Add DelegatingSerializer

[mluckam]

Currently there is a DelegatingDeserializer, but no DelegatingSerializer. Was there a reason for not including a DelegatingSerializer? It would be nice to have a DelegatingSerializer for symmetry.

I am currently using the DelegatingDeserializer to add validation to the deserialize process, see example project. Since there is not a DelegatingSerializer, the same strategy cannot be applied to serialization. This is just one use case, but I figure others may also utilize deserialization strategies with the DelegatingDeserializer that could also be applied to serialization if there was a DelegatingSerializer.

Welcome any thoughts regarding the request.

[cowtowncoder]

Did you see StdDelegatingSerializer.java? Would that not work?

I am not sure why DelegatingDeserializer.java was NOT renamed as StdDelegatingDeserializer.java in 3.0 (would have made sense), but looks like 2.x -> 3.x there was some refactoring as 2.x had both DelegatingDeserializer.java and StdDelegatingDeserializer.java, of which latter was actually renamed as StdConvertingDeserializer.java.

EDIT: looks like StdDelegatingSerializer.java is a "converting" one, assuming existence of Converter, and what you want is "plain" delegating serializer.
And what'd be really good would be:

• Addition of "plain" DelegatingSerializer like you propose,

but also for 3.1:

• Deprecating StdDelegatingSerializer (as name), create StdDelegatingSerializer in its place -- effectively renaming this implementation, make old one extend new one (or vice versa).

to address naming inconsistency.

[github-actions]

🧪 Code Coverage Report

Metric	Coverage	Change
Instructions	80.17%	📉 -0.040%
Branches	73.24%	📉 -0.040%

Coverage data generated from JaCoCo test results

[cowtowncoder]

Looks good, just need to figure out how exactly to untangle naming of "delegating vs converting" (de)serializers.

I think plain Delegating[De]Serializer vs StdConvertin[De]Serializer is the goal, just need to keep 3.1 compatible with 3.0 -- 3.1 will be strict baseline for 3.x, but I don't want gratuitous breakages between 3.0 and 3.1 (but deprecating 3.0 types is fine).

[cowtowncoder]

Quick note: I renamed existing class in #5631 / #5632 so there's nothing this PR needs to do wrt naming -- DelegatingSerializer is fine as-is.

[github-actions]

🧪 Code Coverage Report

Metric	Coverage	Change
Instructions	80.18%	📉 -0.030%
Branches	73.26%	📉 -0.020%

Coverage data generated from JaCoCo test results

[github-actions]

🧪 Code Coverage Report

Metric	Coverage	Change
Instructions	80.23%	📉 -0.030%
Branches	73.33%	📈 +0.000%

Coverage data generated from JaCoCo test results

[github-actions]

🧪 Code Coverage Report

Metric	Coverage	Change
Instructions	80.36%	📉 -0.040%
Branches	73.55%	📉 -0.020%

Coverage data generated from JaCoCo test results

[cowtowncoder]

@mluckam Ok looks good, happy to add! Just one more thing before doing so: unless we've gotten one from your earlier (CLAs are good after submitting one), we'd need CLA.

It's from here: https://github.com/FasterXML/jackson/blob/main/contributor-agreement.pdf
and the usual way is to print, fill & sign, scan (or modify PDF for the same), email to cla at fasterxml dot com.
Once I got it I can merge this in 3.x!

Thank you again for contributing this.

[mluckam]

CLA signed previously as part of #4483

[cowtowncoder]

CLA signed previously as part of #4483

Ah. Yes, I see it now, for M. Luckam. Thanks!

[github-actions]

🧪 Code Coverage Report

Metric	Coverage	Change
Instructions	80.37%	📉 -0.030%
Branches	73.55%	📉 -0.020%

Coverage data generated from JaCoCo test results

[github-actions]

🧪 Code Coverage Report

Metric	Coverage	Change
Instructions	80.38%	📉 -0.020%
Branches	73.56%	📉 -0.010%

Coverage data generated from JaCoCo test results

[github-actions]

🧪 Code Coverage Report

Metric	Coverage	Change
Instructions	80.39%	📉 -0.010%
Branches	73.55%	📉 -0.020%

Coverage data generated from JaCoCo test results

[github-actions]

🧪 Code Coverage Report

Metric	Coverage	Change
Instructions	80.41%	📈 +0.010%
Branches	73.57%	📈 +0.000%

Coverage data generated from JaCoCo test results

[cowtowncoder]

Added some more testing coverage; fixed issues with "withXxx()" methods, merged in for 3.1.0!