Add tests that assert evaluator no longer returns whitespace outputs

It seems like when the executor is called to return an output in 'three buffered operations; the second one produces pure whitespace as output' (paraphrased from Cory's observation/comments). In an attempt to reproduce this issue/behavior, I was unable to do so in our Sandbox env so I added tests here; both for Py and JS. For either langs, based on test results, this does not seem to occur, or no longer exist. That being said, I am not sure I am provoking this bug the appropriate way for either langs.

Bug: T322345