def queue_items(self):
select_queue_sql = """
select id,action,params from hainiu_queue where type=1 and fail_times <= %s limit 0,%s for UPDATE;
"""
update_queue_sql = """
update hainiu_queue set type=0 where id in (%s);
"""
list = []
try:
d = DBUtil(config._HAINIU_DB)
sql = select_queue_sql % (self.fail_times,self.limit)
tuple = d.read_tuple(sql)
if len(tuple) == 0:
return list
queue_ids = ''
for t in tuple:
queue_id = t[0]
url = t[1]
param = '' if t[2] is None else t[2]
queue_ids += str(queue_id) + ','
c = NewsFindConsumer(url, param, queue_id)
list.append(c)
queue_ids = queue_ids[:-1]
d.execute(update_queue_sql % (queue_ids))
except:
self.rl.exception()
d.rollback()
d.commit()
finally:
d.close()
return list
def push_queue_items():
count_news_seed_queue_sql = """select count(*) from hainiu_queue where type=3 and fail_times=0;"""
insert_news_seed_internally_queue_items_sql = """
insert into hainiu_queue (type,action,params) values(3,%s,%s);
"""
count_news_seed_internally_sql = """select count(*) from hainiu_web_seed_internally where status=0 for update;"""
selec_news_seed_internally_sql = """select a_url,param,id from hainiu_web_seed_internally where status=0 limit %s,%s;"""
update_news_seed_internally_sql = """update hainiu_web_seed_internally set status=1 where id in (%s);"""
rl = LogUtil().get_base_logger()
try:
d = DBUtil(config._HAINIU_DB)
queue_total = d.read_one(count_news_seed_queue_sql)[0]
if queue_total != 0:
rl.info('last download_page queue not finish,last queue %s unFinish' % (queue_total))
return
starttime = time.clock()
d = DBUtil(config._HAINIU_DB)
total = long(d.read_one(count_news_seed_internally_sql)[0])
page_size = 1000
page = total / page_size
for i in range(0, page + 1):
sql = selec_news_seed_internally_sql % (0, page_size)
list = d.read_tuple(sql)
values = []
id_values = []
for l in list:
url = l[0]
url = url if url is not None else ''
param = l[1]
param = param if param is not None else ''
values.append((url,param))
id = l[2]
id_values.append(str(id))
if id_values.__len__() != 0:
random.shuffle(values)
d.executemany_no_commit(insert_news_seed_internally_queue_items_sql,values)
ids = ','.join(id_values)
sql = update_news_seed_internally_sql % (ids)
d.execute(sql)
endtime = time.clock()
worksec = int(round((endtime - starttime)))
rl.info('push seed_internally queue finish,total items %s,action time %s\'s' % (total,worksec))
except:
rl.exception()
rl.error(sql)
d.rollback()
finally:
d.close()
def push_queue_items():
count_news_seed_sql = """select count(*) from hainiu_web_seed where status=0;"""
select_news_seed_sql = """select url,category,last_crawl_time from hainiu_web_seed where status=0 limit %s,%s;"""
insert_news_seed_queue_items_sql = """insert into hainiu_queue (type,action,params) values(1,%s,%s);"""
count_news_seed_queue_sql = """select count(*) from hainiu_queue where type=1 and fail_times=0;"""
rl = LogUtil().get_base_logger()
try:
d = DBUtil(config._HAINIU_DB)
queue_total = d.read_one(count_news_seed_queue_sql)[0]
if queue_total != 0:
rl.info('last news_find queue not finish,last queue %s unFinish' %
(queue_total))
return
starttime = time.clock()
total = long(d.read_one(count_news_seed_sql)[0])
page_size = 1000
