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)
class statist():
def __init__(self):
self.load_dicts()
self.db = DBHelper()
def load_dicts(self):
#load domain_college_map
file_path = 'domain_college'
self.domain_college_dict = self.load_file_as_dict(file_path)
#load domain_college_map
file_path = 'sitename_college'
self.site_college_dict = self.load_file_as_dict(file_path)
#load college_list
file_path = 'college_list'
self.college_list = self.load_file_as_list(file_path)
def build_crawler_stat(self):
#load crawed_stats
file_path = 'crawler_stat'
rows = self.load_file_as_list(file_path)
heads = ['College','Request','Response','Suc_Rate','Uniq_Url']
entrys = []
rq_sum = 0
rp_sum = 0
uniq_url_sum = 0
for row in rows:
#TODO try catch the index exceed the row length
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)
entry.append(college)
entry.append(rq)
entry.append(rp)
#zero can not be divided
if rq == '0':
suc_rate = str(100.00) +"%"
else:
#keep two digits
suc_rate =str(round(float(rp)/float(rq)*100,2)) + "%"
entry.append(suc_rate)
entry.append(uniq_url)
entrys.append(entry)
if rq_sum == '0':
suc_rate_aver = str(100.00) + "%s"
else:
suc_rate_aver = str(round(float(rp_sum)/float(rq_sum)*100,2)) + "%s"
sum_entry = ['SUM',str(rq_sum),str(rp_sum),suc_rate_aver,str(uniq_url_sum)]
entrys.append(sum_entry)
return heads,entrys
def build_parse_stat(self):
# need parameter db,table and where clause
tb_full_today = 'extracted_full_info_today'
tb_today = 'refined_info_today'
sql_tmp = "select origin_website_name,count(*) from %s.%s %s group by origin_website_name"
dl = []
# get sitename and parsenum dict
sql = sql_tmp%(DB_1,tb_full_today,' ')
rows = self.db.exe_sql(sql)
dict_1 = dict(rows)
dl.append(dict_1)
#get sitename and error campus info num
where_clause = 'where info_type = 1 and isnull(meeting_time)'
sql = sql_tmp%(DB_1,tb_full_today,where_clause)
rows = self.db.exe_sql(sql)
dict_2 = dict(rows)
dl.append(dict_2)
# get sitename and error recruit info num
where_clause = 'where info_type = 0 and isnull(release_date)'
sql = sql_tmp%(DB_1,tb_full_today,where_clause)
rows = self.db.exe_sql(sql)
dict_3 = dict(rows)
dl.append(dict_3)
#merge the all dicts as a list based on key:sitename
rows = []
for k in dl[0].keys():
row = []
row.append(k)
for d in dl:
if d.get(k) is None:
row.append(0)
else:
row.append(d.get(k))
rows.append(row)
entrys = []
site_tt_sum = 0
err_1_sum = 0
err_2_sum = 0
for row in rows:
entry = []
#TODO exception
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)
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 = str(round((1 - float(err_1 + err_2) / float(site_tt)) * 100,2)) + "%"
entry.append(suc_rate)
entrys.append(entry)
# get sum info
suc_rate_tt = str(round((1 - float(err_1_sum + err_2_sum) / float(site_tt_sum)) * 100,2)) + "%"
sum_entry = ['SUM',site_tt_sum,err_1_sum,err_2_sum,suc_rate_tt]
entrys.append(sum_entry)
heads = ['College','Parse Info','Err Campus Info','Err Recruit Info','Suc_Rate']
return heads,entrys
def build_duplicates_stat(self):
# need parameter db,table and where clause
tb_full_today = 'extracted_full_info_today'
tb_today = 'refined_info_today'
sql_tmp = "select origin_website_name,count(*) from %s.%s %s group by origin_website_name"
dl = []
# get sitename and parsenum dict
sql = sql_tmp%(DB_1,tb_full_today,' ')
