timestamp = get_timestamp(day_delta=1) else: timestamp = sys.argv[1] table_name = "raw_data_%s" % timestamp db = DBUtils() keywords = [] c = db.select(table='brand') for one in c.fetchall(): id,name,buss = one keywords.append(name) site_types = {} c = db.select(table='sites') for one in c.fetchall(): id,name,url,site_type = one site_types[url] = site_type print site_types.items() cursor = db.select(table=table_name,columns=["count(*)"]) (total,) = cursor.fetchone() DELTA = 10000 for start in range(0,total,DELTA): thread.start_new_thread(filter_data,(db,table_name,start,DELTA,timestamp,keywords,site_types)) db.close() print start_time," ",datetime.now()