class SchedulerServer():
'''
schedule server
send request to client
accept request result send to client
'''
def __init__(self, settings):
self.settings = settings
self.mq_class = load_object(settings['SCHEDULER_MEMORY_QUEUE'])
self.mqs = PriorityQueue(self.priority)
self.status = ScheduleStatus()
def has_pending_requests(self):
return len(self) > 0
def push_queue_request(self, request):
self._mq_push(request)
self.status.add_push_queue()
def next_request(self):
request = self.mqs.pop()
if request:
self.status.add_pop_queue()
return request
def __len__(self):
return len(self.mqs)
def _mq_push(self, request):
self.mqs.push(request, -request.priority)
def priority(self, priority):
return self.mq_class()
def open(self, spider):
#打开调度器，绑定spider，实例化mqs，dqs。
self.spider = spider
#PriorityQueue见import部分。
#mqs=memery_queues,queues使用的是自己实现的PriorityQueue。
#队列中没一项都是一个memeryqueue-scrapy.squeues.LifoMemoryQueue
self.mqs = PriorityQueue(self._newmq)
#当指定了JOBDIR，self.dqdir是JOBDIR的子目录。
#self.dqdir = setting['JOBDIR'] + '/request.queue'
#在获取这个文件夹path同时，也创建了这个文件夹。
#self._dq()：返回了一个diskqueue的PriorityQueue，会处理文件中的内容读入到内存。
self.dqs = self._dq() if self.dqdir else None
#返回dupefilter实例。
#RFPDupeFilter没有实现open方法，
#其基类BaseDupeFilter.open(),只有一句pass
#所以这个scheduler.open(),在不改变调度器和去重类的情况下，直接返回的是None
return self.df.open()
def _dq(self):
activef = join(self.dqdir, 'active.json')
if exists(activef):
with open(activef) as f:
prios = json.load(f)
else:
prios = ()
q = PriorityQueue(self._newdq, startprios=prios)
if q:
logger.info("Resuming crawl (%(queuesize)d requests scheduled)",
{'queuesize': len(q)}, extra={'spider': self.spider})
return q
def _dq(self):
activef = join(self.dqdir, 'active.json')
if exists(activef):
with open(activef) as f:
prios = json.load(f)
else:
prios = ()
q = PriorityQueue(self._newdq, startprios=prios)
if q:
log.msg(format="Resuming crawl (%(queuesize)d requests scheduled)",
spider=self.spider, queuesize=len(q))
return q
def __init__(self, crawler, dupefilter, jobdir=None, dqclass=None, mqclass=None, logunser=False, stats=None,run_as_daemon=False):
self.df = dupefilter
self.dqdir = self._dqdir(jobdir)
self.dqclass = dqclass
self.mqclass = mqclass
self.logunser = logunser
self.stats = stats
self.run_as_daemon = run_as_daemon
self.dqs = None
self.mqs = PriorityQueue(self._newmq)
#Scheduler.__init__(self, dupefilter, jobdir, dqclass, mqclass, logunser, stats)
self.redis_handler = redis_handler('localhost', 6379, 0)
self.redis_handler.connect_db()
crawler.signals.connect(self.enqueue, signal=signals.request_scheduled)
def __init__(self, key, scheduler):
self.scheduler = scheduler
self.df = scheduler.dupefilter
self.dqdir = self._dqdir(scheduler.jobdir, key)
self.dqclass = scheduler.dqclass
self.mqclass = scheduler.mqclass
self.logunser = scheduler.logunser
self.stats = scheduler.stats
self.total_concurrency = scheduler.total_concurrency
self.domain_concurrency = scheduler.domain_concurrency
self.ip_concurrency = scheduler.ip_concurrency
self.spider = None
self.mqs = PriorityQueue(self._newmq)
self.dqs = self._dq() if self.dqdir else None
return self.df.open()
def _dq(self):
#self.dqdir = setting['JOBDIR'] + '/request.queue'
#activef = setting['JOBDIR'] + '/request.queue' +'/active.json'
activef = join(self.dqdir, 'active.json')
#如果这个文件存在，则打开，并将文件中的内容，读取进入prios（json结构）
if exists(activef):
with open(activef) as f:
