def run(self, params):
url = params['url']
flag = params['flag']
print(url)
try:
ps = crawl(url)
print(len(ps))
if len(ps) > 10:
uid = store2pg(ps=ps, url=url, flag=flag)
else:
uid = None
if uid:
urls = re.findall('"supdata_whereid":"(.*?)"', ps)
for u in urls:
url = 'http://www.changan.com.cn/news-details.shtml?whereid=%s&column_id=98' % u
log.info('入队列 jz_qymh_pagesource')
queue_job('main_changanqiche.Crawler2', {
'url': url,
'flag': flag
},
queue='jz_qymh_pagesource')
except Exception as e:
print(e)
print('重新入队')
log.info('入队列 jz_qymh_pagesource')
queue_job('main_changanqiche.Crawler1', {
'url': url,
'flag': flag
},
queue='jz_qymh_pagesource')
def run(self, params):
url = params['url']
flag = params['flag']
print(url)
try:
ps = crawl(url)
if len(ps) > 2000:
uid = store2pg(ps=ps, url=url, flag=flag)
else:
uid = None
if uid:
urls = re.compile(
'<a target="_blank" href="(http://www.feelcars.com/.*?html)">',
re.S).findall(ps)
for u in urls:
url = u
log.info('入队列 jz_qckj_pagesource')
queue_job('main_qichetansuowang.Crawler2', {
'url': url,
'flag': flag
},
queue='jz_qckj_pagesource')
except Exception as e:
print(e)
print('重新入队')
log.info('入队列 jz_qckj_pagesource')
queue_job('main_qichetansuowang.Crawler1', {
'url': url,
'flag': flag
},
queue='jz_qckj_pagesource')
def run(self, params):
url = params['url']
flag = params['flag']
print(url)
try:
ps = crawl(url)
if len(str(ps)) > 500:
uid = store2pg(ps=ps, url=url, flag=flag)
else:
uid = None
if uid:
urls = list(
re.compile(
'<p class="title">.*?<a href="(http:.*?)" target="_blank">',
re.S).findall(ps))
for u in urls:
url1 = u
log.info('入队列 jz_cj_pagesource')
queue_job('main_dongfangcaifu.Crawler2', {
'url': url1,
'flag': flag
},
queue='jz_cj_pagesource')
except Exception as e:
print(e)
print('重新入队')
log.info('入队列 jz_cj_pagesource')
queue_job('main_dongfangcaifu.Crawler1', {
'url': url,
'flag': flag
},
queue='jz_cj_pagesource')
def run(self, params):
url = params['url']
flag = params['flag']
print(url)
try:
ps = crawl(url)
print(len(ps))
if len(ps) > 100:
uid = store2pg(ps=ps, url=url, flag=flag)
else:
uid = None
if uid:
urls = re.findall('<a class="newsLink" href="(.*?)">', ps)
for u in urls:
url = 'https://www.lynkco.com.cn' + u
log.info('入队列 jz_qymh_pagesource')
queue_job('main_lingkeqiche.Crawler2', {
'url': url,
'flag': flag
},
queue='jz_qymh_pagesource')
except Exception as e:
print(e)
print('重新入队')
log.info('入队列 jz_qymh_pagesource')
queue_job('main_lingkeqiche.Crawler1', {
'url': url,
'flag': flag
},
queue='jz_qymh_pagesource')
def requeue_job(self, params, fpath=None, nqueue=None):
'''
requeue an unfinished job
run_wrapped的helper function
'''
log.warning('Job Failed, re-queue...%s' % params['url'])
queue_job(fpath, params, queue=nqueue)
def run(self, params):
users = [{"user": x} for x in get_distinct_users()]
print "Queue users"
queue_jobs("bluebirdlib.tasks.GetUserTweets", users, queue="tweets")
hts = [{"search": x} for x in HASHTAGS]
print "Queue hashtags"
queue_jobs("bluebirdlib.tasks.getHashTagTweets", hts, queue="tweets")
print "Queue scheduler"
