wiki_tags.py 文件源码

def get_content_snippet(content, keyword, max_words=30):
    """
    Takes some text. Removes html tags and newlines from it.
    If keyword in this text - returns a short text snippet
    with keyword wrapped into strong tag and max_words // 2 before and after it.
    If no keyword - return text[:max_words].
    """

    def clean_text(content):
        """
        Removes tags, newlines and spaces from content.
        Return array of words.
        """

        # remove html tags
        content = striptags(content)
        # remove newlines
        content = content.replace("\n", " ").split(" ")

        return list(filter(lambda x: x != "", content))

    max_words = int(max_words)

    pattern = re.compile(
        r'(?P<before>.*)%s(?P<after>.*)' % re.escape(keyword),
        re.MULTILINE | re.IGNORECASE | re.DOTALL
    )

    match = pattern.search(content)

    if match:
        words = clean_text(match.group("before"))
        before_words = words[-max_words // 2:]
        words = clean_text(match.group("after"))

        after = " ".join(words[:max_words - len(before_words)])
        before = " ".join(before_words)

        html = "%s %s %s" % (before, striptags(keyword), after)

        kw_p = re.compile(r'(%s)' % keyword, re.IGNORECASE)
        html = kw_p.sub(r"<strong>\1</strong>", html)

        return mark_safe(html)

    return " ".join(clean_text(content)[:max_words])