utils.py 文件源码

def _normalize_answer(s):
    """Normalize string to score answers according to SQuAD dataset scoring rules.

    Remove articles, remove punctuation, fix multiple whitespaces in string, and convert all characters to lowercase.
    """

    def remove_articles(text):
        return re.sub(r'\b(a|an|the)\b', ' ', text)

    def white_space_fix(text):
        return ' '.join(text.split())

    def remove_punc(text):
        exclude = set(string.punctuation)
        return ''.join(ch for ch in text if ch not in exclude)

    def lower(text):
        return text.lower()

    return white_space_fix(remove_articles(remove_punc(lower(s))))