def __init__(self, thread_count=None):
"""
基于内存队列的爬虫，不支持分布式
:param thread_count: 线程数
"""
super(AirSpider, self).__init__()
for key, value in self.__class__.__custom_setting__.items():
setattr(setting, key, value)
log.reload()
self._thread_count = (setting.SPIDER_THREAD_COUNT
if not thread_count else thread_count)
self._memory_db = MemoryDB()
self._parser_controls = []
def __init__(
self,
redis_key=None,
thread_count=None,
begin_callback=None,
end_callback=None,
delete_keys=(),
auto_stop_when_spider_done=None,
auto_start_requests=None,
send_run_time=True,
batch_interval=0,
wait_lock=True,
task_table=None,
):
"""
@summary: 调度器
---------
@param redis_key: 爬虫request及item存放reis中的文件夹
@param thread_count: 线程数，默认为配置文件中的线程数
@param begin_callback: 爬虫开始回调函数
@param end_callback: 爬虫结束回调函数
@param delete_keys: 爬虫启动时删除的key，类型: 元组/bool/string。 支持正则
@param auto_stop_when_spider_done: 爬虫抓取完毕后是否自动结束或等待任务，默认自动结束
@param auto_start_requests: 爬虫是否自动添加任务
@param send_run_time: 发送运行时间
@param batch_interval: 抓取时间间隔 默认为0 天为单位 多次启动时，只有当前时间与第一次抓取结束的时间间隔大于指定的时间间隔时，爬虫才启动
@param wait_lock: 下发任务时否等待锁，若不等待锁，可能会存在多进程同时在下发一样的任务，因此分布式环境下请将该值设置True
@param task_table: 任务表， 批次爬虫传递
---------
@result:
"""
super(Scheduler, self).__init__()
for key, value in self.__class__.__custom_setting__.items():
setattr(setting, key, value)
log.reload()
self._redis_key = redis_key or setting.REDIS_KEY
if not self._redis_key:
raise Exception(
"""
redis_key 为redis中存放request与item的目录。不能为空，
可在setting中配置，如 REDIS_KEY = 'test'
或spider初始化时传参, 如 TestSpider(redis_key='test')
"""
)
self._request_buffer = RequestBuffer(redis_key)
self._item_buffer = ItemBuffer(redis_key, task_table)
self._collector = Collector(redis_key)
self._parsers = []
self._parser_controls = []
self._parser_control_obj = PaserControl
self._auto_stop_when_spider_done = (
auto_stop_when_spider_done
if auto_stop_when_spider_done is not None
else setting.AUTO_STOP_WHEN_SPIDER_DONE
)
self._auto_start_requests = (
auto_start_requests
if auto_start_requests is not None
else setting.SPIDER_AUTO_START_REQUESTS
)
self._send_run_time = send_run_time
self._batch_interval = batch_interval
self._begin_callback = (
begin_callback
if begin_callback
else lambda: log.info("\n********** feapder begin **********")
)
self._end_callback = (
end_callback
if end_callback
else lambda: log.info("\n********** feapder end **********")
)
self._thread_count = (
setting.SPIDER_THREAD_COUNT if not thread_count else thread_count
)
self._spider_name = redis_key
self._project_name = redis_key.split(":")[0]
self._tab_spider_time = setting.TAB_SPIDER_TIME.format(redis_key=redis_key)
self._tab_spider_status = setting.TAB_SPIDER_STATUS.format(redis_key=redis_key)
self._tab_requests = setting.TAB_REQUSETS.format(redis_key=redis_key)
self._tab_failed_requests = setting.TAB_FAILED_REQUSETS.format(
redis_key=redis_key
)
self._is_notify_end = False # 是否已经通知结束
self._last_task_count = 0 # 最近一次任务数量
self._redisdb = RedisDB()
self._project_total_state_table = "{}_total_state".format(self._project_name)
self._is_exist_project_total_state_table = False
# Request 缓存设置
Request.cached_redis_key = redis_key
Request.cached_expire_time = setting.RESPONSE_CACHED_EXPIRE_TIME
delete_keys = delete_keys or setting.DELETE_KEYS
if delete_keys:
self.delete_tables(delete_keys)
self._last_check_task_status_time = 0
self.wait_lock = wait_lock