Tatu Saloranta

Yes, that sounds good in general. One practical consideration: I suspect most of handling needs to be in `jackson-databind`, specifically in `TypeFactory`. This type is only declared here in `jackson-core`...

Maybe we can experiment with #1? A little bit of duplication is fine, although I guess it is easier to comment when seeing more concrete case. Changes to `TypeFactory` might...

Hmmh. First of all, thank you for reporting this issue. Not quite sure how to feel about this as it requires two coercions -- from String to Float; from Float...

One option is to call `DeserializationContext.getParser()` which should always return non-null parser instance. In case of buffering (`TokenBuffer`) it seems, however, that feature flags are not copied from the "real"...

Not sure there's much need to support fast parser for `TextNode` tho, come to think of it -- as long as `DoubleNode` supports it. Or is it common to serialize...

Right, I understand the route. Was more curious as to whether it is used by anyone -- it does explicit coercion from JSON String to `double`. But I guess you...

Since 2.13 has been released marking as 2.14 (earliest) due to need for API additions and possible compatibility concerns for some maximum limits. Also: I think this issue is still...

Some thinking out aloud: * Limits for total input document size can probably be checked at points when input blocks are read (blocking) or fed (async): this is what Woodstox...

Hmmh. Interesting. I can see how/why property introspector gets confused: the problem comes from multiple factors... 1. Implied name (ONLY found if parameter names module registered) for 2nd argument is...

I will have to say that it is quite possible that combination of polymorphic type in `Optional` for root value may be unsupportable. I would strongly recommend not using that...