Kata

python new test framework is required. updated in this fork

The output in the example of the description is bad, no asterisk appears in it, due to bad markdown formatting ([['a', 'a', 'a', 'a', 'a', 'a', 'a', 'a'], ['a', '', '-', '-', '', 'a', 'a', 'a'], ['a', '|', 'a', 'a', '|', 'a', 'a', 'a'], ['b', '', '-', '-', '', 'b', 'b', 'b'], ['b', 'b', 'b', 'b', 'b', 'b', 'b', 'b']] should be [['a', 'a', 'a', 'a', 'a', 'a', 'a', 'a'], ['a', '*', '-', '-', '*', 'a', 'a', 'a'], ['a', '|', 'a', 'a', '|', 'a', 'a', 'a'], ['b', '*', '-', '-', '*', 'b', 'b', 'b'], ['b', 'b', 'b', 'b', 'b', 'b', 'b', 'b']].

haskell : https://www.codewars.com/kumite/5be897312a37b2b47c000a84?sel=5be897312a37b2b47c000a84

The when block is to check for a weird / rare edge case that your ranges in the randoms will sometimes generate ~4-5 times out of a thousand in haskell with the specifications. If you uncomment the when block you can see it.

I'm trying this on ruby and can't seem to figure it out. It's rated 7 so I'm guessing I'm missing something fairly simple?