def get_yesterday_log_filename():
# medical01_20171123_info_logger.log
yesterday = get_yesterday_date()
dir_name = "/home/classify/workspace/material_recommendation/info_log/"
prefix1 = "medical01_"
prefix2 = "medical02_"
log_file_name1 = dir_name + prefix1 + yesterday + "_info_logger.log"
log_file_name2 = dir_name + prefix2 + yesterday + "_info_logger.log"
ana_file_name1 = log_file_name1 + "_analysis.csv"
ana_file_name2 = log_file_name2 + "_analysis.csv"
bdp_file_name1 = log_file_name1 + "_analysis.bdpinput"
bdp_file_name2 = log_file_name2 + "_analysis.bdpinput"
bdp_file_name = dir_name + yesterday + "_info_log.bdpinput"
return log_file_name1, ana_file_name1, bdp_file_name1, log_file_name2, ana_file_name2, bdp_file_name2, bdp_file_name
import os
os.environ['DJANGO_SETTINGS_MODULE'] = 'settings'
import sys
import json
import time
from general_utils.time_utils import get_yesterday_date
from recommend.manager.feed_data_helper import get_all_yesterday_user_id
from recommend.manager.recommend_resource import recommend_news_kernel
from add_data_to_solr.manager.user_topn_news import add_a_part
from recommend.daily_scripts.utils import get_parti_solr_filename, get_parti_uid_filename
cut_num = 4
yesterday_date = get_yesterday_date()
def save_uids(test=False):
'''
获取前一天活跃用户id，分成4份存在文件里(每份5000多个uid）
:return:
'''
all_valid_uids = get_all_yesterday_user_id(test=test)
all_valid_uids = [str(x) for x in all_valid_uids]
l = len(all_valid_uids)
cut_index = [x * l / cut_num for x in range(cut_num + 1)]
for i in range(cut_num):
file_name = get_parti_uid_filename(i, 'news')
def get_caled_user_topn_topics_yesterday(uid):
key = "%s|%s" % (get_yesterday_date(), uid)
return get_caled_user_topn_topics(key)