class DBTestCase(unittest.TestCase):
def setUp(self):
self.db = DBHelper()
self.samples =self.db.get_results_by_clms(columns='*',table=
'samples',database=DB_test,isdict=True)
def tearDown(self):
self.db = None
def test_get_max_group_id(self):
res =self.db.get_max_group_id('refined_info_2','xyzb')
res_2 =self.db.get_max_group_id('info_template','xyzb')
print res
print res_2
def test_batch_insert(self):
entrys =self.db.get_results_by_clms(columns='*',table=
'refined_info',database=DB_1,isdict=True)[:12056]
table = 'info_tmp1'
self.db.recr_table(table,DB_1,'info_template',DB_1);
self.db.batch_insert(entrys,table,DB_1)
def test_batch_get(self):
entrys =self.db.get_results_by_clms(columns='*',table=
'refined_info',database=DB_1,isdict=True)[:2]
for key, value in entrys[1].items():
print key,value,type(value)
class DMGTestCase(unittest.TestCase):
def setUp(self):
self.sb = submitter()
self.dd = data_drawer()
self.db = DBHelper()
self.samples =self.db.get_results_by_clms(columns='*',table=
'samples',database=DB_test,isdict=True)
def tearDown(self):
self.sb = None
self.dd = None
self.db = None
def test_submit_samples(self):
failed_ids = self.sb.submit_to_RB(self.samples)
self.assertTrue(len(self.samples) > 0)
print failed_ids
def test_submit_large_entrys(self):
entrys = self.db.get_results_by_clms(columns='*', table='refined_info',
database=DB_1, isdict=True)[:]
faild_ids = self.sb.submit_to_RB(entrys)
self.assertTrue(len(self.samples) > 0)
print failed_ids
def test_new_added_fields(self):
for entry in self.samples:
entry['last_mod_time'] = 10086
entry['group_id'] = 87217
entry['collegeName'] = 'ouyangming'
failed_ids = self.sb.submit_to_RB(self.samples)
def test_retrive(self):
entrys = self.dd.get_entrys(2000,is_full=False)
for entry in entrys:
#self.assertIn(entry['info_id'],info_ids)
for key,value in entry.items():
print key,value
print "###############################"
def test_submit_error_handle(self):
error_url = "http://www.panguso.com"
self.sb.table = 'samples'
self.sb.db = DB_test
self.sb.deal_submit_req(self.samples[:0],error_url)
def test_post_error(self):
posturl = P
class GroupTestCase(unittest.TestCase):
def setUp(self):
self.dc = datachecker()
self.db = DBHelper()
self.samples =self.db.get_results_by_clms(columns='*',table=
'samples',database=DB_test,isdict=True)
def tearDown(self):
self.dc = None
self.samples = None
def test_group_id(self):
table_new = 'samples_2'
table_old = 'info'
entrys = self.dc.diff_tables(table_old,table_new,DB_test,DB_test)
rm_field(INFO_ID,entrys)
datastore.update_info_table(entrys,table_old,DB_test)
class ExtractTestCase(unittest.TestCase):
def setUp(self):
self.db = DBHelper()
columns = ['recruit_title']
self.recruit_titles = self.db.get_results_by_clms(columns,'refined_info',DB_1)
def tearDown(self):
self.db = None
def test_extract_name(self):
prefix = '$#@&'
for recruit_title in self.recruit_titles[:]:
comp_name = extract.extract_compname(recruit_title[0])
if comp_name.startswith(prefix):
recruit_title = prefix + " " + recruit_title[0]
print "[Origin]:%s"%recruit_title
print "[CompNM]:%s"%comp_name
class prechecker:
def __init__(self):
self.filter = filter()
self.db_helper = DBHelper()
self.cmp_table = 'refined_list_info'
self.table = 'extracted_info'
self.cmp_clms = [COMPANY_NAME,MEETING_TIME,MEETING_LOCATION,ORIGIN_URL,RELEASE_DATE,RECRUIT_TITLE]
def rm_dup_list_info(self,tb_new,db_new,tb_old,db_old):
new_list = self.db_helper.get_results_by_clms("*",tb_new,db_new,True)
old_list = self.db_helper.get_results_by_clms("*",tb_old,db_old,True)
old_dict = {}
for entry in old_list:
url = entry.get(ORIGIN_URL)
if url is not None:
old_dict[url] = None
updates = []
for entry in new_list:
url = entry.get(ORIGIN_URL)
if url in old_dict:
continue
else:
updates.append(entry)
old_dict[url] = None
fields = entry.keys()
self.db_helper.batch_insert(updates,"refined_list_info",DB_1,fields)
return updates
def repair_data(self,entrys,cmp_entrys=None):
if cmp_entrys == None:
cmp_entrys = self.cmp_entrys
LOG.info("repairing Data...")
