python new test framework is required. updated in this fork

Haskell: the reference solution expects Right [] for n = 2.

I think this Kata should be a Kyu 6 or maybe Kyu 5. In comparison to other Kyu 7 Katas it is costly.

Also, if given a floating point number, return the list of pernicious numbers with the number floored (rounded down).

This is an idiotic requirement that is also not fulfilable because float range is different from int range. A number bigger than maximum safe ingeter will cause tons of problems.

Haskell : https://www.codewars.com/kumite/5bf9e1a83efceecbbd00014d?sel=5bf9e1a83efceecbbd00014d

Note this is a bit more involved in Haskell because :

• no builtin toBinary
• no builtin isPrime
• you have to rollUp polymorphic functions (float or int arguments)

Because of this, I did a lot of #3 in the solution setup, I can remove it all if you want, but that really pushes it into a deep yellow for Haskell.

In my opinion this is at least 6th lvl kata. It require A LOT of knowledge and practice.

That is a great kata and I hope being able to solve it in a more glorious way one day (that is to say that my solution is to be improved). Thanks for the riddle.

In the description, the binary representation of 25 is not 11010, it's 11001. I think.

C#-Translation kumited!

https://www.codewars.com/kumite/57aa542053ba33a831001879

Please check and approve it!

(I approved the kata for you! :-))

Something wrong with the test cases in the python version, because I can complete the kata without handling floats (see my solution).

Also, add checking of numbers up to at least 32-64 bits...

Nice kata. :-) Thanx.