prios = json.load(f)
#如果不存在这个文件，则prios为空集合。
else:
prios = ()
q = PriorityQueue(self._newdq, startprios=prios)
if q:
logger.info("Resuming crawl (%(queuesize)d requests scheduled)",
{'queuesize': len(q)},
extra={'spider': self.spider})
return q
def __init__(self, settings):
self.settings = settings
self.mq_class = load_object(settings['SCHEDULER_MEMORY_QUEUE'])
self.mqs = PriorityQueue(self.priority)
self.status = ScheduleStatus()
PyPI: https://pypi.python.org/pypi/queuelib
Github: https://github.com/scrapy/queuelib
Document: 没有
"""
from queuelib import PriorityQueue, FifoDiskQueue
if __name__ == "__main__":
import string
import random
import time
def randstr():
res = list()
for _ in range(8):
res.append(random.choice(string.ascii_letters))
return "".join(res).encode("utf-8")
tasks = [(randstr(), random.randint(1, 5))for i in range(1000)]
st = time.clock()
qfactory = lambda priority: FifoDiskQueue('queue-dir-%s' % priority)
pq = PriorityQueue(qfactory)
for task in tasks:
pq.push(*task)
for _ in range(len(pq)):
pq.pop()
print("queuelib takes %.6f sec." % (time.clock() - st,))
class Scheduler(object):
#Scheduler实例有三个主要的属性或者说是数据结构：
#1 self.dupefilter 去重功能实例。
#2 self.dqs 磁盘队列管理器，里面按优先级为关键字，每一个优先级对应一个request队列。
#4 self.mqs 内存队列管理器，同上。
#__init__方法是通过类方法from_crawler来调用的，也就是一般都要一个crawler才方便实例化。
def __init__(self,
dupefilter,
jobdir=None,
dqclass=None,
mqclass=None,
logunser=False,
stats=None):
self.df = dupefilter
self.dqdir = self._dqdir(jobdir)
self.dqclass = dqclass
self.mqclass = mqclass
#这个东西是从setting读取出来的，用来干嘛呢？
self.logunser = logunser
#crawler的stats，这个属性是做什么的呢？
self.stats = stats
@classmethod
def from_crawler(cls, crawler):
settings = crawler.settings
#DUPEFILTER_CLASS = 'scrapy.dupefilters.RFPDupeFilter'
dupefilter_cls = load_object(settings['DUPEFILTER_CLASS'])
dupefilter = dupefilter_cls.from_settings(settings)
#SCHEDULER_DISK_QUEUE = 'scrapy.squeues.PickleLifoDiskQueue'
dqclass = load_object(settings['SCHEDULER_DISK_QUEUE'])
#SCHEDULER_MEMORY_QUEUE = 'scrapy.squeues.LifoMemoryQueue'
mqclass = load_object(settings['SCHEDULER_MEMORY_QUEUE'])
#LOG_UNSERIALIZABLE_REQUESTS：False
logunser = settings.getbool('LOG_UNSERIALIZABLE_REQUESTS')
return cls(dupefilter, job_dir(settings), dqclass, mqclass, logunser,
crawler.stats)
def has_pending_requests(self):
#len(self)会调用该类中__len__方法。计量内存和硬盘队列的总长。
return len(self) > 0
def __len__(self):
#如果self.dqs（diskqueue)不为空，则加上self.dqs,否则返回self.mqs
return len(self.dqs) + len(self.mqs) if self.dqs else len(self.mqs)
def open(self, spider):
#打开调度器，绑定spider，实例化mqs，dqs。
self.spider = spider
#PriorityQueue见import部分。
#mqs=memery_queues,queues使用的是自己实现的PriorityQueue。
#队列中没一项都是一个memeryqueue-scrapy.squeues.LifoMemoryQueue
self.mqs = PriorityQueue(self._newmq)
#当指定了JOBDIR，self.dqdir是JOBDIR的子目录。
#self.dqdir = setting['JOBDIR'] + '/request.queue'
#在获取这个文件夹path同时，也创建了这个文件夹。
#self._dq()：返回了一个diskqueue的PriorityQueue，会处理文件中的内容读入到内存。
self.dqs = self._dq() if self.dqdir else None
#返回dupefilter实例。
#RFPDupeFilter没有实现open方法，
#其基类BaseDupeFilter.open(),只有一句pass
#所以这个scheduler.open(),在不改变调度器和去重类的情况下，直接返回的是None
return self.df.open()
def close(self, reason):
if self.dqs:
#PriorityQueue.close()方法，获取到每一个prios，并关闭dqs中的每一个队列。
prios = self.dqs.close()
with open(join(self.dqdir, 'active.json'), 'w') as f:
json.dump(prios, f)
#mqs不做处理。
#self.df.close():如果JOBDIR存在则将，则关闭指定的request.seen文件。否则不做任何操作。
return self.df.close(reason)
