feature/RowProcessor-replaceNewlinesWithSpaces

[nezda]

Description

Add optional boolean flag RowProcessor.replaceNewlinesWithSpaces that controls whether or not to replace newlines with spaces in values before passing them to field processors (don't change default behavior by defaulting this to true).

Motivation

TextFieldProcessors may wish to be aware of newlines in the input and encode them as features. Example encoded as a unit test.

[JackSullivan]

This looks good to me. I originally wrote this, and revisiting it now, I'm not convinced that the default behavior is correct. That is, we should not be altering the incoming data in RowProcessor before it is fed to the feature processors. In a future version we'll probably want to change that, but to avoid breaking changes this is probably the best interim solution. The constructor situation on RowProcessor is also in great need of harmonization, probably through consolidation into a builder pattern.

Thanks for putting in the work to put it in, and for bringing this code to our attention again.

[Craigacp]

Thanks. As Jack said, we probably want to revisit this as part of a larger refactor to make RowProcessor a little simpler to use, but in the meantime this seems like a fine change.