def deal_data(self,raw_data):
entrys = []
site_tt_sum = 0
err_1_sum = 0
err_2_sum = 0
for row in raw_data:
try:
entry = []
site = row[0]
site_tt = row[1]
err_1 = row[2]
err_2 = row[3]
site_tt_sum += site_tt
err_1_sum += err_1
err_2_sum += err_2
college = self.site_college_dict.get(site)
if college is None:
LOG.error('[%s] name not found!'%site)
college = site
entry.append(college)
entry.append(site_tt)
entry.append(err_1)
entry.append(err_2)
# no need to prevent site_tt_num as zero :impossible
suc_rate = self.get_rate(site_tt - err_1 - err_2,site_tt)
entry.append(suc_rate)
entrys.append(entry)
except Exception,e:
LOG.error(e)
def deal_data(self,raw_data):
entrys = []
full_sum = 0
refined_sum = 0
for row in raw_data:
try:
entry = []
site = row[0]
full_tt = row[1]
refined = row[2]
full_sum += full_tt
refined_sum += refined
college = self.site_college_dict.get(site)
if college is None:
LOG.error('[%s] name not found!'%site)
college = site
entry.append(college)
entry.append(full_tt)
entry.append(refined)
# no need to prevent site_tt_num as zero :impossible
suc_rate = self.get_rate(refined,full_tt)
entry.append(suc_rate)
entrys.append(entry)
except Exception,e:
LOG.error(e)
def insert_to_campusTalk(entrys):
if len(entrys) == 0:
return
LOG.info("Startging to Store to multi tables")
table = T_campusTalk
id = None
count = 0
for entry in entrys:
if count % 2000 == 0:
LOG.info("[%s] Has been inserted!"%count)
collegeID = insert_subtable(entry,T_collegeInfo)
if collegeID is None:
LOG.error("INSERT TO [CollegeInfo] FAILS:%s"\
%("|".join(entry.values())))
return None
campanyID = insert_subtable(entry,T_companyInfo)
if campanyID is None:
LOG.error("INSERT TO [CampanyInfo] FAILS:%s"\
%("|".join(entry.values())))
return None
new_entry = {}
new_entry[COLLEGE_ID] = collegeID
new_entry[COMPANY_ID] = campanyID
for key,value in entry.items():
if value is None:
continue
if ct_field_map[key][0] == table:
new_entry[ct_field_map[key][1]] = value
if len(new_entry.keys()) > 0:
id = db_helper.insert_entry(new_entry,table)
count += 1
return id
def load_city_map(filepath):
city_id_map = {}
lines = []
try:
file = codecs.open(filepath,encoding='utf-8')
lines = file.readlines()
except Exception as e:
LOG.error("Doesn't find the city list file!")
LOG.error(e)
if len(lines) == 0:
return city_id_map
for line in lines:
line = line.strip("\n")
names = []
#编号，城市名，别名
(id,name,alias)= line.split("\t")
id = int(id)
names.append(name)
alias = alias.split(";")
for name in alias:
names.append(name)
for name in names:
name = name.strip('\"')
if len(name.strip()) == 0:
continue
city_id_map.update({name:id})
return city_id_map
def deal_data(self,raw_data):
entrys = []
tt_sum = 0
cam_tt_sum = 0
rec_tt_sum = 0
fut_cam_tt_sum = 0
for row in raw_data:
try:
entry = []
site = row[0]
tt = row[1]
cam_tt = row[2]
rec_tt = row[3]
fut_cam_tt = row[4]
tt_sum += tt
cam_tt_sum += cam_tt
rec_tt_sum += rec_tt
fut_cam_tt_sum += fut_cam_tt
college = self.site_college_dict.get(site)
if college is None:
LOG.error('[%s] name not found!'%site)
college = site
entry.append(college)
entry.append(tt)
entry.append(cam_tt)
entry.append(rec_tt)
entry.append(fut_cam_tt)
entrys.append(entry)
except Exception,e:
LOG.error(e)
def load_ids_from_rec(self):
file = None
info_ids = []
try:
file = open(FAILED_INFOID_REC,'r')
info_ids_str = file.read()
info_ids = info_ids_str.split(";")
except Exception as e:
LOG.error(e)
finally:
if file:file.close()
return info_ids
def map_job_type(self, value):
value = value.strip()
job_type_id = 0
if len(value) > 0:
if value == "兼职":
job_type_id = 1
elif value == "实习":
job_type_id = 2
elif value == "全职":
job_type_id = 0
else:
LOG.error("The Job Type Value is not valid,[%s]" % (value))
return job_type_id
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)
def send_request(self,retry_time=3):
count = 0
res = None
resp = None
while(count < retry_time):
LOG.info("Send Request Round: [%s]" %(count))
try:
resp = urllib2.urlopen(self.url)
res = resp.read()
break
except Exception,e:
LOG.error(e)
time.sleep(1)
count += 1
def dmg_to_proto_entry(self, entry):