queue_job("bluebirdlib.tasks.Scheduler", {}, queue="tweets")
return 0
def initialize_jobs(bucket_name):
setup_context()
jobs_count = 0
conn = S3Connection(AWS_ACCESS_KEY, AWS_SECRET_KEY)
bucket = Bucket(connection=conn, name=bucket_name)
for key in bucket.list():
queue_job("tasks.Download", {
"bucket_name": bucket_name,
"key_name": key.key
},
queue=download_queue)
jobs_count += 1
return jobs_count
def run(self, params):
#初始化的时候，选用页面长度为30
#每日增量，只要一页就可以
url_s1 = ('http://www.feelcars.com/category/xinnengyuan/page/%s',
'新能源')
end = 4
for i in range(1, end):
url = url_s1[0] % str(i)
log.info('入队列 jz_qckj_pagesource')
queue_job('main_qichetansuowang.Crawler1', {
'url': url,
'flag': url_s1[-1]
},
queue='jz_qckj_pagesource')
def main():
parser = argparse.ArgumentParser(description='Runs a task')
cfg = config.get_config(parser=parser, config_type="run", sources=("file", "env", "args"))
cfg["is_cli"] = True
set_current_config(cfg)
if len(cfg["taskargs"]) == 1:
params = json.loads(cfg["taskargs"][0]) # pylint: disable=no-member
else:
params = {}
# mrq-run taskpath a 1 b 2 => {"a": "1", "b": "2"}
for group in utils.group_iter(cfg["taskargs"], n=2):
if len(group) != 2:
print("Number of arguments wasn't even")
sys.exit(1)
params[group[0]] = group[1]
if cfg["queue"]:
ret = queue_job(cfg["taskpath"], params, queue=cfg["queue"])
print(ret)
else:
worker_class = load_class_by_path(cfg["worker_class"])
job = worker_class.job_class(None)
job.set_data({
"path": cfg["taskpath"],
"params": params,
"queue": cfg["queue"]
})
job.datestarted = datetime.datetime.utcnow()
set_current_job(job)
ret = job.perform()
print(json_stdlib.dumps(ret, cls=MongoJSONEncoder)) # pylint: disable=no-member
def api_job_action():
params = {k: v for k, v in iteritems(request.form)}
if params.get("status") and "-" in params.get("status"):
params["status"] = params.get("status").split("-")
return jsonify({"job_id": queue_job("mrq.basetasks.utils.JobAction",
params,
queue=get_current_config()["dashboard_queue"])})
def api_job_action():
params = {k: v for k, v in request.form.iteritems()}
if params.get("status") and "-" in params.get("status"):
params["status"] = params.get("status").split("-")
return jsonify({"job_id": queue_job("mrq.basetasks.utils.JobAction",
params,
queue=get_current_config()["dashboard_queue"])})
def run(self, params):
params1 = (
'https://www.lynkco.com.cn/Brand/News/NewsMore?pageIndex=%s',
'新闻潮讯')
#初始化的时候，选用页面长度为30
for u in [params1]:
if u[-1] == '新闻潮讯':
end = 8
for i in range(1, end):
url = u[0] % str(i)
log.info('入队列 jz_qymh_pagesource')
queue_job('main_lingkeqiche.Crawler1', {
'url': url,
'flag': u[-1]
},
queue='jz_qymh_pagesource')
def run(self, params):
key_name = params["key_name"]
_, extension = os.path.splitext(key_name)
if (extension == ".jpg"):
conn = S3Connection(AWS_ACCESS_KEY, AWS_SECRET_KEY)
bucket = conn.get_bucket(params["bucket_name"])
key = bucket.get_key(key_name)
key.get_contents_to_filename(key_name)
log.info("Succesfully downloaded file from s3 bucket %s", key_name)
queue_job("tasks.Write", {"key_name": key_name}, queue=write_queue)
