class RedisFeed(object):
def __init__(self, crawlid, spiderid, url, urls_file, priority, port, host,
custom):
self.name = "redis_feed"
self.crawlid = crawlid
self.spiderid = spiderid
self.url = url
self.urls_file = urls_file
self.priority = priority
self.port = port
self.host = host
self.custom = custom
self.inc = 0
self.failed_count, self.failed_rate, self.sucess_rate = 0, 0, 0
if self.custom:
from custom_redis.client import Redis
else:
from redis import Redis
self.redis_conn = Redis(host=self.host, port=self.port)
self.clean_previous_task(self.crawlid)
@classmethod
def parse_args(cls):
parser = argparse.ArgumentParser(description="usage: %prog [options]")
parser.add_argument('-rh',
"--redis-host",
dest="host",
type=str,
default="127.0.0.1",
help="Redis host to feed in. ")
parser.add_argument('-rp',
"--redis-port",
dest="port",
type=int,
default=6379,
help="Redis port to feed in. ")
parser.add_argument('-u',
'--url',
type=str,
help="The url to crawl, a list of products. ")
parser.add_argument('-uf',
'--urls-file',
type=str,
help="The urlsfile to crawl, single product. ")
parser.add_argument('-c',
'--crawlid',
required=True,
type=str,
help="An unique Id for a crawl task. ")
parser.add_argument('-s',
'--spiderid',
required=True,
type=str,
help="The website you wanna crawl. ")
parser.add_argument('-p',
'--priority',
type=int,
default=100,
help="Feed in the task queue with priority. ")
parser.add_argument('--custom',
action="store_true",
help="Use the custom redis whether or not. ")
return cls(**vars(parser.parse_args()))
def clean_previous_task(self, crawlid):
failed_keys = self.redis_conn.keys("failed_download_*:%s" % crawlid)
for fk in failed_keys:
self.redis_conn.delete(fk)
self.redis_conn.delete("crawlid:%s" % crawlid)
self.redis_conn.delete("crawlid:%s:model" % crawlid)
def start(self):
sucess_rate, failed_rate = 0, 0
# item抓取
if self.urls_file:
with open(self.urls_file) as f:
lst = f.readlines()
lines_count = len(lst)
for index, url in enumerate(lst):
json_req = '{"url":"%s","crawlid":"%s","spiderid":"%s","callback":"parse_item", "priority":%s}' % (
url.strip("\357\273\277\r\n"), self.crawlid,
self.spiderid, self.priority)
self.failed_count += self.feed(self.get_name(), json_req)
sucess_rate, failed_rate = self.show_process_line(
lines_count, index + 1, self.failed_count)
self.redis_conn.hset("crawlid:%s" % self.crawlid,
"total_pages", lines_count)
self.redis_conn.expire("crawlid:%s" % self.crawlid,
2 * 24 * 60 * 60)
# 分类抓取
else:
url_list = self.url.split(" ")
lines_count = len(url_list)
for index, url in enumerate(url_list):
json_req = '{"url":"%s","crawlid":"%s","spiderid":"%s","callback":"parse","priority":%s}' % (
url.strip(),
self.crawlid,
self.spiderid,
self.priority,
)
self.failed_count += self.feed(self.get_name(), json_req)
sucess_rate, failed_rate = self.show_process_line(
lines_count, index + 1, self.failed_count)
print("\ntask feed complete. sucess_rate:%s%%, failed_rate:%s%%" %
(sucess_rate, failed_rate))
def get_name(self):
return "{sid}:item:queue".format(sid=self.spiderid)
def feed(self, queue_name, req):
if self.custom:
from custom_redis.client.errors import RedisError
else:
from redis import RedisError
try:
self.redis_conn.zadd(queue_name, req, -self.priority)
return 0
except RedisError:
traceback.print_exc()
return 1
def show_process_line(self, count, num, failed):
per = count / 100
success = num - failed
