19.2.py

import re
with open('input/19.txt') as f:
  content = f.read().split("\n\n")
  rules = content[0].split("\n")
  messages = content[1].split("\n")

def make_regex(rules, chars, i, depth=0):
  # base rule
  if i in chars:
    return chars[i]

  # lol
  if depth == 14:
    return ''

  terms = []
  # make a regex with ORs
  for term in rules[i]:
    terms.append(''.join(make_regex(rules, chars, t, depth+1) for t in term))
  return '(' + '|'.join(terms) + ')'

def parse_into_regex(rules):
  parsed_rules = {}
  characters = {}

  # go through all rules and build up lists of terms
  for r in rules:
    i, rule = r.split(': ')
    i = int(i)

    # convert new rules
    if i == 8:
      rule = '42 | 42 8'
    elif i == 11:
      rule = '42 31 | 42 11 31'

    # save our direct characters
    if '"' in rule:
      characters[i] = rule[1]
      continue

    # get the terms grouped on either side of the OR
    for side in rule.split('|'):
      terms = []
      for term in side.strip().split(' '):
        terms.append(int(term))
      if i not in parsed_rules:
        parsed_rules[i] = []
      parsed_rules[i].append(terms)

  # print(parsed_rules, characters)

  return make_regex(parsed_rules, characters, 0)

regex = parse_into_regex(rules)
print(len([match for match in messages if re.match('^' + regex + '$', match)]))