parser.py

import os
import re
import sys
import unidecode

default_regex = re.compile(r"\n[\t ]*([A-Z0-9'\- ]+[A-Z0-9'\-])(?: \(.+\))?(?: \[.+\])?\n(([\t ]+(?!\(|\[|\s))?(?(3).*(?:\n\3[\t ]*\S+.*)*|[\t ]+\S+.*(?:\n[\t ]+\S+.*)*))")

def parse_script(filename, directory='scripts', regex=default_regex):
    with open(os.path.join(directory, filename), encoding='ISO-8859-1') as content_file:
        matches = re.finditer(regex, unidecode.unidecode(re.sub(u'u\0092', '\'', content_file.read())))
        lines = []
        for m in matches:
          name = m.group(1)
          line = re.sub('\s+', ' ', re.sub('\(.+?\)|\[.+?\]', '', m.group(2), 0, re.DOTALL)).strip()
          if len(line):
            lines.append((name, line))
        return lines

def line_to_words(line):
  
  return re.findall(r'\'?[\w]+(?:[\'\u2019-][\w]+)*\'?', line)