page = total / page_size
for i in range(0, page + 1):
sql = select_news_seed_sql % (i * page_size, page_size)
list = d.read_tuple(sql)
values = []
for l in list:
url = l[0]
publisher = get_tld(url)
publisher = publisher[0:publisher.index((
'.'))] if publisher.__contains__('.') else publisher
param = {}
param['category'] = l[1]
param['publisher'] = publisher
param = json.dumps(param, ensure_ascii=False)
values.append((url, param))
if values.__len__() != 0:
random.shuffle(values)
d.executemany(insert_news_seed_queue_items_sql, values)
endtime = time.clock()
worksec = int(round((endtime - starttime)))
rl.info(
'push news_find queue finish,total items %s,action time %s\'s' %
(total, worksec))
except:
rl.exception()
rl.error(sql)
d.rollback()
finally:
d.close()
def push_queue_items():
inert_sql = """
insert into hainiu_queue (type,params,action) values(1,%s,%s);
"""
count_sql = """
select count(1) from hainiu_queue where type=1;
"""
select_sql = """
select id from hainiu_queue where type=1 limit %s,%s;
"""
rl = LogUtil().get_base_logger()
try:
d = DBUtil(config._HAINIU_DB)
sql = inert_sql
insert_list = [("aaa", "bbb"), ("dffddf", "awwee")]
d.executemany(sql, insert_list)
sql = count_sql
queue_total = d.read_one(sql)[0]
print "queue_total", queue_total
page_size = 10
page = (queue_total / page_size) + 1
print "page", page
for i in range(0, page):
sql = select_sql % (i * page_size, page_size)
select_list = d.read_tuple(sql)
print "page", i
for record in select_list:
id = record[0]
print id
except:
rl.exception()
rl.error(sql)
d.rollback()
finally:
d.close()
def redis2Hdfs():
select_xpath_rule_sql = """select host,xpath,type from stream_extract_xpath_rule where host='%s' and status=0"""
rl = LogUtil().get_base_logger()
try:
d = DBUtil(_ZZ_DB)
start = 0
is_finish = True
host_set = set()
f = FileUtil()
t = TimeUtil()
time_str = t.now_time(format='%Y%m%d%H%M%S')
#local_xpath_file_path = '/user/zengqingyong17/spark/xpath_cache_file' + time_str
local_xpath_file_path = 'E:/python_workspaces/data/xpath/xpath_file' + time_str
starttime = time.clock()
r = redis.Redis('nn1.hadoop', '6379', db=6)
while is_finish:
values = set()
rs = r.scan(start, "total_z:*", 10)
# 新游标
start = rs[0]
if start ==0:
is_finish = False
# print rs
for i in rs[1]:
host = i.split(":")[1]
total_key = i
txpath_key = 'txpath_z:%s' % host
fxpath_key = 'fxpath_z:%s' % host
total = r.get(total_key)
# 降序排序获得次数(0,1)
txpath = r.zrevrange(txpath_key, 0, 1)
row_format = "%s\t%s\t%s\t%s"
if txpath:
txpath_num = int(r.zscore(txpath_key, txpath[0]))
if txpath.__len__() == 2:
# 返回txpath_key 中txpath[1]的数值
txpath_num_1 = int(r.zscore(txpath_key, txpath[1]))
txpath_num_1 = txpath_num_1 if txpath_num_1 is not None else 0
if txpath_num / float(total) >= 0.8:
values.add(row_format % (host, txpath[0], 'true', '0'))
host_set.add(host)
else:
if txpath_num >= 100:
values.add(row_format % (host, txpath[0], 'true', '0'))
host_set.add(host)
if txpath_num_1 is not None and txpath_num_1 >= 100:
values.add(row_format % (host, txpath[1], 'true', '0'))
host_set.add(host)
# 获得fxpath_key的全部值
fxpath = r.smembers(fxpath_key)
if fxpath:
# print 'fxpath:%s' % fxpath
for fx in fxpath:
values.add(row_format % (host, fx, 'false', '1'))
host_set.add(host)
