find_names_brute.py 文件源码

def setup(self, config):
    """
    Load name model (word list) and compile regexes for stop characters.

    :param config: Configuration object.
    :type config: ``dict``
    """
    reference_model = os.path.join(
        config[helper.CODE_ROOT], config[helper.NAME_MODEL])

    self.stopper = regex.compile(('(%s)' % '|'.join([
        'and', 'or', 'og', 'eller', r'\?', '&', '<', '>', '@', ':', ';', '/',
        r'\(', r'\)', 'i', 'of', 'from', 'to', r'\n', '!'])),
        regex.I | regex.MULTILINE)

    self.semistop = regex.compile(
        ('(%s)' % '|'.join([','])), regex.I | regex.MULTILINE)
    self.size_probability = [0.000, 0.000, 0.435, 0.489, 0.472, 0.004, 0.000]
    self.threshold = 0.25
    self.candidates = defaultdict(int)

    with gzip.open(reference_model, 'rb') as inp:
      self.model = json.loads(inp.read().decode('utf-8'))

    self.tokenizer = regex.compile(r'\w{2,20}')