class Scheduler(object):
"""Redis-based scheduler"""
def __init__(self, redis, persist, queue_key):
self.server = redis
self.persist = persist
self.queue_key = queue_key
# in-memory queue
self.own_queue = []
def __len__(self):
return len(self.queue)
@classmethod
def from_settings(cls, settings):
redisConfig = settings.get('REDIS_CONFIG', REDIS_CONFIG)
persist = settings.get('SCHEDULER_PERSIST', SCHEDULER_PERSIST)
queue_key = settings.get('SCHEDULER_QUEUE_KEY', QUEUE_KEY)
server = redis.Redis(redisConfig['host'], redisConfig['port'])
return cls(server, persist, queue_key)
@classmethod
def from_crawler(cls, crawler):
settings = crawler.settings
return cls.from_settings(settings)
def open(self, spider):
self.spider = spider
self.queue = SpiderQueue(self.server, spider, self.queue_key)
self.df = RFPDupeFilter(self.server, DUPEFILTER_KEY % {'spider': spider.name})
# notice if there are requests already in the queue
if not self.persist:
self.df.clear()
self.queue.clear()
if len(self.queue):
spider.log("Resuming crawl (%d requests scheduled)" % len(self.queue))
def close(self, reason):
pass
def enqueue_request(self, request):
if not request.dont_filter and self.df.request_seen(request):
return
if self.spider.logined:
self.queue.push(request)
else:
self.own_queue.append(request)
def next_request(self):
if self.spider.logined:
return self.queue.pop()
if len(self.own_queue) > 0:
return self.own_queue.pop()
def has_pending_requests(self):
if self.spider.logined:
return len(self) > 0
return len(self.own_queue)
def open(self, spider):
self.spider = spider
self.queue = SpiderQueue(self.server, spider, self.queue_key)
self.df = RFPDupeFilter(self.server, DUPEFILTER_KEY % {'spider': spider.name})
# notice if there are requests already in the queue
if not self.persist:
self.df.clear()
self.queue.clear()
if len(self.queue):
spider.log("Resuming crawl (%d requests scheduled)" % len(self.queue))
