Fun kata. Null byte edge cases were somewhat finicky, but fulfilling to iron out. It'd be fun to set up a fully streaming/piped decoder of this too.

Add tests and expand description.

Can you explain the (r - 4) part of this? The solution is extremely elegant, but I don't understand why it works. I understand why division by 2 is necessary, but why subtraction in increments of 4?

Oh, I agree with you. It was useful to have to figure out how to do it in Python 2. Makes me appreciate the innate unicode support of Python 3.

Try decoding/encoding the string/your results.

I would make the "it" clear in the failure of the .isEmpty() method test. "CWStack() class should return none if .itsEmpty() is True". Also, True and False should be capped in the instructions.

Thanks. I finally figured out how to make it work, though I think I went about it the wrong way.

Edit: Yea, I did it completely the wrong way. I'll remember that next time!

Can you remove trailing spaces from the tests with multiple words? They're contentless and not part of the kata, but are preventing my algorithm from being confirmed. For instance (I'm creating this example): "bovine-malarky; " is being evaluated for as "b4e-m5y; " so "b4e-m5y;" fails. There's no reason to preserve the trailing space (and the problem doesn't say that you should).

Ditto. Has anyone finished this with Python?

May be worth making it explicit that the parameters will always be ints (or passing strings to see if people are checking for type).

Add a colon at the end of the first line.

I figured out what the issue was. I was thinking in terms of .search and not .match. Thanks tho!