new_entry = {}
convert_int_keys = set([INFO_ID, GROUP_ID, HAS_HUKOU, HAS_EXAM, HAS_RESUME, INFO_TYPE, CLICK_RATE])
convert_time_keys = set(
[MEETING_TIME, RELEASE_DATE, RESUME_START_DATE, RESUME_END_DATE, EXAM_TIME, INTERVIEW_TIME, LAST_MOD_TIME]
)
for key, value in entry.items():
if value is None:
continue
if isinstance(value, basestring):
if value.strip() == "":
continue
# convert to int count:9
if key in convert_int_keys:
try:
value = int(value)
except ValueError:
LOG.error("ValueError,key is [%s] value is [%s]" % (key, value))
# convert to id count:2
elif key == COLLEGE_NAME:
college_id = self.map_college(value)
new_entry[COLLEGE_ID] = college_id
elif key == RECRUIT_CITIES:
try:
value = self.map_cities(value)
except:
LOG.error("In Map cities[%s]:%s" % (key, value))
# convert to type:
if key == JOB_TYPE:
value = self.map_job_type(value)
if key == COMPANY_TYPE:
value = self.map_company_type(value)
# convert to list count:1
elif key == WORK_PLACE:
value = self.deal_work_place(value)
# convert time to timestamp count:6
elif key in convert_time_keys:
try:
if not isinstance(value, datetime.datetime):
value = datetime.datetime.strptime(value, DATEFORMAT)
value = get_timestamp(value)
except Exception, e:
LOG.error("[%s]:[%s]" % (key, e))
# last_mod_time has micro-second scale
elif key == LAST_MOD_TIME:
try:
value = datetime.datetime.strptime(value, DATEFORMAT)
value = get_timestamp(value)
except Exception, e:
LOG.error("[%s]:[%s]" % (key, e))
def word_seg_list(self,protobufs):
res = []
if not check_list(protobufs):
LOG.info("Do Not Call Word Segging,For the input list is:%s"%(protobufs))
return res
try:
self.transport.open()
LOG.info("Begin RPC Word Segging,[%s] To Be Segged!"
%(len(protobufs)))
res = self.client.word_seg_list(protobufs)
self.transport.close()
LOG.info("Finish RPC Word Segging,[%s] Entrys Have Been Segged!"
%(len(res)))
self.rec_protobuf(res)
except Exception,e:
LOG.error(e)
def load_table(filepath=None):
try:
domtree = ElementTree.parse(filepath)
root = domtree.getroot()
tables = root.findall('table')
table_info = {}
for table in tables:
columns = []
table_name = table.find('name').text
columns_list = table.find('columns').findall('column')
for column in columns_list:
columns.append(column.text)
table_info[table_name] = columns
except Exception,e:
LOG.error(e)
sys.exit(-1)
def index_list_string(self, protobufs):
if not check_list(protobufs):
LOG.info("Do Not Call Indexing, For the input list is:%s" % (protobufs))
return
count = 0
try:
self.transport.open()
is_ready = self.start_index()
if not is_ready:
LOG.error("Index Server Is Not Ready!")
return 0
count = self.client.put_list_string(protobufs)
LOG.info("[%s] Entrys Has Been Successfully Indexed." % (count))
self.stop_index()
self.transport.close()
except Exception, e:
LOG.error(e)
def serilize_entrys(self,entrys):
protobufs = []
LOG.info("Begin Serilizing Entrys,[%s] Entrys To Be Serilized!"
%(len(entrys)))
for entry in entrys:
mdoc = merged_doc_pb2.MergedDocInfo()
new_entry = self.convertor.dmg_to_proto_entry(entry)
for key,value in new_entry.items():
try:
if isinstance(value,list):
getattr(mdoc,key).extend(value)
else:
setattr(mdoc,key,value)
except Exception,e:
LOG.error("[%s]:%s" % (key,value))
LOG.error(e)
protobuf = mdoc.SerializeToString()
protobufs.append(protobuf)
def load_columns(self, filepath=None):
columns_list = []
try:
dom_tree = ElementTree.parse(filepath)
root = dom_tree.getroot()
columns = root.findall("field")
if columns is None or len(columns) == 0:
LOG.error("No columns found in xml conf [%s]" % (filepath))
sys.exit(-1)
columns_count = 0
for column in columns:
columns_list.append(column.text)
columns_count += 1
LOG.info("Total Load [%d] columns in conf [%s]" % (columns_count, filepath))
except Exception as e:
LOG.error(e)
sys.exit(-1)
return columns_list
def write_ids_to_rec(self):
LOG.info("Recording Failed IDs Into File [%s]" %(FAILED_INFOID_REC))
if not (len(self.failed_ids)>0):
return
# record the failed_ids to file
# check the record file dir if exists ,if not create it
dir= os.path.dirname(FAILED_INFOID_REC)
file = None
