Approved

C fork

• adding tests for both lowercase and uppercase letters
• adding tests for negative numbers, since most languages test them
• making initial code compile
• better assertion messages

Fixed.

python new test framework

Approved

The answer is almost definitely "it's platform dependent" and "how much does the JIT like you?".

I can change the author solution to be the other one if people want though, I just think that the current one is cute.

For an 8kyu kata, I would prefer the single, simple, descriptively named API call.

Node 12+ has been enabled now

Ta!

Constraints: 1 ≤ a[i] ≤ 10,000

that does not seem to be true, as there is a fixed test [34_241, 23_434, 2_341] in all languages

please add the performance tag; the naive solution times out

Node 18 support

This can be not easy to estimate because there are two opposing mechanisms at play:

• the fold approach is elementary, simple operations, but possibly in bytecode, which needs to be executed on the VM,
• the string approach is a potentially complex operation (strings mean dealing with languages, locales, formats, etc.), but possibly executed as a native code.

So you have either a very basic logic but possibly in a higer-level layer of execution, versus a more complex mechanism but possibly closer to the metal. And as it usually comes to performance tuning, it's difficult to tell exactly without measuring.

Scala translation

Merged.