success_rate = success * 100.0 / count
failed_rate = failed * 100.0 / count
str_success_rate = "%.2f%% " % success_rate
str_failed_rate = "%.2f%% " % failed_rate
if num >= self.inc:
self.inc += per
if sys.platform == "win32":
import ctypes
std_out_handle = ctypes.windll.kernel32.GetStdHandle(-11)
color_ctl = ctypes.windll.kernel32.SetConsoleTextAttribute
color_ctl(std_out_handle, 2)
print("\r", str_success_rate, "")
color_ctl(std_out_handle, 32)
print(int(success_rate * 30 / 100) * ' ', "")
if int(failed_rate):
color_ctl(std_out_handle, 64)
print(int(failed_rate * 30 / 100) * ' ', "")
color_ctl(std_out_handle, 0)
color_ctl(std_out_handle, 4)
print(str_failed_rate, "")
color_ctl(std_out_handle, 7)
else:
print("\r", str_success_rate, "")
print(
"%s%s" %
(int(success_rate * 50 / 100) * '\033[42m \033[0m',
int(failed_rate * 50 / 100) * '\033[41m \033[0m'),
str_failed_rate)
return success_rate, failed_rate
class Scheduler(Logger):
# 记录当前正在处理的item, 在处理异常时使用
present_item = None
def __init__(self, crawler):
self.settings = crawler.settings
self.set_logger(crawler)
if self.settings.get("CUSTOM_REDIS"):
from custom_redis.client import Redis
else:
from redis import Redis
self.redis_conn = Redis(self.settings.get("REDIS_HOST"),
self.settings.get("REDIS_PORT"))
self.queue_name = "%s:*:queue"
self.queues = {}
@classmethod
def from_crawler(cls, crawler):
return cls(crawler)
def open(self, spider):
self.spider = spider
self.queue_name = self.queue_name % spider.name
spider.set_redis(self.redis_conn)
spider.set_logger(self.logger)
def request_to_dict(self, request):
headers = dict([(item[0].decode("ascii"), item[1])
for item in request.headers.items()])
req_dict = {
'url':
request.url,
'method':
request.method,
'headers':
headers,
'body':
request.body,
'cookies':
request.cookies,
'meta':
request.meta,
'_encoding':
request._encoding,
'dont_filter':
request.dont_filter,
'callback':
None if request.callback is None else request.callback.__name__,
'errback':
None if request.errback is None else request.errback.__name__,
}
return req_dict
@enqueue_request_method_wrapper
def enqueue_request(self, request):
req_dict = self.request_to_dict(request)
key = "{sid}:item:queue".format(sid=req_dict['meta']['spiderid'])
self.redis_conn.zadd(key, json.dumps(req_dict, cls=P22P3Encoder),
-int(req_dict["meta"]["priority"]))
self.logger.debug("Crawlid: '{id}' Url: '{url}' added to queue".format(
id=req_dict['meta']['crawlid'], url=req_dict['url']))
@next_request_method_wrapper
def next_request(self):
queues = self.redis_conn.keys(self.queue_name)
if queues:
queue = random.choice(queues)
self.logger.info("length of queue %s is %s" %
(queue, self.redis_conn.zcard(queue)))
item = None
if self.settings.get("CUSTOM_REDIS"):
item = self.redis_conn.zpop(queue)
else:
pipe = self.redis_conn.pipeline()
pipe.multi()
pipe.zrange(queue, 0, 0).zremrangebyrank(queue, 0, 0)
result, count = pipe.execute()
# 1.1.8 add
if result:
item = result[0]
if item:
item = json.loads(item)
self.present_item = item
headers = item.get("headers", {})
body = item.get("body")
if item.get("method"):
method = item.get("method")
else:
method = "GET"
try:
req = Request(item['url'],
method=method,
body=body,
headers=headers)
except ValueError:
req = Request('http://' + item['url'],
method=method,
body=body,
headers=headers)
if 'callback' in item:
cb = item['callback']
if cb and self.spider:
cb = getattr(self.spider, cb)
req.callback = cb
