Performance is not only about short code or long code. You need to think about the time and space complexity. I'm having the same issue while it's an 0(n). I'm trying to find the best way to do this

Groovy Translation

Rust version lacks random tests

Your solution has a bug and it will not work in pycharm if you run it more than once.

Bugs in your solution are not a kata issue.

great!

It would be nice to show the input when the test fails.

If you use lodash, it can be _.sample(charset) and no problems with indexing :)

EDIT: fixed

Write a function that takes in a string and replaces all the letters with their respective positions in the English alphabet

There is nothing stated about whether or not the input should only be English letters + spaces + ..., so any character can be used in test cases. And using .indexOf(x) with padding instead of .indexOf(x) + 1 in this case is incorrect, so that should be tested as well.

It's fixed for the underscore character, but the general issue still stays that a filler character can be used and won't be detected, because detecting it would require testing of the whole character space. I would say that using a character outside of the alphabet used by the problem as padding is not incorrect. What do you think?

fixed (see JS fork above)

Done

Merged

JS fork

• updated to Node 18
• replaced legacy testing framework with Mocha/Chai
• resolved this issue

OP solved it, closing