Closed cowtowncoder closed 2 years ago
Ugh. Should have read #151 PR first, it was NOT a test fix but actual change; and change does not make sense from performance perspective -- it would make match more reliable but very expensive. Ideally we'd have perf test/benchmark to prove this but from what I see the implementation is quite inefficient and negates point of caching (esp. when key type is used for HashMaps...)
Reverts FasterXML/jackson-jaxrs-providers#151