def enqueue_request(self, request):
#去重标志为真，dupefilter中出现过当前request：
#不做操作，记录日志。
if not request.dont_filter and self.df.request_seen(request):
self.df.log(request, self.spider)
return False
#如果进入disk队列，就不会进入内存队列。
#入列先尝试disk，disk没有再尝试内存。
dqok = self._dqpush(request)
if dqok:
#self.stats是干嘛用的。
self.stats.inc_value('scheduler/enqueued/disk', spider=self.spider)
else:
self._mqpush(request)
self.stats.inc_value('scheduler/enqueued/memory',
spider=self.spider)
self.stats.inc_value('scheduler/enqueued', spider=self.spider)
return True
def next_request(self):
#出列先尝试从内存队列读取，内存队列没有，再从
request = self.mqs.pop()
if request:
self.stats.inc_value('scheduler/dequeued/memory',
spider=self.spider)
else:
request = self._dqpop()
if request:
self.stats.inc_value('scheduler/dequeued/disk',
spider=self.spider)
if request:
self.stats.inc_value('scheduler/dequeued', spider=self.spider)
return request
#优先级队列入列，确定使用哪一个队列的标志是request中的priority属性。
def _dqpush(self, request):
#self.dqs为空，也就是磁盘队列没有实例化，直接返回。
if self.dqs is None:
return
try:
#把request请求放入到对应的优先级队列中。
#request_to_dict方法可以把request实例，转换成一个字典：
#d = {'url': request.url.decode('ascii'), # urls should be safe (safe_string_url)
#'callback': cb,
#'errback': eb,
#'method': request.method,
#'headers': dict(request.headers),
#'body': request.body,
#'cookies': request.cookies,
#'meta': request.meta,
#'_encoding': request._encoding,
#'priority': request.priority,
#'dont_filter': request.dont_filter,}
reqd = request_to_dict(request, self.spider)
#字典根据优先级入列。
self.dqs.push(reqd, -request.priority)
except ValueError as e: # non serializable request
if self.logunser:
logger.error(
"Unable to serialize request: %(request)s - reason: %(reason)s",
{
'request': request,
'reason': e
},
exc_info=True,
extra={'spider': self.spider})
return
else:
#如果没有错误，则返回真。
return True
def _mqpush(self, request):
#内存队列，就直接入列，不转换成字典。
self.mqs.push(request, -request.priority)
def _dqpop(self):
if self.dqs:
d = self.dqs.pop()
if d:
return request_from_dict(d, self.spider)
def _newmq(self, priority):
return self.mqclass()
def _newdq(self, priority):
return self.dqclass(join(self.dqdir, 'p%s' % priority))
#当self.dqdir有值的时候，才会调用该方法
def _dq(self):
#self.dqdir = setting['JOBDIR'] + '/request.queue'
#activef = setting['JOBDIR'] + '/request.queue' +'/active.json'
activef = join(self.dqdir, 'active.json')
#如果这个文件存在，则打开，并将文件中的内容，读取进入prios（json结构）
if exists(activef):
with open(activef) as f:
prios = json.load(f)
#如果不存在这个文件，则prios为空集合。
else:
prios = ()
q = PriorityQueue(self._newdq, startprios=prios)
if q:
logger.info("Resuming crawl (%(queuesize)d requests scheduled)",
{'queuesize': len(q)},
extra={'spider': self.spider})
return q
def _dqdir(self, jobdir):
if jobdir:
dqdir = join(jobdir, 'requests.queue')
if not exists(dqdir):
os.makedirs(dqdir)
return dqdir
class Scheduler(object):
def __init__(self, dupefilter, jobdir=None, dqclass=None, mqclass=None, logunser=False, stats=None):
self.df = dupefilter
self.dqdir = self._dqdir(jobdir)
self.dqclass = dqclass
self.mqclass = mqclass
self.logunser = logunser
self.stats = stats
@classmethod
def from_crawler(cls, crawler):
settings = crawler.settings
dupefilter_cls = load_object(settings['DUPEFILTER_CLASS'])
dupefilter = dupefilter_cls.from_settings(settings)
dqclass = load_object(settings['SCHEDULER_DISK_QUEUE'])
mqclass = load_object(settings['SCHEDULER_MEMORY_QUEUE'])