LOG.info("Entrys to Repair size is [%s],cmp_entrys size is [%s]"%(len(entrys),len(cmp_entrys)))
cmple_info_dict = collections.defaultdict(dict)
for entry in cmp_entrys:
origin_url = entry.get(ORIGIN_URL)
if origin_url != None:
cmple_info_dict[origin_url].update(entry)
for entry in entrys:
origin_url = entry.get(ORIGIN_URL)
if origin_url in cmple_info_dict:
for clm in cmple_info_dict[origin_url]:
value = entry.get(clm)
if value is None:
new_value = cmple_info_dict[origin_url][clm]
value = new_value
entry[clm] = value
return entrys
def pre_process(self):
self.rm_dup_list_info('extracted_list_info',DB_1,self.cmp_table,DB_1)
cmp_entrys = self.db_helper.get_results_by_clms(self.cmp_clms,self.cmp_table,DB_1,isdict=True)
entrys = self.db_helper.get_results_by_clms("*",self.table,DB_1,isdict=True)
entrys = self.repair_data(entrys,cmp_entrys)
entrys = self.filter.rm_college_from_loc(entrys)
self.db_helper.exe_sql('delete from %s.%s'%(DB_1,self.table))
if len(entrys) > 0 :
fields = entrys[0].keys()
self.db_helper.batch_insert(entrys,self.table,DB_1,fields)
from common.DataBaseHelper import DBHelper
import sys
reload(sys)
sys.setdefaultencoding('utf-8')
import re
import codecs
db_helper = DBHelper()
res = db_helper.get_results_by_clms('*','site_domain','xyzb_test')
def get_dict_from_file(file):
file = codecs.open(file,'r','utf-8')
list = []
rs = re.compile('[ ]+')
for line in file.readlines():
line = rs.split(line)
list.append(line)
file.close()
return list
def test_match():
file = open('domain_college','w')
for line in site_stats:
domain = line[0].strip()
domain_pat = re.compile(domain)
suc = False
for site_domain in site_domains:
if domain_pat.search(site_domain):
file.write("%s\t%s\n"%(domain,site_domains[site_domain]))
suc = True
break
entry[key] = value
entry_json = json.dumps(entry)
self.dumpjson_to_file(entry_json)
return res
def dumpjson_to_file(self, item):
# res = json.dumps(json_list)
res = str(item)
output_file = "json_" + datetime.datetime.now().strftime("%m_%d_%H%M")
output = os.path.join(self.output_dir, output_file)
try:
json_file = open(output, "a")
json_file.write(res)
json_file.write("\n")
json_file.close()
except IOError, e:
LOG.error(e)
sys.exit(-1)
finally:
if json_file:
json_file.close()
JD = JsonDumper(output_dir)
if __name__ == "__main__":
db_helper = DBHelper()
entrys = db_helper.get_results_by_clms(columns="*", table="info_2", database="xyzb", isdict=True)[:]
jd = JsonDumper(output_dir)
jd.dump_entrys(entrys)
import array
reload(sys)
sys.setdefaultencoding("utf-8")
host = "10.10.211.101"
port = 9090
dw = data_drawer()
#entrys = dw.get_entrys()
mdoc = None
res = []
db_helper = DBHelper()
entrys = db_helper.get_results_by_clms(columns="*",table='refined_info',isdict=True)[:]
city_id = load_city_map(CITY_FILE_PATH)
college_id = load_college_map(COLLEGE_FILE_PATH)
missed_colleges = set()
#using protobuff to serilize entrys
def serilize_entrys(entrys):
res = []
for entry in entrys:
mdoc = merged_doc_pb2.MergedDocInfo()
for key,value in entry.items():
# 空值的跳过
if value is None:
continue
class IndexTestCase(unittest.TestCase):
def setUp(self):
self.db = DBHelper()
self.samples =self.db.get_results_by_clms(columns='*',table=
'samples',database=DB_test,isdict=True)
self.segger = segger()
self.indexer = indexer()
self.dd = data_drawer()
def tearDown(self):
self.db = None
self.samples = None
self.segger = None
self.indexer = None
def test_convert_to_protoentry(self):
self.segger.serilize_entrys(self.samples)
def test_serilize_entrys(self):
entrys =self.db.get_results_by_clms(columns='*',table=
'refined_info',database=DB_1,isdict=True)
res = self.segger.serilize_entrys(entrys[:10])
res = self.segger.word_seg_list(res)
retdoc = merged_doc_pb2.MergedDocInfo()
for entry in res:
retdoc.ParseFromString(entry)
for i in range(0, len(retdoc.terms_info.term_infos)):
