no message

Pyspark_job/sqoop_import/ods_bs_category_asin_detail.py

@@ -18,6 +18,9 @@ if __name__ == '__main__':
     d2 = f'0{d2}' if int(d2) < 10 else f'{d2}'
     db_type = 'postgresql_14'
     import_tb = f"{site_name}_bs_category_asin_detail_month_{d1}_{d2}"
+    if date_type == 'day':
+        import_tb = f"{site_name}_bs_category_asin_detail_day_{date_info.replace('-', '_')}"
+
     query = f"""
         select 
             id, asin, null as week, best_sellers_rank, created_time as created_at, updated_time as updated_at, last_herf, all_best_sellers_href

Pyspark_job/sqoop_import/ods_keepa_finder_asin.py

@@ -4,6 +4,13 @@ import sys
 sys.path.append(os.path.dirname(sys.path[0]))
 
 from utils.secure_db_client import get_remote_engine
+from datetime import datetime, timedelta
+
+
+def is_saturday(date_str: str, fmt: str = "%Y-%m-%d") -> bool:
+    dt = datetime.strptime(date_str, fmt)
+    # weekday(): 周一=0 ... 周六=5 周日=6
+    return dt.weekday() == 5
 
 site_name = sys.argv[1]  # 参数1：站点
 date_type = sys.argv[2]  # 参数2：类型：day/week/4_week/month/quarter
@@ -24,12 +31,18 @@ partition_dict = {
 }
 sql_query = f"select id, request_id, asin, tracking_since from keepa_finder_asin where 1=1 " \
             f"and \$CONDITIONS"
-engine.sqoop_raw_import(
-        query=sql_query,
-        hive_table=hive_table,
-        partitions=partition_dict,
-        m=1,
-    )
 
+is_saturday_flag = is_saturday(date_info)
+
+if is_saturday_flag:
+    print(f"{date_info}是周六, 因此需要进行导入")
+    engine.sqoop_raw_import(
+            query=sql_query,
+            hive_table=hive_table,
+            partitions=partition_dict,
+            m=1,
+        )
+else:
+    print(f"{date_info}不是周六, 因此不需要进行导入")