Fix for https:/DominoKit/domino-jackson/issues/68

[tobias-hammerschmidt]

This is a proposal for fixing #68. I'm using this within my project and it allows a proper roundtrip of NaN/Infinity values from frontend to backend and back to the frontend.

[vegegoku]

Thank you for the PR
Would you please include some test cases for this.

[tobias-hammerschmidt]

I just added some tests for NaN/Infinity (de)serialization of double values. The client side test for deserializing "NaN" fails though and I haven't sorted out how to fix this. I'm using deserialization of NaN values in my project which is transpiled using j2cl-maven-plugin and haven't observed any issue. As long as the NaN handling is within the transpiled javascript code everything seems to be working.

As a side note maybe it makes sense to also fix the handling of NaN/Infinity (de)serialization for float values?

[vegegoku]

I just added some tests for NaN/Infinity (de)serialization of double values. The client side test for deserializing "NaN" fails though and I haven't sorted out how to fix this. I'm using deserialization of NaN values in my project which is transpiled using j2cl-maven-plugin and haven't observed any issue. As long as the NaN handling is within the transpiled javascript code everything seems to be working.

As a side note maybe it makes sense to also fix the handling of NaN/Infinity (de)serialization for float values?

I am not sure about the error you had, but I un-commented your assertion and did a build with tests and it passed.

I have one note thought, Can you please make the PR against the development branch rather than the master branch? this might require you to rebase your own fork/branch.

[tobias-hammerschmidt]

Ok I'll uncomment the single assertion in question, close this PR and create a new one which targets the development branch.