The problem is that you should first of all check to see if the sequence is invalid (=has invalid characters); one of the text cases deliberately give you something like '____' just to test that.

I am not the one who wrote the Python version, but basically any version works that way on purpose.

I hope that was clear enough and the issue can be considered fixed :)

I enjoyed this kata. My only advice would be to add a few more relevant categories so that it can be found easier ;)