else:
#TODO handle compressed and other file types
log.warn(
"Currently unable to handle file extension type for file %s",
key_name)
os.remove(key_name)
def api_queue_job(task): #存在参数相互依赖的关系 无法使用 ApiSchemaWrapper
queue = request.args.get('queue', '').strip()
args = request.args.to_dict()
params, err = fixTaskParams(MRQ_TASK_DICT, task, args)
rst = ApiErrorBuild()
rst = {"job_id": queue_job(task, params, queue=queue if queue else None)} if err is None else err
return jsonify(rst)
def exec_push_work(self, url):
# import subprocess
# modify arguments
# 這些是會傳下去給spider的函數 除了spider需要用的之外，還包括所有栈名以及其任務的路徑
# 如果有更改栈名請於此更改
args = {
'url': url,
'spiderTask': 'spider.spider.__Spider',
'spiderqueue': 'crawl_posts',
'parseTask': 'parser.parse_posts.__Parser',
'parsequeue': 'parse_posts'
}
# task = ['spider.spider_crawl.LcSpider']
# command = ['mrq-run'] + task + args
# '--queue', 'crawl_posts'
# subprocess.Popen(command)
queue_job(args['spiderTask'], args, queue=args['spiderqueue'])
def run(self, params):
params1 = (
'https://www.changan.com.cn/news-changan?page=%s&year=%s&keyword=&type=0&ajax_req=1&t=1584689024944',
'长安动态')
params2 = ('http://www.changan.com.cn/company.shtml', '合资合作')
#初始化的时候，选用页面长度为30
for u in [params1]:
if u[-1] == '长安动态':
year = datetime.datetime.now().strftime('%Y')
#每周一次，一次一页（8篇）
for page in range(1, 2):
url = u[0] % (str(page), str(year))
log.info('入队列 jz_qymh_pagesource')
queue_job('main_changanqiche.Crawler1', {
'url': url,
'flag': u[-1]
},
queue='jz_qymh_pagesource')
def _crawl(self, batch):
responses = self._downloader.download(batch)
if self._time_sleep:
time.sleep(self._time_sleep)
for response in responses:
self._processor.logger.info(response)
callback = response.request.callback(response)
if isinstance(callback, types.GeneratorType):
pipe = self._queue.get_pipe()
for item in callback:
if isinstance(item, Request):
# logger.info("push request to queue..." + str(item))
if self._should_follow(item):
self._queue.push_pipe(item, pipe)
elif isinstance(item, PipeItem):
# 如果返回对象是pipeItem，则用对应的pipeline处理
self._process_count += 1
for pipe_name in item.pipe_names:
queue_job(PIPELINE_TASK,
PipelineArgs(pipe_name, item.result),
queue=PIPELINE)
if self.test:
if self._process_count > 0:
return
elif isinstance(
item,
Violet): # 如果返回的是tuple，即详情页的processor和详情页的请求信息
queue_job(CRAWLER_TASK,
CrawlArgs(item.processor, item.request),
queue=CRAWLER)
else:
raise Exception('not return correct value!!!')
pipe.execute()
elif isinstance(callback, Request):
# logger.info("push request to queue..." + str(back))
if self._should_follow(callback):
self._queue.push(callback)
elif isinstance(callback, PipeItem):
# 如果返回对象是pipeItem，则用对应的pipeline处理
self._process_count += 1
for pipe_name in callback.pipe_names:
queue_job(PIPELINE_TASK,
PipelineArgs(pipe_name, callback.result),
queue=PIPELINE)
elif isinstance(callback,
Violet): # 如果返回的是tuple，即详情页的processor和详情页的请求信息
queue_job(CRAWLER_TASK,
CrawlArgs(item.processor, item.request),
queue=CRAWLER)
else:
# # 如果返回对象不是pipeItem，则默认用每个pipeline处理
raise Exception('not return correct value!!!')