if 'errback' in item:
eb = item['errback']
if eb and self.spider:
eb = getattr(self.spider, eb)
req.errback = eb
if 'meta' in item:
item = item['meta']
# defaults not in schema
if 'curdepth' not in item:
item['curdepth'] = 0
if "retry_times" not in item:
item['retry_times'] = 0
for key in item.keys():
req.meta[key] = item[key]
if 'useragent' in item and item['useragent'] is not None:
req.headers['User-Agent'] = item['useragent']
if 'cookie' in item and item['cookie'] is not None:
if isinstance(item['cookie'], dict):
req.cookies = item['cookie']
elif isinstance(item['cookie'], (str, bytes)):
req.cookies = parse_cookie(item['cookie'])
return req
def close(self, reason):
self.logger.info("Closing Spider", {'spiderid': self.spider.name})
def has_pending_requests(self):
return False
class Scheduler(Logger):
def __init__(self, crawler):
self.settings = crawler.settings
self.set_logger(crawler)
if self.settings.get("CUSTOM_REDIS"):
from custom_redis.client import Redis
else:
from redis import Redis
self.redis_conn = Redis(self.settings.get("REDIS_HOST"),
self.settings.get("REDIS_PORT"))
self.queue_name = "%s:*:queue"
self.queues = {}
self.extract = tldextract.extract
@classmethod
def from_crawler(cls, crawler):
return cls(crawler)
def open(self, spider):
self.spider = spider
self.queue_name = self.queue_name%spider.name
spider.set_redis(self.redis_conn)
spider.set_logger(self.logger)
def request_to_dict(self, request):
req_dict = {
'url': request.url.decode('ascii'),
'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,
'callback': None if request.callback is None else request.callback.func_name,
'errback': None if request.errback is None else request.errback.func_name,
}
return req_dict
def enqueue_request(self, request):
req_dict = self.request_to_dict(request)
ex_res = self.extract(req_dict['url'])
key = "{sid}:{dom}.{suf}:queue".format(
sid=req_dict['meta']['spiderid'],
dom=ex_res.domain,
suf=ex_res.suffix)
self.redis_conn.zadd(key, json.dumps(req_dict), -int(req_dict["priority"]))
self.logger.debug("Crawlid: '{id}' Url: '{url}' added to queue"
.format(id=req_dict['meta']['crawlid'],
url=req_dict['url']))
def next_request(self):
queues = self.redis_conn.keys(self.queue_name)
if queues:
queue = random.choice(queues)
self.logger.info("length of queue %s is %s" %
(queue, self.redis_conn.zcard(queue)))
if self.settings.get("CUSTOM_REDIS"):
item = self.redis_conn.zpop(queue)
else:
pipe = self.redis_conn.pipeline()
pipe.multi()
pipe.zrange(queue, 0, 0).zremrangebyrank(queue, 0, 0)
result, count = pipe.execute()
item = result[0]
if item:
item = json.loads(item)
try:
req = Request(item['url'])
except ValueError:
req = Request('http://' + item['url'])
if 'callback' in item:
cb = item['callback']
if cb and self.spider:
cb = getattr(self.spider, cb)
req.callback = cb
if 'errback' in item:
eb = item['errback']
if eb and self.spider:
eb = getattr(self.spider, eb)
req.errback = eb
if 'meta' in item:
item = item['meta']
# defaults not in schema
if 'curdepth' not in item:
item['curdepth'] = 0
if "retry_times" not in item:
item['retry_times'] = 0
for key in item.keys():
req.meta[key] = item[key]
if 'useragent' in item and item['useragent'] is not None:
req.headers['User-Agent'] = item['useragent']
if 'cookie' in item and item['cookie'] is not None:
if isinstance(item['cookie'], dict):
req.cookies = item['cookie']
elif isinstance(item['cookie'], basestring):