logunser = settings.getbool('LOG_UNSERIALIZABLE_REQUESTS')
return cls(dupefilter, job_dir(settings), dqclass, mqclass, logunser, crawler.stats)
def has_pending_requests(self):
return len(self) > 0
def open(self, spider):
self.spider = spider
self.mqs = PriorityQueue(self._newmq)
self.dqs = self._dq() if self.dqdir else None
return self.df.open()
def close(self, reason):
if self.dqs:
prios = self.dqs.close()
with open(join(self.dqdir, 'active.json'), 'w') as f:
json.dump(prios, f)
return self.df.close(reason)
def enqueue_request(self, request):
if not request.dont_filter and self.df.request_seen(request):
self.df.log(request, self.spider)
return False
dqok = self._dqpush(request)
if dqok:
self.stats.inc_value('scheduler/enqueued/disk', spider=self.spider)
else:
self._mqpush(request)
self.stats.inc_value('scheduler/enqueued/memory', spider=self.spider)
self.stats.inc_value('scheduler/enqueued', spider=self.spider)
return True
def next_request(self):
request = self.mqs.pop()
if request:
self.stats.inc_value('scheduler/dequeued/memory', spider=self.spider)
else:
request = self._dqpop()
if request:
self.stats.inc_value('scheduler/dequeued/disk', spider=self.spider)
if request:
self.stats.inc_value('scheduler/dequeued', spider=self.spider)
return request
def __len__(self):
return len(self.dqs) + len(self.mqs) if self.dqs else len(self.mqs)
def _dqpush(self, request):
if self.dqs is None:
return
try:
reqd = request_to_dict(request, self.spider)
self.dqs.push(reqd, -request.priority)
except ValueError as e: # non serializable request
if self.logunser:
logger.error("Unable to serialize request: %(request)s - reason: %(reason)s",
{'request': request, 'reason': e},
exc_info=True, extra={'spider': self.spider})
return
else:
return True
def _mqpush(self, request):
self.mqs.push(request, -request.priority)
def _dqpop(self):
if self.dqs:
d = self.dqs.pop()
if d:
return request_from_dict(d, self.spider)
def _newmq(self, priority):
return self.mqclass()
def _newdq(self, priority):
return self.dqclass(join(self.dqdir, 'p%s' % priority))
def _dq(self):
activef = join(self.dqdir, 'active.json')
if exists(activef):
with open(activef) as f:
prios = json.load(f)
else:
prios = ()
q = PriorityQueue(self._newdq, startprios=prios)
if q:
logger.info("Resuming crawl (%(queuesize)d requests scheduled)",
{'queuesize': len(q)}, extra={'spider': self.spider})
return q
def _dqdir(self, jobdir):
if jobdir:
dqdir = join(jobdir, 'requests.queue')
if not exists(dqdir):
os.makedirs(dqdir)
return dqdir
def push(self, request, priority=0):
slot = request.get('meta', {}).get('scheduler_slot', None)
if slot not in self.pqueues:
self.pqueues[slot] = PriorityQueue(self.qfactory)
self._slots.append(slot)
self.pqueues[slot].push(request, priority)
class Scheduler(object):
def __init__(self,
dupefilter,
jobdir=None,
dqclass=None,
mqclass=None,
logunser=False,
stats=None):
self.df = dupefilter
self.dqdir = self._dqdir(jobdir)
self.dqclass = dqclass
self.mqclass = mqclass
self.logunser = logunser
self.stats = stats
@classmethod
def from_crawler(cls, crawler):
settings = crawler.settings
dupefilter_cls = load_object(settings['DUPEFILTER_CLASS'])
dupefilter = dupefilter_cls.from_settings(settings)
dqclass = load_object(settings['SCHEDULER_DISK_QUEUE'])
mqclass = load_object(settings['SCHEDULER_MEMORY_QUEUE'])
logunser = settings.getbool('LOG_UNSERIALIZABLE_REQUESTS')
