should move the allOf component under a oneOf and add {"type": "null"} to the oneOf, such that null validates against the schema.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 97.62%. Comparing base (0db7586) to head (33d377d).
⚠️ Report is 3 commits behind head on main.

Additional details and impacted files

@@           Coverage Diff           @@
##             main     #230   +/-   ##
=======================================
  Coverage   97.61%   97.62%           
=======================================
  Files          56       56           
  Lines        5240     5256   +16     
=======================================
+ Hits         5115     5131   +16     
  Misses        125      125           

Flag	Coverage Δ
unittests	97.62% <100.00%> (+<0.01%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[daveshanley]

This PR is safe and correctly handles the edge case of OpenAPI 3.0 schemas with both nullable: true and allOf.

Safety for OpenAPI 3.1/3.2: The transformation is gated behind version < 3.05 in schema_compiler.go:76-78, so it only affects OpenAPI 3.0 specs. OpenAPI 3.1+ uses native JSON Schema type arrays instead of the nullable keyword, so this code path is never triggered.

Semantic correctness: The transformation from allOf + nullable to oneOf: [{allOf: ...}, {type: null}] correctly expresses 'value must satisfy allOf constraints OR be null' in JSON Schema.

Code quality: Good test coverage with edge cases (existing oneOf arrays). The recursive transform handles nested schemas correctly.

Minor note: There's slight redundancy where null appears in both the type array and oneOf, but this is harmless and doesn't affect validation correctness.

^^ this was claude, I never asked it to do this, but well, it decided to. might as well leave it as it's not wrong.

[alexrjones]

Thanks @daveshanley!

Minor note: There's slight redundancy where null appears in both the type array and oneOf, but this is harmless and doesn't affect validation correctness.

I'm away from my computer at the moment, but I checked what the behaviour would be (on jsonschema.dev) if we skipped adding null to the type array for this case, and IINM doing so causes a null value to fail validation, because null is not of type object - so I think specifying null in both places is required