These two clarifications were helpful for me:

1. For negative numbers you should find positive divisors.
2. Kata considers every number is its own divisor (ie. 107 is a divisor of 107)

It will be nice if those two notes could be added to description.

Happy Coding!!!

I thought I had the solution right but I’m getting this error:

'11, 11, 2(, 3\x1f, 5\r, 8\xf2, =\xc5, E}, R\x08, gK, '

should equal

'1, 1, 2, 3, 5, 8, 13, 21, 34, 55'

This test has ‘\n’ as input.

Other tests work fine as long as they don’t include characters with ‘\’.

I’m I using a wrong codification?
I used chr() and ord() to change ascii<->string.

I'm a little disappointed with this kata. Such a good idea but I didn't like the implementation. These are the problems I found:

1. The problem was not clearly explained.
2. The definition of rows and columns is reversed.
3. The test are not random, you can just capture input and learn output
4. There are no loops in the given maps (it will be interesting to see if the miner can avoid loops)
5. When you get a time out it doesn't provide any information about what test case went wrong (I believe this is a general problem in Codewars)
6. Difficulty should be 4 kyu
7. There is a test-case which is wrong, it gives a map fill with walls: _m = [[0], [0], [0], [1]] which should really be _m = [[1], [1], [1], [1]]

I have to say, nevertheless, it's been so much fun, that I don't care about any of that. The idea was really good, and I had a great time walking around the issues. Codewars rocks!!!