return cls(dupefilter, job_dir(settings), dqclass, mqclass, logunser,
crawler.stats)
def has_pending_requests(self):
return len(self) > 0
def open(self, spider):
self.spider = spider
self.mqs = PriorityQueue(self._newmq)
self.dqs = self._dq() if self.dqdir else None
return self.df.open()
def close(self, reason):
if self.dqs:
prios = self.dqs.close()
with open(join(self.dqdir, 'active.json'), 'w') as f:
json.dump(prios, f)
return self.df.close(reason)
def enqueue_request(self, request):
if not request.dont_filter and self.df.request_seen(request):
self.df.log(request, self.spider)
return False
dqok = self._dqpush(request)
if dqok:
self.stats.inc_value('scheduler/enqueued/disk', spider=self.spider)
else:
self._mqpush(request)
self.stats.inc_value('scheduler/enqueued/memory',
spider=self.spider)
self.stats.inc_value('scheduler/enqueued', spider=self.spider)
return True
def next_request(self):
request = self.mqs.pop()
if request:
self.stats.inc_value('scheduler/dequeued/memory',
spider=self.spider)
else:
request = self._dqpop()
if request:
self.stats.inc_value('scheduler/dequeued/disk',
spider=self.spider)
if request:
self.stats.inc_value('scheduler/dequeued', spider=self.spider)
return request
def __len__(self):
return len(self.dqs) + len(self.mqs) if self.dqs else len(self.mqs)
def _dqpush(self, request):
if self.dqs is None:
return
try:
reqd = request_to_dict(request, self.spider)
self.dqs.push(reqd, -request.priority)
except ValueError as e: # non serializable request
if self.logunser:
logger.error(
"Unable to serialize request: %(request)s - reason: %(reason)s",
{
'request': request,
'reason': e
},
exc_info=True,
extra={'spider': self.spider})
return
else:
return True
def _mqpush(self, request):
self.mqs.push(request, -request.priority)
def _dqpop(self):
if self.dqs:
d = self.dqs.pop()
if d:
return request_from_dict(d, self.spider)
def _newmq(self, priority):
return self.mqclass()
def _newdq(self, priority):
return self.dqclass(join(self.dqdir, 'p%s' % priority))
def _dq(self):
activef = join(self.dqdir, 'active.json')
if exists(activef):
with open(activef) as f:
prios = json.load(f)
else:
prios = ()
q = PriorityQueue(self._newdq, startprios=prios)
if q:
logger.info("Resuming crawl (%(queuesize)d requests scheduled)",
{'queuesize': len(q)},
extra={'spider': self.spider})
return q
def _dqdir(self, jobdir):
if jobdir:
dqdir = join(jobdir, 'requests.queue')
if not exists(dqdir):
os.makedirs(dqdir)
return dqdir
class Scheduler(object):
def __init__(self, dupefilter, jobdir=None, dqclass=None, mqclass=None, logunser=False, stats=None):
self.df = dupefilter
self.dqdir = self._dqdir(jobdir)
self.dqclass = dqclass
self.mqclass = mqclass
self.logunser = logunser
self.stats = stats
@classmethod
def from_crawler(cls, crawler):
settings = crawler.settings
dupefilter_cls = load_object(settings['DUPEFILTER_CLASS'])
dupefilter = dupefilter_cls.from_settings(settings)
dqclass = load_object(settings['SCHEDULER_DISK_QUEUE'])
mqclass = load_object(settings['SCHEDULER_MEMORY_QUEUE'])
logunser = settings.getbool('LOG_UNSERIALIZABLE_REQUESTS')
return cls(dupefilter, job_dir(settings), dqclass, mqclass, logunser, crawler.stats)
def has_pending_requests(self):
return len(self) > 0
def open(self, spider):
self.spider = spider
self.mqs = PriorityQueue(self._newmq)
self.dqs = self._dq() if self.dqdir else None
return self.df.open()
