There are no random tests.

If no such word exists, return "No word found".

For error conditions, returning a magic value that might be a valid return value in non-error conditions is not a good idea.

Throw an error, or maybe have solver return "" instead ( I haven't solved, and I have not read the description closely to see if "" really makes sense ).

If we follow the best practice, a function should always return the same type. This will make it easy to translate the problem to languages such as C++ and Java. I would also encourage authors to use type hints in the Solution Setup code.

from typing import Tuple, List


def help_jack(protons: int, neutrons: int) -> Tuple[int, int, List[int], int]:
    # code here

or if we allow the function to return strings too, then:

from typing import Any


def help_jack(protons: int, neutrons: int) -> Any:
    # code here

• don't round (see issue below)
• put a dict of atomic masses in preloaded
• use all elements
• use numbers > 9 (most solutions already handle large numbers)

This isn't really true, is it?

Sort the substrings first by length, then alphabetical order

• No Random Tests
• Kata Concept Lacks Novelty
• Possible Duplicate