clean, but the intital split/filter repetition seems a bit high in performance cost and very reduceable.

It seems that everybody solving this question gets stuck on the ordering after getting 90% done with the kata (even me). the sort order is sorted lastly alphabetically.

If you're at that point; take a step back and clear your head and then you'll get it.

Good Kata.

Nevermind. I''m going to try it one more time and come back to this question.

Sorry. I forgot the code point reference. That solution isn't viable for the tests after. (updated in original question.)

but 'y' isn't before 'h' or 'r' Alphabetically.

And a later test mixes '1' '2' and '=' as meaningless to the order.

{"Are they here", "yes, they are here"} should return {"2:eeeee/2:yy/=:hh/=:rr"}

I think there is a pattern here that I'm not fully aware of because "Lexographic Order" means "Alphabetical"; but I assume you mean "In the order it appears in the strings."

However; 'h' and 'r' both appear before 'y'; if I assume that I'm ordering based on the reverse of one of the strings. AND 'r' appears before 'h' if I assume Min(appearance_order).

Is this puzzle saying that we should order the output based on the forward order of the second string AND THEN the forward order of the first string? If so "Reverse Lexographic Order" is slightly misleading.

(Edits are fixing grammar, typos, and word choices)

Edit: A later test mixes 1,2 and = prefixes in an uncertain order; which means that there is an order of operations assumed in the final solution that isn't offered in the problem statement itself. If you're assuming a different type of ordering based on which part of the process you're in; it should be clarified.

This assumption is made by "Code Point Order" being said after "Reverse Lexigraphic Order".

Which means you're assuming that there is a particular design pattern that should be used here; despite not referencing it or allowing for the users to solve the problem as written.

in JavaScript: test doesn't check if is formatted to two decimal places, despite it stating the requirement in the description.

i dont understand whats going here...

your elegant solution makes me feel lacking.

after finishing, I felt dumb with the polynomial i divised...

response to old question on performance:

my understanding is because the divisions are all repetition and the Math.Log is all repetition.

combined altogether, youre doing approximately more loops than what you think jt should be doing.

Ah. I found the Node version selector.

I meant; returning the original 'cells' input throws the syntax error whole halting the test.

JS throws error from test cases.