from functools import cache

with open('input.txt', 'r') as file:
    data: list[str] = file.readlines()

parts: list[str] = data[0].strip().split(', ')

starting_letter: dict[str, str] = {}

for part in parts:
    first_char = part[0]

    if first_char not in starting_letter:
        starting_letter[first_char] = [part]
    else:
        starting_letter[first_char].append(part)

designs = [design.strip() for design in data[2:]]

# this solution is pretty slow and stupid but it works, i think some variation of KMP would be faster? like a graph based search with a large amount of pre-processing
# Could maybe use memoization on substrings
@cache
def search_designs(design: str) -> bool:
    if not design:
        return True

    first_char = design[0]

    if first_char not in starting_letter:
        return False

    possible_matches = starting_letter[first_char]

    possible = []
    for possible_match in possible_matches:
        if possible_match == design[:len(possible_match)]:
            possible.append(design.replace(possible_match, ''))

    return any([search_designs(m) for m in possible])


possible_count = 0
for i, design in enumerate(designs):
    if search_designs(design):
        print(f'Design {i} finished search')
        possible_count += 1

print(f'Possible Designs: {possible_count}')