tornado_fetcher.py 文件源码

def can_fetch(self, user_agent, url):
        parsed = urlsplit(url)
        domain = parsed.netloc
        if domain in self.robots_txt_cache:
            robot_txt = self.robots_txt_cache[domain]
            if time.time() - robot_txt.mtime() > self.robot_txt_age:
                robot_txt = None
        else:
            robot_txt = None

        if robot_txt is None:
            robot_txt = RobotFileParser()
            try:
                response = yield gen.maybe_future(self.http_client.fetch(
                    urljoin(url, '/robots.txt'), connect_timeout=10, request_timeout=30))
                content = response.body
            except tornado.httpclient.HTTPError as e:
                logger.error('load robots.txt from %s error: %r', domain, e)
                content = ''

            try:
                content = content.decode('utf8', 'ignore')
            except UnicodeDecodeError:
                content = ''

            robot_txt.parse(content.splitlines())
            self.robots_txt_cache[domain] = robot_txt

        raise gen.Return(robot_txt.can_fetch(user_agent, url))