update

task/spark_test.py

@@ -91,11 +91,15 @@ for t in range(2, task_days):
     yesterday_str_format = (now + datetime.timedelta(days=-1)).strftime("%Y-%m-%d")
     one_week_age_str = (now + datetime.timedelta(days=-7)).strftime("%Y%m%d")
     new_urser_device_id_sql = r"""
-         select cl_type,app_version,count(app_version) as count_num from 
-         ( select *,rank() over(partition by cl_id order by partition_date desc) as ranking from
-         (select cl_type,app_version,cl_id,partition_date from online.bl_hdfs_maidian_updates where partition_date >= "20200201" group by cl_type,app_version,cl_id,partition_date) 
-         ) where ranking =1 group by cl_type,app_version
-    """.format(start_date="20210101", end_date="20200222", today_str_format=today_str_format,tomorrow_str=tomorrow_str)
+SELECT *
+                         FROM online.bl_hdfs_maidian_updates
+                         WHERE partition_date >= '{partition_day}'
+                           AND partition_date < '{end_date}'
+                           AND ((action in ('on_click_topic_card','on_click_diary_card','search_result_click_infomation_item')
+                               AND page_name in ('search_result_more','search_result_diary','search_result_post'))
+                               or (action = 'on_click_card' AND params['card_content_type'] in ('answer','diary') AND page_name in ('search_result_more','search_result_diary','search_result_question_answer')))
+                        limit 200
+    """.format(partition_day="20210101", end_date="20200102",tomorrow_str=tomorrow_str)
 
   #   print(new_urser_device_id_sql)
     new_urser_device_id_df = spark.sql(new_urser_device_id_sql)