print 'term_sign:', retdoc.terms_info.term_infos[i].term_sign
print 'term_weight:',retdoc.terms_info.term_infos[i].term_weight
print '##########'
def test_record_protos(self):
file = open("proto.output",'wb')
res = self.segger.serilize_entrys(self.samples[3:10])
res = self.segger.word_seg_list(res)
print "length after segging res is ",len(res)
retdoc = merged_doc_pb2.MergedDocInfo()
for entry in res:
arr = array.array('I')
length = len(entry)
li = [length]
arr.fromlist(li)
file.write(arr.tostring())
file.write(entry)
file.close()
def test_serilize_field(self):
#entrys =self.db.get_results_by_clms(columns='*',table=
# 'info_tes',database=DB_2,isdict=True)
res = self.segger.serilize_entrys(self.samples)
res = self.segger.word_seg_list(res)
def test_word_seg_list(self):
protobufs = self.segger.serilize_entrys(self.samples[:2])
res = self.segger.word_seg_list(protobufs)
self.indexer.index_list_string(res)
def test_index_start_stop(self):
protobufs = self.segger.serilize_entrys(self.samples[:2])
#res = self.segger.word_seg_list(protobufs)
self.indexer.transport.open()
print self.indexer.start_index()
print self.indexer.stop_index()
self.indexer.transport.close()
#print self.indexer.index_list_string(res)
def test_index_list_string(self):
entrys =self.db.get_results_by_clms(columns='*',table=
'refined_info',database=DB_1,isdict=True)[:]
protobufs = self.segger.serilize_entrys(entrys)
res = self.segger.word_seg_list(protobufs)
self.indexer.index_list_string(res)
def test_split_city(self):
conv = TypeConvertor()
str= u"北京"
print str.strip().split(',')
print conv.city_id.get(str)
ids = conv.map_cities(str)
print ids
def test_0_entrys_index(self):
protobufs = self.segger.serilize_entrys([])
res = self.segger.word_seg_list(protobufs)
self.indexer.index_list_string(res)
def test_city_id(self):
conv = TypeConvertor()
print conv.city_id
def test_college_id(self):
conv = TypeConvertor()
print conv.map_college(u"清华大学")
self.assertEqual(1,conv.map_college(u"北京大学"))
class scheduler():
""" To schedule the data flow
"""
def __init__(self):
self.db = DBHelper()
self.dc = datachecker()
self.sb = submitter()
self.dd = data_drawer()
self.sg = segger()
self.id = indexer()
self.pc = prechecker()
self.dmg_on = DMG_ON
self.test_on = TEST_ON
if len(sys.argv) < 4:
self.table_old = "refined_info"
self.db_old = DB_1
self.table_new = "extracted_info"
self.db_new = DB_1
else:
self.table_old = sys.argv[1]
self.db_old = sys.argv[2]
self.table_new = sys.argv[3]
self.db_new = sys.argv[4]
def data_flow(self):
threads = []
self.pc.pre_process()
#remove duplicates
entrys = self.dc.diff_tables(self.table_old,self.table_new,self.db_old,self.db_new)
#insert the new coming entrys into database
self.db.batch_insert(entrys,self.table_old,self.db_old)
if (self.test_on):
t = threading.Thread(target=self.db.batch_insert,args=(entrys,self.table_old,DB_test))
t.start()
threads.append(t)
if(self.dmg_on):
#submite the updates to dmg
self.sb.deal_submit_req(entrys)
#get the all data from dmg
entrys = dd.get_entrys()
self.db.recr_table(TB_DMG,DB_1,TB_TEMPLATE,DB_1)
t = threading.Thread(target=self.db.batch_insert,args=(entrys,TB_DMG,DB_1))
t.start()
threads.append(t)
else:
entrys = self.db.get_results_by_clms(columns='*',table= self.table_old,
database=self.db_old,isdict=True)
entrys = sorted(entrys,key=self.sort_func,reverse=True)
#serilize entrys to protobuffers
entrys = self.sg.serilize_entrys(entrys)
#remote call word segging
entrys = self.sg.word_seg_list(entrys)
# #remote call indexing
self.id.index_list_string(entrys)
# wait until all threads are over
for t in threads:
t.join()
def sort_func(self,entry):
info_type = entry.get(INFO_TYPE)
if info_type == 1:
time = entry.get(MEETING_TIME)
else:
time = entry.get(RELEASE_DATE)
if time is None:
return 0
else :
return get_timestamp(time)