sql = select_xpath_rule_sql % host
list_rule = d.read_tuple(sql)
for rule in list_rule:
type = rule[2]
if type == 0:
values.add(row_format % (rule[0], rule[1], 'true', '2'))
host_set.add(host)
elif type == 1:
values.add(row_format % (rule[0], rule[1], 'false', '3'))
host_set.add(host)
f.write_file_line_pattern(local_xpath_file_path, values, "a")
#上传到HDFS的XPATH配置文件目录
# c.upload("/user/qingniu/xpath_cache_file/", local_xpath_file_path)
endtime = time.clock()
worksec = int(round((endtime - starttime)))
rl.info('total host %s,action time %s\'s' % (host_set.__len__(), worksec))
except:
rl.exception()
d.rollback()
finally:
d.close()
def xpath_config_file():
select_xpath_rule_sql = """select host,xpath,type from stream_extract_xpath_rule where host='%s' and status=0"""
rl = LogUtil().get_base_logger()
try:
# _HAINIU_DB = {'HOST': '192.168.137.190', 'USER': '******', 'PASSWD': '12345678', 'DB': 'hainiucrawler',
# 'CHARSET': 'utf8', 'PORT': 3306}
d = DBUtil(config._HAINIU_DB)
# d = DBUtil(_HAINIU_DB)
r = redis.Redis('nn1.hadoop', 6379, db=6)
# r = redis.Redis('redis.hadoop', 6379, db=6)
f = FileUtil()
t = TimeUtil()
c = Client("http://nn1.hadoop:50070")
time_str = t.now_time(format='%Y%m%d%H%M%S')
# local_xpath_file_path = '/Users/leohe/Data/input/xpath_cache_file/xpath_file' + time_str
local_xpath_file_path = '/home/qingniu/xpath_cache_file/xpath_file' + time_str
start_cursor = 0
is_finish = True
starttime = time.clock()
host_set = set()
while is_finish:
values = set()
limit = r.scan(start_cursor, 'total:*', 10)
if limit[0] == 0:
is_finish = False
start_cursor = limit[0]
for h in limit[1]:
host = h.split(":")[1]
total_key = h
txpath_key = 'txpath:%s' % host
fxpath_key = 'fxpath:%s' % host
total = r.get(total_key)
txpath = r.zrevrange(txpath_key, 0, 1)
row_format = "%s\t%s\t%s\t%s"
if txpath:
# print 'txpath:%s' % txpath
txpath_num = int(r.zscore(txpath_key, txpath[0]))
if txpath.__len__() == 2:
txpath_num_1 = int(r.zscore(txpath_key, txpath[1]))
txpath_num_1 = txpath_num_1 if txpath_num_1 is not None else 0
# print 'txpath_max_num:%s' % txpath_num
if txpath_num / float(total) >= 0.8:
values.add(row_format % (host, txpath[0], 'true', '0'))
host_set.add(host)
else:
if txpath_num >= 1:
values.add(row_format %
(host, txpath[0], 'true', '0'))
host_set.add(host)
if txpath_num_1 is not None and txpath_num_1 >= 1:
values.add(row_format %
(host, txpath[1], 'true', '0'))
host_set.add(host)
fxpath = r.smembers(fxpath_key)
if fxpath:
# print 'fxpath:%s' % fxpath
for fx in fxpath:
values.add(row_format % (host, fx, 'false', '0'))
host_set.add(host)
sql = select_xpath_rule_sql % host
list_rule = d.read_tuple(sql)
for rule in list_rule:
type = rule[2]
if type == 0:
values.add(row_format %
(rule[0], rule[1], 'true', '2'))
host_set.add(host)
elif type == 1:
values.add(row_format %
(rule[0], rule[1], 'false', '3'))
host_set.add(host)
f.write_file_line_pattern(local_xpath_file_path, values, "a")
#上传到HDFS的XPATH配置文件目录
c.upload("/user/qingniu/xpath_cache_file/", local_xpath_file_path)
endtime = time.clock()
worksec = int(round((endtime - starttime)))
rl.info('total host %s,action time %s\'s' %
(host_set.__len__(), worksec))
except:
rl.exception()
d.rollback()
finally:
d.close()
