day19/monstermessages2.py

   1 #!/usr/bin/python3
   2
   3 import sys
   4 import re
   5
   6 filename="p2_12.txt"
   7 if len(sys.argv) > 1:
   8     filename=sys.argv[1]
   9
  10 in_rules=True
  11 rules={}
  12 data=[]
  13
  14 for line in open(filename, "r"):
  15     line = line.rstrip()
  16     if line == "":
  17         in_rules=False
  18         continue
  19
  20     if in_rules:
  21         (rule_number,rule) = line.split(":")
  22         rule_number=int(rule_number)
  23         rule=rule.strip()
  24         rules[rule_number] = [s.strip() for s in rule.split("|")]
  25     else:
  26         data.append(line)
  27
  28 def get_rule_regex(rulenumber):
  29     regex=""
  30     first=False
  31     for rule in rules[rulenumber]:
  32         if rule[0] != '"':
  33             # not a character, so we're going to go down the wishing well
  34             for number in rule.split(" "):
  35                 number=int(number)
  36                 regex+=get_rule_regex(number)
  37             regex+="|"
  38         else:
  39             regex+=rule[1:-1]
  40     if regex[-1] == "|":
  41         regex=regex[0:-1]
  42     if len(rules[rulenumber]) > 1 or len(rules[rulenumber][0]) > 3:
  43         regex="("+regex+")"
  44
  45     return regex
  46
  47 # because we're doing insane things, lets build the regex for rules[8] first, which is literally 42 multiple times
  48 #rules[8]=["42", "42 8"]
  49 #rules[11]=["42 31", "42 11 31"]
  50
  51 rules[8]=['"((' + get_rule_regex(42) + ')(?P<rule8>(' + get_rule_regex(42) + ')*))"']
  52 rules[11]=['"((' + get_rule_regex(42) + '))(?P<rule11>.*)((' + get_rule_regex(31) + '))"']
  53
  54 # ok - so all the rules are in, lets build a huge regex that gets all the rules
  55 pattern="^"+get_rule_regex(0)+"$"
  56
  57 match_count=0
  58
  59 for line in data:
  60     matches = re.match(pattern, line)
  61     if matches:
  62         if matches.group('rule11'):
  63             part=matches.group('rule8') + matches.group('rule11')
  64             rule_31_matches=0
  65             rule_42_matches=0
  66             while len(part) > 0:
  67                 # remove the end, and keep count, we need at least that many rule 42s to
  68                 # be removed from the front
  69                 remove_pattern=get_rule_regex(31)+"$"
  70                 new_part=re.sub(remove_pattern, '', part)
  71                 if len(new_part) == len(part):
  72                     # try removing a rule 42 from the start
  73                     remove_pattern="^" + get_rule_regex(42)
  74                     new_part=re.sub(remove_pattern, '', part)
  75                     if len(new_part) == len(part):
  76                         # if nothing got removed, and we're not 0 length, we're not a match
  77                         break
  78                     else:
  79                         part=new_part
  80                         rule_42_matches+=1
  81                 else:
  82                     rule_31_matches+=1
  83                     part=new_part
  84             else:
  85                 if rule_42_matches >= rule_31_matches:
  86                     match_count+=1
  87         else:
  88             match_count+=1
  89
  90 print("There were {} matches in total".format(match_count))