update

task/daily_search_word_count.py

@@ -78,7 +78,7 @@ for t in range(0, task_days):
     SELECT query,
            partition_date,
            all_search_uv as all_search_uv, --全部搜索uv
-           t3.all_search_pv as all_search_pv --全部搜索pv
+           all_search_pv as all_search_pv --全部搜索pv
     FROM (
              --搜索pvuv
              SELECT query
@@ -181,14 +181,12 @@ for t in range(0, task_days):
              GROUP BY query,partition_date
          ) t3 order by all_search_uv desc
     
-    """.format(start_date=today_str, end_date=yesterday_str)
+    """.format(start_date=yesterday_str, end_date=today_str)
 
     print(sql_search_ctr)
     search_ctr_df = spark.sql(sql_search_ctr)
-    # spam_pv_df.createOrReplaceTempView("dev_view")
     search_ctr_df.show(1)
     sql_res = search_ctr_df.collect()
-    # date_str = yesterday_date.strftime("%Y%m%d")
     tag_names_list_week = []
     for name in sql_res:
         word = name.get("query")
@@ -208,9 +206,9 @@ for t in range(0, task_days):
         pid = hashlib.md5((partition_date + word).encode("utf8")).hexdigest()
         insert_sql_tuple = (word, nums, uv, pid, date_str)
         insert_list.append(insert_sql_tuple)
-        # print(insert_sql)
+        print(insert_sql)
         if count % 100 == 0:
-            # cursor.execute("set names 'UTF8'")
+            cursor.execute("set names 'UTF8'")
             res = cursor.executemany(insert_sql, insert_list)
             db.commit()
             # print(res)