Thank you!, I should find time to keep doing more kata :), but is nice to know that you enjoy it

The Things is that only %$&/#·@|º\ª chars are considered noise. I will change the explanation to be more crear.

I have delete it... and I can't accept it again. Please, try to do it again.

Thank you and sorry for the inconvenience

ops... I don't know how I can find your new version :-S

Array.prototype.includes is a ES7 method an it runs ok on chrome and firefox, but it is not supported by the current version of babel running on this environment, that is why your code can't run as you expect.

So this is not a test case error and you must avoid Array.prototype.includes in your code in CodeWars. You can open an issue asking for include a polyfill in CW environment here.

CoffeeScript translation Kumited

Yeah, thats solve the issue... but I have a new issue for you. I'm going to post it. We keep talking :)

You are not testing if our implementations are using cache to avoid calculate count for already caculated word counts.

I think I have solved this issue :)

What do you think I should do?, warning about don't use the operator ... because babel is going to transpile it to a for loop, or make a regex whick allows only this kind of loops created by babel?

What happens is that when the code is transpiled to ES5 with babel, ...predicates is transpiled to

  for (var _len = arguments.length, predicates = Array(_len), _key = 0; _key < _len; _key++) {
    predicates[_key] = arguments[_key];
  }

So the code transpiled has a for loop... I have to see who I can avoid this exception.

Thank you :)

Hi, the thing is that if the input string is empty, or if there is no match between your regexp and the input string str.match(/YOUR REGEXP/g) is going to return null instead of an array, so you can't execute join("") as a method of null. I think that's all I can say :). Good Luck

I have changed tests and I have realised that your solution was wrong ;-)