rows = self.db.exe_sql(sql)
dict_1 = dict(rows)
dl.append(dict_1)
#get sitename and error campus info num
sql = sql_tmp%(DB_1,tb_today,' ')
rows = self.db.exe_sql(sql)
dict_2 = dict(rows)
dl.append(dict_2)
#merge the all dicts as a list based on key:sitename
rows = []
for k in dl[0].keys():
row = []
row.append(k)
for d in dl:
if d.get(k) is None:
row.append(0)
else:
row.append(d.get(k))
rows.append(row)
entrys = []
full_sum = 0
refined_sum = 0
for row in rows:
entry = []
#TODO exception
site = row[0]
full_tt = row[1]
refined = row[2]
full_sum += full_tt
refined_sum += refined
college = self.site_college_dict.get(site)
entry.append(college)
entry.append(full_tt)
entry.append(refined)
# no need to prevent site_tt_num as zero :impossible
suc_rate = str(round(float(refined) / float(full_tt) * 100,2)) + "%"
entry.append(suc_rate)
entrys.append(entry)
# get sum info
suc_rate_tt = str(round(float(refined_sum) / float(full_sum) * 100,2)) + "%"
sum_entry = ['SUM',full_sum,refined_sum,suc_rate_tt]
entrys.append(sum_entry)
heads = ['College','Full Info','Refined Info','Uniqs_Prop']
return heads,entrys
def build_total_stat(self):
# need parameter db,table and where clause
tb_refined = 'refined_info'
sql_tmp = "select origin_website_name,count(*) from %s.%s %s group by origin_website_name"
dl = []
# get sitename and parsenum dict
sql = sql_tmp%(DB_1,tb_refined,' ')
rows = self.db.exe_sql(sql)
dict_1 = dict(rows)
dl.append(dict_1)
#get sitename and error campus info num
where_clause = "where info_type = 1"
sql = sql_tmp%(DB_1,tb_refined, where_clause)
rows = self.db.exe_sql(sql)
dict_2 = dict(rows)
dl.append(dict_2)
#get sitename and error campus info num
where_clause = "where info_type = 0"
sql = sql_tmp%(DB_1,tb_refined, where_clause)
rows = self.db.exe_sql(sql)
dict_3 = dict(rows)
dl.append(dict_3)
#get sitename and error campus info num
date_str = datetime.datetime.now().strftime('%Y-%m-%d')
where_clause = "where info_type = 1 and meeting_time > '%s'"%date_str
sql = sql_tmp%(DB_1,tb_refined, where_clause)
rows = self.db.exe_sql(sql)
dict_4 = dict(rows)
dl.append(dict_4)
#merge the all dicts as a list based on key:sitename
rows = []
for k in dl[0].keys():
row = []
row.append(k)
for d in dl:
if d.get(k) is None:
row.append(0)
else:
row.append(d.get(k))
rows.append(row)
entrys = []
tt_sum = 0
cam_tt_sum = 0
rec_tt_sum = 0
fut_cam_tt_sum = 0
for row in rows:
entry = []
#TODO exception
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)
entry.append(college)
entry.append(tt)
entry.append(cam_tt)
entry.append(rec_tt)
entry.append(fut_cam_tt)
entrys.append(entry)
# get sum info
sum_entry = ['SUM',tt_sum,cam_tt_sum,rec_tt_sum,fut_cam_tt_sum]
entrys.append(sum_entry)
heads = ['College','Total_Info','Campus_Info','Recruit_Info','Future_Info']
return heads,entrys
def to_dict(self,rows):
data = [(row[0][0],row[0][1]) for row in rows if len (row[0]) > 1 ]
return zip(data)
def load_file_as_dict(self,file_path):
map = {}
file = codecs.open(file_path,'r','utf-8')
for line in file.readlines():
line = line.split('\t')
key = line[0].strip()
value = line[1].strip()
map[key] = value
file.close()
return map
def load_file_as_list(self,file_path):
li = []
file = codecs.open(file_path,'r','utf-8')
for line in file.readlines():
li.append(line.strip().split('\t'))
file.close()
return li
