not enough random tests.

your solution is incorrect:

shelve_trophies(2, 6, ['5.1.4', '4.#.1', '1.x.2', '1.1.3', '0.1.2', '4.1.5', '5.1.2'])
expected 2 but was 1

• 4.#.1 => 4 is on shelf 0
• 4.1.5 or 5.1.4 => 5 is on 1
• 5.1.2 => 2 is on 0
• 0.1.2 => 0 is on 1
• 1.x.2 => 1 is on 1
• 1.1.3 => 3 is on 0

all trophies positions are perfectly determined and there is only one solution.

your solution isn't robust enough:

on these arguments: 5, 4, ['3.#.4', '3.1.0', '2.#.0', '1.x.2', '3.x.2', '1.#.3', '1.1.2'] 

STDERR
Traceback (most recent call last):
  File "main.py", line 95, in <module>
    test.assert_equals(n_shelves, answer(shelves, trophies, rules))
  File "main.py", line 52, in answer
    [1])]
IndexError: list index out of range

Hi,

Sounds interesting. But you forgot to give some information:

• trophies seems to actually be 0-indexed.
• what about the shelves? I guess it's the same?
• should all shelves be used at least once or not?
• max number of trophies per shelf?

Aside of that:

• you need more fixed tests (tests that are actually testing specific edge cases)
• you need more random tests (and call them that way, not just "real tests". Those tests clearly aren't reality related... ;) )