ABA主题标签优化

Pyspark_job/sqoop_export/dwt_st_theme_agg.py

 import os
 import sys
-import random
-import string
 
 sys.path.append(os.path.dirname(sys.path[0]))
+
 from utils.db_util import DBUtil
 from utils.ssh_util import SSHUtil
-from utils.common_util import CommonUtil,DateTypes
+from utils.common_util import CommonUtil
 
 if __name__ == '__main__':
     site_name = CommonUtil.get_sys_arg(1, None)
     date_type = CommonUtil.get_sys_arg(2, None)
     date_info = CommonUtil.get_sys_arg(3, None)
 
-
     #  获取最后一个参数
     test_flag = CommonUtil.get_sys_arg(len(sys.argv) - 1, None)
     print(f"执行参数为{sys.argv}")
@@ -22,8 +20,6 @@ if __name__ == '__main__':
         db_type = 'postgresql_test'
         print("导出到测试库中")
     else:
-        # db_type = "postgresql"
-        # print("导出到PG库中")
         db_type = "postgresql_cluster"
         print("导出到PG集群库库中")
 
@@ -39,12 +35,11 @@ if __name__ == '__main__':
 
     # 保证幂等性，先删除原始表同周期的数据
     sql = f"""
-             drop table if exists {export_tb};
-             create table if not exists {export_tb} 
-                    (
-                    like {export_master_tb}  including ALL 
-             );
-        """
+        drop table if exists {export_tb};
+        create table if not exists {export_tb} (
+            like {export_master_tb}  including ALL 
+        );
+    """
     DBUtil.engine_exec_sql(engine, sql)
 
     # 导出agg表
@@ -67,7 +62,13 @@ if __name__ == '__main__':
             "pattern_num_rate",
             "date_info",
             "created_time",
-            "updated_time"
+            "updated_time",
+            "pattern_word_num",
+            "pattern_category_id",
+            "pattern_category_current_id",
+            "pattern_rank",
+            "pattern_rank_change_rate",
+            "pattern_rank_rate_of_change"
         ],
         partition_dict={
             "site_name": site_name,