def close(self, reason):
if self.dqs:
prios = self.dqs.close()
with open(join(self.dqdir, 'active.json'), 'w') as f:
json.dump(prios, f)
return self.df.close(reason)
def enqueue_request(self, request):
if not request.dont_filter and self.df.request_seen(request):
self.df.log(request, self.spider)
return
dqok = self._dqpush(request)
if dqok:
self.stats.inc_value('scheduler/enqueued/disk', spider=self.spider)
else:
self._mqpush(request)
self.stats.inc_value('scheduler/enqueued/memory', spider=self.spider)
self.stats.inc_value('scheduler/enqueued', spider=self.spider)
def next_request(self):
request = self.mqs.pop()
if request:
self.stats.inc_value('scheduler/dequeued/memory', spider=self.spider)
else:
request = self._dqpop()
if request:
self.stats.inc_value('scheduler/dequeued/disk', spider=self.spider)
if request:
self.stats.inc_value('scheduler/dequeued', spider=self.spider)
return request
def __len__(self):
return len(self.dqs) + len(self.mqs) if self.dqs else len(self.mqs)
def _dqpush(self, request):
if self.dqs is None:
return
try:
reqd = request_to_dict(request, self.spider)
self.dqs.push(reqd, -request.priority)
except ValueError, e: # non serializable request
if self.logunser:
log.msg(format="Unable to serialize request: %(request)s - reason: %(reason)s",
level=log.ERROR, spider=self.spider,
request=request, reason=e)
return
else:
class Slot(object):
def __init__(self, key, scheduler):
self.scheduler = scheduler
self.df = scheduler.dupefilter
self.dqdir = self._dqdir(scheduler.jobdir, key)
self.dqclass = scheduler.dqclass
self.mqclass = scheduler.mqclass
self.logunser = scheduler.logunser
self.stats = scheduler.stats
self.total_concurrency = scheduler.total_concurrency
self.domain_concurrency = scheduler.domain_concurrency
self.ip_concurrency = scheduler.ip_concurrency
self.spider = None
self.mqs = PriorityQueue(self._newmq)
self.dqs = self._dq() if self.dqdir else None
return self.df.open()
def close(self, reason):
if self.dqs:
prios = self.dqs.close()
with open(join(self.dqdir, 'active.json'), 'w') as f:
json.dump(prios, f)
return self.df.close(reason)
def has_pending_requests(self):
return len(self) > 0
def enqueue_request(self, request):
if not request.dont_filter and self.df.request_seen(request):
self.df.log(request, self.spider)
return
dqok = self._dqpush(request)
if dqok:
self.stats.inc_value('scheduler/enqueued/disk', spider=self.spider)
else:
self._mqpush(request)
self.stats.inc_value('scheduler/enqueued/memory', spider=self.spider)
self.stats.inc_value('scheduler/enqueued', spider=self.spider)
return len(self)
def next_request(self):
request = self.mqs.pop()
if request:
self.stats.inc_value('scheduler/dequeued/memory', spider=self.spider)
else:
request = self._dqpop()
if request:
self.stats.inc_value('scheduler/dequeued/disk', spider=self.spider)
if request:
self.stats.inc_value('scheduler/dequeued', spider=self.spider)
return request
def __len__(self):
return len(self.dqs) + len(self.mqs) if self.dqs else len(self.mqs)
def _dqpush(self, request):
if self.dqs is None:
return
try:
reqd = request_to_dict(request, self.spider)
self.dqs.push(reqd, -request.meta['priority'])
except ValueError as e: # non serializable request
if self.logunser:
log.msg(format="Unable to serialize request: %(request)s - reason: %(reason)s",
level=log.ERROR, spider=self.spider,
request=request, reason=e)
return
else:
return True
def _mqpush(self, request):
self.mqs.push(request, -request.meta['priority'])
def _dqpop(self):
if self.dqs:
d = self.dqs.pop()
if d:
return request_from_dict(d, self.spider)
def _newmq(self, priority):
return self.mqclass()
def _newdq(self, priority):
return self.dqclass(join(self.dqdir, 'p%s' % priority))
def _dq(self):
activef = join(self.dqdir, 'active.json')
if exists(activef):
with open(activef) as f:
priiios = json.load(f)
else:
prios = ()
q = PriorityQueue(self._newdq, startprios=prios)
if q:
log.msg(format="Resuming crawl (%(queuesize)d requests scheduled)",
spider=self.spider, queuesize=len(q))
return q
def _dqdir(self, jobdir, key):
if jobdir:
dqdir = join(jobdir, 'requests.queue.%s'%key)
if not exists(dqdir):
os.makedirs(dqdir)
return dqdir
class Scheduler(object):
def __init__(self,
dupefilter,
jobdir=None,
dqclass=None,
mqclass=None,
logunser=False,
stats=None):
self.df = dupefilter
self.dqdir = self._dqdir(jobdir)
self.dqclass = dqclass
self.mqclass = mqclass
self.logunser = logunser
self.stats = stats
@classmethod
def from_crawler(cls, crawler):
settings = crawler.settings
dupefilter_cls = load_object(settings['DUPEFILTER_CLASS'])
dupefilter = dupefilter_cls.from_settings(settings)
dqclass = load_object(settings['SCHEDULER_DISK_QUEUE'])
mqclass = load_object(settings['SCHEDULER_MEMORY_QUEUE'])
logunser = settings.getbool('LOG_UNSERIALIZABLE_REQUESTS')
return cls(dupefilter, job_dir(settings), dqclass, mqclass, logunser,
crawler.stats)
def has_pending_requests(self):
return len(self) > 0
def open(self, spider):
self.spider = spider
self.mqs = PriorityQueue(self._newmq)
self.dqs = self._dq() if self.dqdir else None
return self.df.open()
def close(self, reason):
if self.dqs:
prios = self.dqs.close()
with open(join(self.dqdir, 'active.json'), 'w') as f:
json.dump(prios, f)
return self.df.close(reason)
def enqueue_request(self, request):
if not request.dont_filter and self.df.request_seen(request):
self.df.log(request, self.spider)
return
dqok = self._dqpush(request)
if dqok:
self.stats.inc_value('scheduler/enqueued/disk', spider=self.spider)
else:
self._mqpush(request)
self.stats.inc_value('scheduler/enqueued/memory',
spider=self.spider)
self.stats.inc_value('scheduler/enqueued', spider=self.spider)
def next_request(self):
request = self.mqs.pop()
if request:
self.stats.inc_value('scheduler/dequeued/memory',
spider=self.spider)
else:
request = self._dqpop()
if request:
self.stats.inc_value('scheduler/dequeued/disk',
spider=self.spider)
if request:
self.stats.inc_value('scheduler/dequeued', spider=self.spider)
return request
def __len__(self):
return len(self.dqs) + len(self.mqs) if self.dqs else len(self.mqs)
def _dqpush(self, request):
if self.dqs is None:
return
try:
reqd = request_to_dict(request, self.spider)
self.dqs.push(reqd, -request.priority)
except ValueError, e: # non serializable request
if self.logunser:
log.msg(
format=
"Unable to serialize request: %(request)s - reason: %(reason)s",
level=log.ERROR,
spider=self.spider,
request=request,
reason=e)
return
else:
def open(self, spider):
self.spider = spider
self.mqs = PriorityQueue(self._newmq)
self.dqs = self._dq() if self.dqdir else None
return self.df.open()
def open(self, spider):
self.spider = spider
self.mqs = PriorityQueue(self._newmq)
self.dqs = self._dq() if self.dqdir else None
return self.df.open()