def wait_for_job(path, params, **kwargs):
job_id = queue_job(path, params, **kwargs)
while True:
time.sleep(5)
res = get_job_result(job_id)
if res["status"] == "success":
return res.get("result")
elif res["status"] not in ["queued", "started", "interrupt"]:
raise Exception("Job %s was in status %s" %
(path, res.get("status")))
def wait_for_job(path, params, **kwargs):
job_id = queue_job(path, params, **kwargs)
while True:
time.sleep(5)
res = get_job_result(job_id)
if res["status"] == "success":
return res.get("result")
elif res["status"] not in ["queued", "started", "interrupt"]:
raise Exception("Job %s was in status %s" % (
path, res.get("status")
))
def run(self, params):
#初始化的时候，选用页面长度为30
#每日增量，只要一页就可以
url_s1 = ('http://finance.eastmoney.com/news/cjjsp_%s.html', '经济时评')
url_s2 = ('http://finance.eastmoney.com/news/cgnjj_%s.html', '国内经济')
url_s3 = ('http://finance.eastmoney.com/news/cgjjj_%s.html', '国际经济')
# for i in range(1,4):
for i in range(1, 26):
url1 = url_s1[0] % str(i)
log.info('入队列 jz_cj_pagesource')
queue_job('main_dongfangcaifu.Crawler1', {
'url': url1,
'flag': url_s1[-1]
},
queue='jz_cj_pagesource')
url2 = url_s2[0] % str(i)
log.info('入队列 jz_cj_pagesource')
queue_job('main_dongfangcaifu.Crawler1', {
'url': url2,
'flag': url_s2[-1]
},
queue='jz_cj_pagesource')
url3 = url_s3[0] % str(i)
log.info('入队列 jz_cj_pagesource')
queue_job('main_dongfangcaifu.Crawler1', {
'url': url3,
'flag': url_s3[-1]
},
queue='jz_cj_pagesource')
def create_job():
taskpath = request.form['path']
taskparams = json.loads(request.form['params'])
if taskpath.startswith("admin"):
return None
if g.user.is_authenticated():
taskparams["user"] = str(g.user.id)
job_id = queue_job("tasks.%s" % taskpath, taskparams)
return json.dumps({"job_id": str(job_id)})
def create_job():
taskpath = request.form['path']
taskparams = json.loads(request.form['params'])
if taskpath.startswith("admin"):
return None
if g.user.is_authenticated():
taskparams["user"] = str(g.user.id)
job_id = queue_job("tasks.%s" % taskpath, taskparams)
return json.dumps({"job_id": str(job_id)})
def send_crawl_task(sid, tel, flow_type, province, city, timeout=5, sleep_time=1):
alive = check_crawler_alive(sid)
if alive:
return False
else:
# send job queue to start crawler
params = {
'sid': sid,
'tel': tel,
'flow_type': flow_type,
'province': province,
'city': city
}
job_id = queue_job(config.TASK_PATH, params, queue=config.QUEUE_NAME)
if len(str(job_id))>11:
db['sid_info'].update_one(
{'sid': sid}, {'$set': {'job_id':job_id}}, upsert=True)
return True
else:
return False
def run(self, params):
url = params['url']
flag = params['flag']
try:
info = sess.query(Jz_dongfangcaifu_PageSource).filter_by(
url=url).first()
sess.rollback()
if not info:
ps = crawl(url)
if len(str(ps)) > 500 and '返回' not in str(ps):
uid = store2pg(ps=ps, url=url, flag=flag)
else:
uid = None
if uid:
log.info('入队列 jz_cj_parse')
queue_job('main_dongfangcaifu.Parse', {
'url': url,
'flag': flag
},
queue='jz_cj_parse')
else:
print('新闻已存在')
log.info('入队列 jz_cj_parse')
queue_job('main_dongfangcaifu.Parse', {
'url': url,
'flag': flag
},
queue='jz_cj_parse')
except Exception as e:
print(e)
print('重新入队')
log.info('入队列 jz_cj_pagesource')
queue_job('main_dongfangcaifu.Crawler2', {
'url': url,
'flag': flag
},
queue='jz_cj_pagesource')
def run(self, params):
url = params['url']
flag = params['flag']
try:
info = sess.query(Jz_lingkeqiche_PageSource).filter_by(
url=url).first()
if not info:
ps = crawl(url)
if len(ps) > 100:
uid = store2pg(ps=ps, url=url, flag=flag)
else:
uid = None
if uid:
log.info('入队列 jz_qymh_parse')
queue_job('main_lingkeqiche.Parse', {
'url': url,
'flag': flag
},
queue='jz_qymh_parse')
else:
print('新闻已存在')
log.info('入队列 jz_qymh_parse')
queue_job('main_lingkeqiche.Parse', {
'url': url,
'flag': flag
},
queue='jz_qymh_parse')
except Exception as e:
print(e)
print('重新入队')
log.info('入队列 jz_qymh_pagesource')
queue_job('main_lingkeqiche.Crawler2', {
'url': url,
'flag': flag
},
queue='jz_qymh_pagesource')
def run(self, params):
url = params['url']
flag = params['flag']
try:
info = sess.query(Jz_qichetansuowang_PageSource).filter_by(
url=url).first()
sess.rollback()
if not info:
ps = crawl(url)
if len(str(ps)) > 2000:
uid = store2pg(ps=ps, url=url, flag=flag)
else:
uid = None
if uid:
log.info('入队列 jz_qckj_parse')
queue_job('main_qichetansuowang.Parse', {
'url': url,
'flag': flag
},
queue='jz_qckj_parse')
else:
print('新闻已存在，并入解析')
queue_job('main_qichetansuowang.Parse', {
'url': url,
'flag': flag
},
queue='jz_qckj_parse')
except Exception as e:
print(e)
print('重新入队')
log.info('入队列 jz_qckj_pagesource')
queue_job('main_qichetansuowang.Crawler2', {
'url': url,
'flag': flag
},
queue='jz_qckj_pagesource')
def run(self, params):
return queue_job(params["path"], params["params"])
from mrq import context
from mrq.job import queue_job
context.setup_context()
result = queue_job("crawler.Fetch", {
"url": "http://docs.python-requests.org",
"from": "whatever.com"
},
queue="crawl")
print(result)
def run(self, params):
return queue_job(params["path"], params["params"], queue=params.get("queue"))
def run(self, params):
url = params['url']
flag = params['flag']
try:
info = sess1.query(Jz_changanqiche_content).filter_by(
url=url).first()
sess.rollback()
if not info:
info_2 = sess.query(Jz_changanqiche_PageSource).filter_by(
url=url).first()
sess.rollback()
ps = info_2.pagesource
ps_uid = info_2.uid
# author = re.findall('',ps)
# author = author[0] if author else None
public_time = re.findall('><span id="love_number">(.*?)</span',
ps)
public_time = public_time[0].strip() if public_time else None
content1 = re.compile(
'<div class="news-details-main">(.*?)<div class="details-main-btn"',
re.S).findall(ps)
content2 = content1[0] if content1 else None
pic = re.findall('src="(.*?)" ', content2) if content2 else []
for i in range(len(pic)):
pic[i] = 'https:' + pic[i]
pic = ';'.join(set(pic))
content2 = content2.replace('<br/>', '\n').replace(
'<img src', '[img src').replace('jpg"/>', 'jpg"/]')
content = re.sub('<.*?>', '',
content2.replace(' ', '')).replace(
'$(".content img").wrap("");',
'').strip()
title = re.findall('<h2>(.*?)</h2>', ps)[0]
meta_keywords = re.compile(
'<meta name="keywords" content="(.*?)">',
re.S).findall(ps)[0].strip()
hid = store2pg_parse(url=url,
author=None,
public_time=public_time,
page_source=ps_uid,
content=content,
website_name='长安汽车',
channel_name=flag,
title=title,
topic=None,
tag=meta_keywords,
meta_keywords=None,
pic=pic,
flag=None)
if hid:
print('完成')
else:
print('新闻解析已存在')
except Exception as e:
print(e)
if e != "'NoneType' object has no attribute 'replace'":
print('重新入队')
log.info('入队列 jz_qymh_parse')
queue_job('main_changanqiche.Parse', {
'url': url,
'flag': flag
},
queue='jz_qymh_parse')
def requeue_job(self, params, fpath=None, nqueue=None):
'''
Requeue a failed job
'''
log.warning('Job Failed, re-queue...%s' % params['url'])
queue_job(fpath, params, queue=nqueue)
def parser_job(self, params, fpath=None, nqueue=None):
'''
Add to the queue of the Parser task
'''
queue_job(fpath, params, queue=nqueue)
# -*- coding: utf-8 -*-
"""
Created with IntelliJ IDEA.