if not os.path.isdir(dir):
os.mkdir(dir)
try:
file = open(FAILED_INFOID_REC,'w')
file.write(";".join([str(id) for id in self.failed_ids]))
LOG.info("Record Successfully Totally [%s] entrys"
%(len(self.failed_ids)))
except Exception as e:
LOG.error(e)
finally:
if file: file.close()
def submit_to_RB(self,entrys,posturl=POSTURL):
""" submit data to dmg
"""
failed_count = 0
suc_count = 0
suc_ids = []
failed_ids = []
failed_entrys = []
count = 0
state = ""
res = None
for entry in entrys[:]:
#just for test should delet after testing
count += 1
info_id = entry[INFO_ID]
entry_json = self.construct_json_data(entry)
resjson = {"resjson":json.dumps(entry_json)}
try:
state =self.post(posturl,resjson)
res = json.loads(state)['result']
if (res == 0):
LOG.warning("[%s] Submitted Error!" %(info_id))
LOG.warning(state)
failed_count += 1
if (res == 1): suc_count += 1
except Exception as e:
LOG.error(e)
failed_count += 1
LOG.error(state)
failed_ids.append(info_id)
failed_entrys.append(entry)
# LOG.debug("Post one entry into RB the result is: [%s],Suc_Count=\
# [%s],Fail_count=[%s]" % (res,suc_count,failed_count))
if (suc_count + 1) % 1000 == 0 or (failed_count + 1 ) % 100 == 0:
LOG.info("Post Entrys To DMG,Suc:[%s] Failed:[%s]"
%(suc_count,failed_count))
self.failed_ids = failed_ids
self.failed_entrys = failed_entrys
LOG.info("Successfully Submitted To DMG [%s],Failed:[%s]"
%(suc_count,failed_count))
def deal_data(self, raw_data):
entrys = []
rq_sum = 0
rp_sum = 0
uniq_url_sum = 0
rows = self.prepare_data()
for row in raw_data:
# TODO try catch the index exceed the row length
try:
entry = []
dm = row[0]
rq = row[1]
rp = row[2]
uniq_url = row[3]
rq_sum += int(rq)
rp_sum += int(rp)
uniq_url_sum += int(uniq_url)
college = self.domain_college_dict.get(dm)
if college is None:
LOG.error("[%s] domain not found" % dm)
college = dm
entry.append(college)
entry.append(rq)
entry.append(rp)
# zero can not be divided
suc_rate = self.get_rate(rp, rq)
entry.append(suc_rate)
entry.append(uniq_url)
entrys.append(entry)
except Exception, e:
LOG.error(e)
suc_rate_aver = self.get_rate(rp_sum, rq_sum)
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
# 数据库读出的long值改为int，info_id字段除外
if isinstance(value,long):
if(key != INFO_ID):
value = int(value)
# 时间转化成时间戳int类型
if isinstance(value,datetime.datetime):
value = get_timestamp(value)
# 对recruit_cities和编号进行映射
if key == RECRUIT_CITIES:
for city in value.strip().split(","):
if len(city) > 0:
id = city_id.get(value)
if id is None:
LOG.error("City Name [%s] Not Found!"%value)
else:
mdoc.recruit_cities.append(id)
continue
# 对collegeName和编号进行映射
if key == COLLEGE_NAME:
value = value.strip()
if len(value) > 0:
id = college_id.get(value)
if id is None:
LOG.error("College Name [%s] Not Found!"%value)
missed_colleges.add(value)
#默认为0
mdoc.college_ID = 1
else:
#填充protobuffer中的college_id 字段
mdoc.college_ID = id
#处理work_place值
if key == 'work_place':
for place in value.strip().split(","):
if len(place) > 0:
mdoc.work_place.append(place)
LOG.info(";".join(mdoc.work_place))
continue
try:
#计算term-weight需要gbk编码的value
if isinstance(value,str):
value = value.decode('utf-8').encode('gbk')
setattr(mdoc, key, value)
mdoc.origin_url_sign = 0
mdoc.recruit_title_sign = 0
mdoc.info_text_sign = 0
mdoc.group_id = 1
except Exception,e:
LOG.error(e)
LOG.error("key is [%s] and value is[%s]"%(key,value))
serilized = mdoc.SerializeToString()
res.append(serilized)
except Exception, e:
LOG.error("[%s]:[%s]" % (key, e))
# last_mod_time has micro-second scale
elif key == LAST_MOD_TIME:
try:
value = datetime.datetime.strptime(value, DATEFORMAT)
value = get_timestamp(value)
except Exception, e:
LOG.error("[%s]:[%s]" % (key, e))
# else convert to string
else:
try:
value = value
except Exception as e:
LOG.error("[s%]:[%s]" % (key, e))
new_entry[key] = value
# set the default key and value
# new_entry['normalized_url_sign'] = 0
# new_entry['recruit_title_sign'] = 0
# new_entry['info_text_sign'] = 0
return new_entry
# convert cities to cityid_list
def map_cities(self, value):
city_ids = []
for city in value.strip().split(","):
if len(city) > 0:
