class UserGrabber:
def __init__(self):
self.db = Database(MONGODB)
self.ID_queue = queue.LifoQueue()
def __grab(self,uid,mode):
if not self.db.connected:
self.db.connect()
check = self.db.select({
'uid':uid,
'mode':mode,
},tname=FINISHED_TABLE,c_map=False)
if check:
logger.info(f'当前用户uid:{uid} 已经被爬取过 [{check[0].get("done_time")}].')
return 1
user = TTUser(uid)
followings = user.get_followings(MDB=1)
for i in followings:
self.ID_queue.put_nowait(i.get('user_id'))
logger.info(f'加载 用户ID：{uid} 关注的uid {len(followings)} 个进入队列.')
logger.info(f'开始爬取用户ID：{uid} 的头条数据.MODE[{mode}]')
if mode in [ARTICLE,VIDEO,WEITT]:
user.get_published(ALL=True,MDB=1,MODE=mode)
elif mode == 'all':
for i in [ARTICLE,VIDEO,WEITT]:
user.get_published(ALL=True, MDB=1, MODE=i)
else:
raise ValueError(f'头条用户链式抓取模式 mode 参数值错误:{mode}')
self.db.save({
'uid':uid,
'mode':mode,
'done_time':time_to_date(time.time()),
},tname=FINISHED_TABLE)
return 1
def run_forever(self,mode=ARTICLE):
if self.ID_queue.empty():
self.ID_queue.put_nowait(ENTER_USER_ID)
while 1:
if not self.ID_queue.empty():
uids = []
threads = []
res = []
for i in range(MAX_THREADS):
try:
uids.append(self.ID_queue.get_nowait())
except:
break
logger.info(f'已装载 {len(uids)} 个uid进入队列.')
for _,uid in enumerate(uids):
threads.append(GrabThread(self.__grab,args=(uid,mode)))
for i in threads:
i.start()
for i in threads:
i.join()
res.append(i.get_result())
count = sum([i for i in res if isinstance(i,int)])
logger.info(f'此次抽取队列uid爬取任务 实际完成 {count} 个.')
def search(self,
keyword,
count=COUNT_SEARCH,
USER=False,
VIDEO=False,
ALL=True,
MDB=None,
strict=False):
if USER:
tab = 4
elif VIDEO:
tab = 2
else:
tab = 1
tab_kind = {1: '综合', 2: '视频', 4: '用户'}
amount = 0
offset = 0
results = []
retries = MAX_RETRY
dbname = MONGODB['search']
if MDB:
if isinstance(MDB, Database) and not MDB.connected:
MDB.connect()
elif not isinstance(MDB, Database):
MDB = Database(MONGODB)
MDB.connect()
MDB.use_db(dbname)
while 1:
params = params_for_search(keyword, tab=tab, offset=offset)
response = send_request('get',
API_SEARCH,
params=params,
JSON=True,
retries=retries,
DATA=1,
headers=self.headers)
print(API_SEARCH)
print(params)
print(self.headers)
data = response.get('data')
if bool(data):
offset = response.get('offset')
for item in data:
if not ALL:
if amount >= count:
logger.info(f'[采集完毕] 已达到搜索要求的{count}条数据.[OK]')
return results
if MDB:
tname = f'{keyword}-{tab_kind[tab]}'
_id = item.get('id')
asks = MDB.select({'id': {"=": _id}}, tname=tname)
if asks:
continue
MDB.save(item, tname=tname)
if strict and USER:
name = item.get('name')
if name == keyword:
logger.info(f'[搜索匹配成功]Strict 模式下搜索到相关用户!')
return item
results.append(item)
amount += 1
logger.info(f'此次已搜索:{keyword} {tab_kind[tab]}数据 {amount} 条.')
if response.get('has_more') != 0:
retries = -1
else:
logger.info(
f'搜索关键词:{keyword} {tab_kind[tab]}数据采集完毕. 此次采集总数:{amount}.')
return results
def wrapper(self,*args,**kwargs):
res = func(self,*args,**kwargs)
params_func = res.get('params_func')
more = res.get('more','has_more')
more_out = res.get('more_out')
variables = res.get('var',{})
handler = res.get('condition_handle',{})
req_kwargs = res.get('request_kwargs',{})
args = res.get('extra_args',{})
res_args = res.get('res_args',{})
db_setup = res.get('db_setup',{})
var_outer = res.get('var_outer')
cleaner = res.get('cleaner')
data_out = res.get('data_out')
item_out = res.get('item_out')
item_callback = res.get('item_callback')
data_wrap = res.get('data_wrap',True)
count = kwargs.get('count',COUNT_HOTNEWS)
MDB = kwargs.get('MDB')
ALL = kwargs.get('ALL')
var_keys = [i for i in variables.keys()]
var_values = [i for i in variables.values()]
retries = MAX_RETRY
amount = 0
results = []
while 1:
params = params_func(*var_values,**args,**res_args)
if method.lower() == 'post':
req_kwargs.update({
'data': params
})
if 'params' in req_kwargs:
req_kwargs.pop('params')
else:
req_kwargs.update({
'params':params
})
if 'data' in req_kwargs:
req_kwargs.pop('data')
response = send_request(method, api,
retries=retries,
**req_kwargs)
if data_wrap:
data = response.get(data_out).get('data') if data_out else response.get('data')
else:
data = response
if bool(data):
if var_outer:
var_values = [response.get(var_outer).get(i) for i in var_keys]
else:
var_values = [response.get(i) for i in var_keys]
if res_args:
res_args.update({
'response':response
})
raw_data = data.get(item_out) if item_out else data
if not raw_data:
logger.info(f'数据抓取完毕. 此次采集总数:{amount}.')
return results
for item in raw_data:
if item_callback and isfunction(item_callback):
cb_res = item_callback(self,item)
if cb_res and not isinstance(cb_res,tuple):
continue
elif isinstance(cb_res,tuple) and cb_res[-1] == 200:
item = cb_res[0]
if not ALL:
if amount >= count:
logger.info(f'[采集完毕] 已达到搜索要求的{count}条数据.[OK]')
return results
if handler:
flags = []
for i in handler.keys():
_func = handler[i][-1]
_param = handler[i][0]
_sec_param = item.get(i)
if _func(_param,_sec_param):
flags.append(1)
else:
flags.append(0)
if all(flags):
logger.info(f'未满足抓取条件,略过,标识:{item.get(db_setup["ticket"])}')
continue
if MDB :
if isinstance(MDB,Database) and not MDB.connected:
MDB.connect()
elif not isinstance(MDB,Database):
MDB = Database(MONGODB)
MDB.connect()
MDB.use_db(db_setup['db'])
if cleaner and callable(cleaner):
item = cleaner(item)
_id = item.get(db_setup['ticket'])
asks = MDB.select({db_setup['ticket']: {"=": _id}}, tname=db_setup['tname'])
if asks:
continue
MDB.save(item, tname=db_setup['tname'])
results.append(item)
amount += 1
tip = f'此次抓取 数据 {amount} 条.' if not MDB else \
f'此次抓取 存入数据库:{db_setup.get("db")} 数据 {amount} 条.表:{db_setup.get("tname")}'
logger.info(tip)
if more_out:
_more = response.get(more_out).get(more)
else:
_more = response.get(more)
if _more:
retries += 1
else:
logger.info(f'数据抓取完毕. 此次采集总数:{amount}.')
return results