class TaskQueue(object):
'''
task queue for scheduler, have a priority queue and a time queue for delayed tasks
'''
processing_timeout = 10 * 60
def __init__(self, rate=0, burst=0):
self.mutex = threading.Lock()
self.priority_queue = PriorityTaskQueue()
self.time_queue = PriorityTaskQueue()
self.processing = PriorityTaskQueue()
self.bucket = Bucket(rate=rate, burst=burst)
@property
def rate(self):
return self.bucket.rate
@rate.setter
def rate(self, value):
self.bucket.rate = value
@property
def burst(self):
return self.burst.burst
@burst.setter
def burst(self, value):
self.bucket.burst = value
def check_update(self):
self._check_time_queue()
self._check_processing()
def _check_time_queue(self):
now = time.time()
self.mutex.acquire()
while self.time_queue.qsize() and self.time_queue.top.exetime < now:
task = self.time_queue.get()
task.exetime = 0
self.priority_queue.put(task)
self.mutex.release()
def _check_processing(self):
now = time.time()
self.mutex.acquire()
while self.processing.qsize() and self.processing.top.exetime < now:
task = self.processing.get()
if task.taskid is None:
continue
task.exetime = 0
self.priority_queue.put(task)
logging.info("[processing: retry] %s" % task.taskid)
self.mutex.release()
def put(self, taskid, priority=0, exetime=0):
now = time.time()
self.mutex.acquire()
if taskid in self.priority_queue:
task = self.priority_queue[taskid]
if priority > task.priority:
task.priority = priority
elif taskid in self.time_queue:
task = self.time_queue[taskid]
if priority > task.priority:
task.priority = priority
if exetime < task.exetime:
task.exetime = exetime
else:
task = InQueueTask(taskid, priority)
if exetime and exetime > now:
task.exetime = exetime
self.time_queue.put(task)
else:
self.priority_queue.put(task)
self.mutex.release()
def get(self):
if self.bucket.get() < 1:
return None
now = time.time()
self.mutex.acquire()
try:
task = self.priority_queue.get_nowait()
self.bucket.desc()
except Queue.Empty:
self.mutex.release()
return None
task.exetime = now + self.processing_timeout
self.processing.put(task)
self.mutex.release()
return task.taskid
def done(self, taskid):
if taskid in self.processing:
self.processing[taskid].taskid = None
def __len__(self):
return self.priority_queue.qsize() + self.time_queue.qsize()
def __contains__(self, taskid):
return taskid in self.priority_queue or taskid in self.time_queue
def __init__(self, rate=0, burst=0):
self.mutex = threading.Lock()
self.priority_queue = PriorityTaskQueue()
self.time_queue = PriorityTaskQueue()
self.processing = PriorityTaskQueue()
self.bucket = Bucket(rate=rate, burst=burst)
def __init__(self, rate=0, burst=0):
self.mutex = threading.Lock()
self.priority_queue = PriorityTaskQueue()
self.time_queue = PriorityTaskQueue()
self.processing = PriorityTaskQueue()
self.bucket = Bucket(rate=rate, burst=burst)
class TaskQueue(object):
'''
task queue for scheduler, have a priority queue and a time queue for delayed tasks
'''
processing_timeout = 10 * 60
def __init__(self, rate=0, burst=0):
self.mutex = threading.Lock()
self.priority_queue = PriorityTaskQueue()
self.time_queue = PriorityTaskQueue()
self.processing = PriorityTaskQueue()
self.bucket = Bucket(rate=rate, burst=burst)
@property
def rate(self):
return self.bucket.rate
@rate.setter
def rate(self, value):
self.bucket.rate = value
@property
def burst(self):
return self.burst.burst
@burst.setter
def burst(self, value):
self.bucket.burst = value
def check_update(self):
self._check_time_queue()
self._check_processing()
def _check_time_queue(self):
now = time.time()
self.mutex.acquire()
while self.time_queue.qsize() and self.time_queue.top.exetime < now:
task = self.time_queue.get()
task.exetime = 0
self.priority_queue.put(task)
self.mutex.release()
def _check_processing(self):
now = time.time()
self.mutex.acquire()
while self.processing.qsize() and self.processing.top.exetime < now:
task = self.processing.get()
if task.taskid is None:
continue
task.exetime = 0
self.priority_queue.put(task)
logging.info("[processing: retry] %s" % task.taskid)
self.mutex.release()
def put(self, taskid, priority=0, exetime=0):
now = time.time()
self.mutex.acquire()
if taskid in self.priority_queue:
task = self.priority_queue[taskid]
if priority > task.priority:
task.priority = priority
self.priority_queue.resort()
elif taskid in self.time_queue:
task = self.time_queue[taskid]
if priority > task.priority:
task.priority = priority
if exetime < task.exetime:
task.exetime = exetime
self.time_queue.resort()
elif taskid in self.processing and self.processing[taskid].taskid:
# force update a processing task is not allowed as there are so many
# problems may happen
return
else:
task = InQueueTask(taskid, priority)
if exetime and exetime > now:
task.exetime = exetime
self.time_queue.put(task)
else:
self.priority_queue.put(task)
self.mutex.release()
def get(self):
if self.bucket.get() < 1:
return None
now = time.time()
self.mutex.acquire()
try:
task = self.priority_queue.get_nowait()
self.bucket.desc()
except Queue.Empty:
self.mutex.release()
return None
task.exetime = now + self.processing_timeout
self.processing.put(task)
self.mutex.release()
return task.taskid
def done(self, taskid):
if taskid in self.processing:
self.processing.queue_dict.pop(taskid).taskid = None
return True
return False
def __len__(self):
return self.priority_queue.qsize() + self.time_queue.qsize()
def __contains__(self, taskid):
if taskid in self.priority_queue or taskid in self.time_queue:
return True
if taskid in self.processing and self.processing[taskid].taskid:
return True
return False
class TaskQueue(object):
'''
task queue for scheduler, have a priority queue and a time queue for delayed tasks
'''
processing_timeout = 10 * 60
def __init__(self, rate=0, burst=0):
self.mutex = threading.RLock()
self.priority_queue = PriorityTaskQueue()
self.time_queue = PriorityTaskQueue()
self.processing = PriorityTaskQueue()
self.bucket = Bucket(rate=rate, burst=burst)
@property
def rate(self):
return self.bucket.rate
@rate.setter
def rate(self, value):
self.bucket.rate = value
@property
def burst(self):
return self.burst.burst
@burst.setter
def burst(self, value):
self.bucket.burst = value
def check_update(self):
'''
Check time queue and processing queue
put tasks to priority queue when execute time arrived or process timeout
'''
self._check_time_queue()
self._check_processing()
def _check_time_queue(self):
now = time.time()
self.mutex.acquire()
while self.time_queue.qsize() and self.time_queue.top and self.time_queue.top.exetime < now:
task = self.time_queue.get_nowait()
task.exetime = 0
self.priority_queue.put(task)
self.mutex.release()
def _check_processing(self):
now = time.time()
self.mutex.acquire()
while self.processing.qsize() and self.processing.top and self.processing.top.exetime < now:
task = self.processing.get_nowait()
if task.taskid is None:
continue
task.exetime = 0
self.priority_queue.put(task)
logger.info("processing: retry %s", task.taskid)
self.mutex.release()
def put(self, taskid, priority=0, exetime=0):
'''Put a task into task queue'''
now = time.time()
task = InQueueTask(taskid, priority, exetime)
self.mutex.acquire()
if taskid in self.priority_queue:
self.priority_queue.put(task)
elif taskid in self.time_queue:
self.time_queue.put(task)
elif taskid in self.processing and self.processing[taskid].taskid:
# force update a processing task is not allowed as there are so many
# problems may happen
pass
else:
if exetime and exetime > now:
self.time_queue.put(task)
else:
self.priority_queue.put(task)
self.mutex.release()
def get(self):
'''Get a task from queue when bucket available'''
if self.bucket.get() < 1:
return None
now = time.time()
self.mutex.acquire()
try:
task = self.priority_queue.get_nowait()
self.bucket.desc()
except Queue.Empty:
self.mutex.release()
return None
task.exetime = now + self.processing_timeout
self.processing.put(task)
self.mutex.release()
return task.taskid
def done(self, taskid):
'''Mark task done'''
if taskid in self.processing:
del self.processing[taskid]
return True
return False
def size(self):
return self.priority_queue.qsize() + self.time_queue.qsize() + self.processing.qsize()
def __len__(self):
return self.size()
def __contains__(self, taskid):
if taskid in self.priority_queue or taskid in self.time_queue:
return True
if taskid in self.processing and self.processing[taskid].taskid:
return True
return False
import requests
from bs4 import BeautifulSoup
import proxy_switcher
from api_helper import CRAWLER_HEADERS
from token_bucket import Bucket
from utils.jsdati import JsdatiApi
from config.config import V2EX_USERNAME, V2EX_PASSWORD, JSDATI_USERNAME, JSDATI_PASSWORD
from utils.notification import wechat_notify
V2EX_INDEX_URL = 'https://www.v2ex.com'
V2EX_SIGNIN_URL = 'https://www.v2ex.com/signin'
V2EX_TOPIC_WEB_URL = 'https://www.v2ex.com/t/{topic_id}'
bucket = Bucket(rate=0.3, burst=1)
dmapi = JsdatiApi(JSDATI_USERNAME, JSDATI_PASSWORD)
def consume_token(func):
@functools.wraps(func)
def wrapper(*args, **kwargs):
while bucket.get() < 1:
time.sleep(0.5)
bucket.desc()
return func(*args, **kwargs)
return wrapper
}
# V2EX API
V2EX_SITE_URL = 'https://www.v2ex.com'
STATS_API_PATH = '/api/site/stats.json'
ALL_NODES_PATH = '/api/nodes/all.json'
NODE_INFO_PATH = '/api/nodes/show.json' # param: `id` or `name`
LATEST_TOPICS_PATH = '/api/topics/latest.json'
TOPIC_INFO_PATH = '/api/topics/show.json' # param: `id`
REPLIES_OF_TOPIC_PATH = '/api/replies/show.json' # param: `topic_id`
MEMBER_INFO_PATH = '/api/members/show.json' # param: `id` or `username`
API_RATE_LIMIT_ONE_HOUR = 120
bucket = Bucket(rate=0.5, burst=1)
def consume_token(func):
@functools.wraps(func)
def wrapper(*args, **kwargs):
while bucket.get() < 1:
time.sleep(0.5)
bucket.desc()
return func(*args, **kwargs)
return wrapper
class APIHelper(object):
"""API service with traffic flow controller"""