Jonas Konrad

Records are not just treated as normal classes by jackson. e.g. the getter names are different, and property names for the constructor are known. If you want the same behavior...

I can't answer authoritatively whether this should be a feature or anything – I don't work on jackson-databind, I just try to help with the issue reports a bit. `constructorDetector`...

@soc the documentation simply means that this is the default value for [the 'lenient' setting](https://fasterxml.github.io/jackson-annotations/javadoc/2.13/com/fasterxml/jackson/annotation/JsonFormat.html#lenient--)

@cowtowncoder I would think it would have an impact on jackson, because `@Inherited` affects the behavior of the normal reflection getAnnotation

Ah that makes sense. There are definitely some micronaut bugs in this area also, since we introspect the annotations ourselves too, but don't follow jacksons annotation inheritance behavior properly. It...

Can you give a test case that does not use spring? But my guess is that DefaultTyping.NON_FINAL is the issue – records are final.

i can reproduce your issue with the new test case, and yes it is because of `DefaultTyping.NON_FINAL`. Changing it to `DefaultTyping.EVERYTHING` resolves this. However it is somewhat problematic because it...

I mean, it has to be returned as *some* type, it can't be Long and Integer at the same time. Unless you configure [USE_LONG_FOR_INTS](https://fasterxml.github.io/jackson-databind/javadoc/2.6/com/fasterxml/jackson/databind/DeserializationFeature.html#USE_LONG_FOR_INTS), it will use Integer for small...

@cowtowncoder using methodhandle you can set fields / call methods using primitives without boxing. Not sure if it would make much difference here though.

Looks to be caused by the fix for #2644. For that fix, the TokenBuffer buffers the exact value because the necessary precision can't be determined beforehand (the target type might...