You must not have read the comments there. I specifically did not spoiler some comments about it, because I thought people should know ( I ran into it myself until somebody pointed it out ).

The test quality is poor; it's 50 random tests of n < 1e6 along with n = 0, 1, 1e8. The original code can pass the tests once it returns the correct answer for n = 1e8 without trying to compute it.

But then bumping the input range of random tests will also cause timeouts. Not even the faster solution finishes all the tests if the 50 random tests are bumped to n < 1e7.

10e6 and 10**6 ( and same for 8 ) are not the same thing.

retired

retired

return an array of [giver, receiver]

edge cases
Input: 2 - too few employees for this fun: expected [ [ 2, 1 ], [ 1, 2 ] ] to equal null

null is not an array. The description should be correct and consistent.

edge cases
Input: 2 - too few employees for this fun: expected [ [ 2, 1 ], [ 1, 2 ] ] to equal null

This is unspecified.

it no longer passes

If all issues are resolved, you can republish the kata.

My (unfinished) code passed without me considering that while writing it.

Kindly check again @rowcased

Have added a few valid swaps for sample tests

I have updated the tests as per the docs. Please give them another look. Have also updated the description to capture no direct exhange of gifts.

There should be at least one sample test with a valid swap.