Description:
User: jinhuichen
Date: 3/19/2018 11:35 AM
Description:
"""
from mrq.job import queue_job
from constants.task_name import CRAWLER_TASK
from constants.queue_name import CRAWLER
from fetchman.pipeline.pipe_item import CrawlArgs
from processors.tuliu_processor import Tuliu_Processor
# from tasks.spider_task import no_queue_task
if __name__ == '__main__':
# res = no_queue_task({"processor": Tuliu_Processor.__name__})
# 启动初始化任务
res = queue_job(CRAWLER_TASK, CrawlArgs(Tuliu_Processor), queue=CRAWLER)
print(res)
def run(self, params):
collection = connections.mongodb_jobs.simple_crawler_urls
response = requests.get(params["url"])
if response.status_code != 200:
log.warning("Got status %s on page %s (Queued from %s)" % (
response.status_code, response.url, params.get("from")
))
return False
# Store redirects
if response.url != params["url"]:
collection.update({"_id": params["url"]}, {"$set": {
"redirected_to": response.url,
"fetched_date": datetime.datetime.now()
}})
document = lxml.html.fromstring(response.content)
document.make_links_absolute(response.url)
queued_count = 0
document_domain = urlparse.urlparse(response.url).netloc
for (element, attribute, link, pos) in document.iterlinks():
link = re.sub("#.*", "", link or "")
if not link:
continue
domain = urlparse.urlparse(link).netloc
# Don't follow external links for this example
if domain != document_domain:
continue
# We don't want to re-queue URLs twice. If we try to insert a duplicate,
# pymongo will throw an error
try:
collection.insert({"_id": link})
except:
continue
queue_job("crawler.Fetch", {
"url": link,
"from": params["url"]
}, queue="crawl")
queued_count += 1
stored_data = {
"_id": response.url,
"queued_urls": queued_count,
"html_length": len(response.content),
"fetched_date": datetime.datetime.now()
}
collection.update(
{"_id": response.url},
stored_data,
upsert=True
)
return True
def run(self, params):
return queue_job(params["path"], params["params"])
def run(self, params):
collection = connections.mongodb_jobs.simple_crawler_urls
response = requests.get(params["url"])
if response.status_code != 200:
log.warning(
"Got status %s on page %s (Queued from %s)" %
(response.status_code, response.url, params.get("from")))
return False
# Store redirects
if response.url != params["url"]:
collection.update({"_id": params["url"]}, {
"$set": {
"redirected_to": response.url,
"fetched_date": datetime.datetime.now()
}
})
document = lxml.html.fromstring(response.content)
document.make_links_absolute(response.url)
queued_count = 0
document_domain = urlparse.urlparse(response.url).netloc
for (element, attribute, link, pos) in document.iterlinks():
link = re.sub("#.*", "", link or "")
if not link:
continue
domain = urlparse.urlparse(link).netloc
# Don't follow external links for this example
if domain != document_domain:
continue
# We don't want to re-queue URLs twice. If we try to insert a duplicate,
# pymongo will throw an error
try:
collection.insert({"_id": link})
except:
continue
queue_job("crawler.Fetch", {
"url": link,
"from": params["url"]
},
queue="crawl")
queued_count += 1
stored_data = {
"_id": response.url,
"queued_urls": queued_count,
"html_length": len(response.content),
"fetched_date": datetime.datetime.now()
}
collection.update({"_id": response.url}, stored_data, upsert=True)
return True
