I'm passing by value here. So the data is copied into a new string and returned. This may be a best practice if you don't want to modify the original string. I would normally take a value like this by reference or use RVO but this looked like a fairly clean implementation. For its use case, passing by value is probably enough here.

Both string and algorithm are included in the test cases already. So we have access to them via the code that runs our code.

Can never overuse templates. 😉😂

?? 😅

I seem to be having issues with random tests for c++.

Expected: equal to xgay iZay !
Actual: xgay iZay!

^ As you can see the expected output has more whitespaces between the words than it should.

Edit: It seems this comment section actually clears up the extra whitespaces so the example doesn't look incorrect. But I can assure you extra whitespace is what showed up in the output.