예제 #1
0
파일: scheduler.py 프로젝트: thorwang/cye
    def open(self, spider):
        self.spider = spider
        self.queue_key = self.get_queue_key(spider)
        self.queue = SpiderQueue(self.redis_cli, spider, self.queue_key)
        self.df = RFPDupeFilter(self.redis_cli,
                                self.get_dupefilter_key(spider))
        if spider is not None:
            spider.log("Queue key of redis (%s)" % self.get_queue_key(spider),
                       log.INFO)
            spider.log(
                "Dupefilter key of redis (%s)" %
                self.get_dupefilter_key(spider), log.INFO)

        # notice if there are requests already in the queue
        if len(self.queue):
            spider.log("Resuming crawl (%d requests scheduled)" %
                       len(self.queue))