Nice kata. Enjoyed it.

Fixed in latest fork

Approved by someone

Approved

python new test framework is required. updated in this fork

Fork addressing this issue here:
https://www.codewars.com/kumite/6503a985c1845360d6bff45b

Would appreciate review; the new order-agnostic equality check isn't efficient but it's more than fast enough for this kata.

Confirming this is an issue in the Java version; will fork and rectify.

'refb l miq hj ko n g p a c d '
This test and some other tests fail because of the ORDER of the elements in the resulting array. But according to the description of the kata the order should not be taken into account!
The fail message is that the test expects the first element to be 'g', but my code makes it a 'c', which is correct too and it's just the second element in the tests, not the first like I have it.

PS: To make the tests order-insensitive you could use this code instead:
assertThat(LoneChar.loneliest("refb l miq hj ko n g p a c d ")).containsExactlyInAnyOrder('g', 'c');

the string can have leading/trailing spaces - you should not count them

And it has been already asked and answered here

how the output will be "g"
assert.deepEqual(loneliest('abcXdXXXefXXgXXXhXiXjXXXXXX'), ['g'], "result");
g = XXgXXX = 5
j = XjXXXXXX=7
Answer should be J instead of g

Java translation

Really funny and challenging kata, compliments.

Not a native in ruby but made an attempt to update test fixture. It works with MRI 3.0 now, let me know if there are any more changes that need to be done

A bit too easy for a 6.

Printing them ?