req.cookies = parse_cookie(item['cookie'])
return req
def close(self, reason):
self.logger.info("Closing Spider", {'spiderid': self.spider.name})
def has_pending_requests(self):
return False
class SpiderFeeder(object):
def __init__(self, crawlid, spiderid, url, urls_file, priority, port, host,
custom):
self.crawlid = crawlid
self.spiderid = spiderid
self.url = url
self.urls_file = urls_file
self.priority = priority
self.port = port
self.host = host
self.custom = custom
self.inc = 0
self.failed_count, self.failed_rate, self.sucess_rate = 0, 0, 0
if self.custom:
from custom_redis.client import Redis
else:
from redis import Redis
self.redis_conn = Redis(host=self.host, port=self.port)
self.clean_previous_task(self.crawlid)
def clean_previous_task(self, crawlid):
failed_keys = self.redis_conn.keys("failed_download_*:%s" % crawlid)
for fk in failed_keys:
self.redis_conn.delete(fk)
self.redis_conn.delete("crawlid:%s" % crawlid)
self.redis_conn.delete("crawlid:%s:model" % crawlid)
def start(self):
success_rate, failed_rate = 0, 0
# item抓取
if self.urls_file:
with open(self.urls_file) as f:
lst = f.readlines()
lines_count = len(lst)
for index, url in enumerate(lst):
req = Request(url=url.strip("\357\273\277\r\n"),
callback="parse_item",
meta={
"crawlid": self.crawlid,
"spiderid": self.spiderid,
"priority": self.priority
})
self.failed_count += self.feed(self.get_name(),
pickle.dumps(req))
success_rate, failed_rate = \
self.show_process_line(
lines_count, index + 1, self.failed_count)
self.redis_conn.hset("crawlid:%s" % self.crawlid,
"total_pages", lines_count)
# 分类抓取
else:
url_list = self.url.split(" ")
lines_count = len(url_list)
for index, url in enumerate(url_list):
req = Request(url=url.strip(),
callback="parse",
meta={
"crawlid": self.crawlid,
"spiderid": self.spiderid,
"priority": self.priority
})
self.failed_count += self.feed(self.get_name(),
pickle.dumps(req))
sucess_rate, failed_rate = self.show_process_line(
lines_count, index + 1, self.failed_count)
print("\ntask feed complete. sucess_rate:%s%%, failed_rate:%s%%" %
(success_rate, failed_rate))
def get_name(self):
return "{sid}:request:queue".format(sid=self.spiderid)
def feed(self, queue_name, req):
if self.custom:
from custom_redis.client.errors import RedisError
else:
from redis import RedisError
try:
self.redis_conn.zadd(queue_name, req, -self.priority)
return 0
except RedisError:
traceback.print_exc()
return 1
def show_process_line(self, count, num, failed):
per = count / 100
success = num - failed
success_rate = success * 100.0 / count
failed_rate = failed * 100.0 / count
str_success_rate = "%.2f%% " % success_rate
str_failed_rate = "%.2f%% " % failed_rate
if num >= self.inc:
self.inc += per
if sys.platform == "win32":
import ctypes
std_out_handle = ctypes.windll.kernel32.GetStdHandle(-11)
color_ctl = ctypes.windll.kernel32.SetConsoleTextAttribute
color_ctl(std_out_handle, 2)
print("\r", str_success_rate, "")
color_ctl(std_out_handle, 32)
print(int(success_rate * 30 / 100) * ' ', "")
if int(failed_rate):
color_ctl(std_out_handle, 64)
print(int(failed_rate * 30 / 100) * ' ', "")
color_ctl(std_out_handle, 0)
color_ctl(std_out_handle, 4)
print(str_failed_rate, "")
color_ctl(std_out_handle, 7)
else:
print("\r", str_success_rate, "")
print(
"%s%s" %
(int(success_rate * 50 / 100) * '\033[42m \033[0m',
int(failed_rate * 50 / 100) * '\033[41m \033[0m'),
str_failed_rate)
return success_rate, failed_rate