Add some fixed test to sample tests too, where there are at least two longest strings to all languages. See the post below.

The tests are fine, your code is not. It fails tests like this one:

assert.strictEqual(longestConsec(["zone", "abigail", "theta", "abigail", "form", "libe", "zas"], 2), "abigailtheta")

It doesn't return the first one.

I don't understand why [0] doesn't return an optional that needs to be unwrapped?

Isn't [0] calling subscript(_:) (see https://developer.apple.com/documentation/swift/dictionary/subscript(_:)-8rfql

I think there might be something wrong with this kata on JavaScript. My code is passing the basic tests. While my code fails the attempt test on codewars, I put the exact same input into repljs.com (with the same random strarr and k that the test is testing) and it passes on repljs.com.

This has to be satire. Right?

Yeah I think that should still be validated. 0 is a valid value for UInt32 so it's still a potential bug. Negative numbers don't need to be checked though since they are not valid UInt32 values.

Count the number of divisors of a positive integer n.

A positive integer is an integer greater than 0.

But if you want to be sure, then yes you could validate the input too. Not needed here tho.

If anyone accidentally calls divisors(0) then this will crash. There needs to be a check like:

guard n != 0 else { return 0 }

However I'm not sure what you should return. Returning 0 would technically be right. 0 has 0 divisors since 0 can't be divided.