Skip RS CTRL-CHAR to support JSON Text Sequence (RFC7464)

[iifawzi]

Hello, This PR tries to resolve #633

This is my first PR for the project, so I could be missing a point. Feel free to guide me through points that might need to be added.

I didn't implement it in the data input parser as it seems like we're skipping the validation of WS

jackson-core/src/main/java/com/fasterxml/jackson/core/json/UTF8DataInputJsonParser.java

Lines 2299 to 2304 in 99f749d

	} else {
	// 06-May-2016, tatu: Could verify validity of WS, but for now why bother.
	// ... but line number is useful thingy
	if (i == INT_CR || i == INT_LF) {
	++_currInputRow;
	}

so not sure if we need to have it now, the same way we do for other parsers

[cowtowncoder]

May look funny but JsonParser.Feature is to be removed from 3.0 yet is needed internally (all StreamReadFeatures and JsonReadFeatures must map back to one entry).
But we don't want users to use JsonParser.Feature any more.

[iifawzi]

It's good to know about that to help take care of it next time.
Thanks for your collaboration on the PRs!

[cowtowncoder]

@iifawzi Implementation looks good; refactored things slightly wrt tests but functionally it was fine.
Did not remember that DataInput-backed parser skips WS validation; agree that we'll go with it for now.

The only concern I have is merging to 3.0/master -- that will be some fun. But I'll take care of it.

Thank you again for contributing this!

EDIT: merging wasn't too bad after all. :)

[iifawzi]

@iifawzi Implementation looks good; refactored things slightly wrt tests but functionally it was fine. Did not remember that DataInput-backed parser skips WS validation; agree that we'll go with it for now.

The only concern I have is merging to 3.0/master -- that will be some fun. But I'll take care of it.

Thank you again for contributing this!

EDIT: merging wasn't too bad after all. :)

Thanks for your collaboration @cowtowncoder, it helps and encourages contributing to the project.
Do you think the next move for this is to allow it in Content as we do for tabs? Otherwise, I might try to take a look on 1144

[cowtowncoder]

@iifawzi thank you! It is great to hear that! And thank you for your contributions so far.

On RS in content: unless there's strong use case for allowing it (that is, some actually needing it for specific use), I'd probably not want to add support for in-content. #1144 would be good to try to figure out.