所有
我正在尝试创建具有两个输入的递归查询:
我有三个工作子查询:
我发现可以在简单的SQL中执行此操作,例如查询:
WITH RECURSIVE x2 (result) AS (
SELECT 1
UNION ALL
SELECT result*2 FROM x2)
SELECT * FROM x2 LIMIT 10;
当我使用2个输入 ITEM_ID + LOT 运行第一个子查询 RECEIVING 时,它会返回一些行,然后就OK了,SQL应该停止。 如果没有,查询应开始在子查询 BATCH 中搜索 BATCH编号,此批次编号用于第三个子查询 INGREDIENT 为 RECEIVING 返回新ITEM_ID + LOT 。
Diagram for recursive subquery
示例:
INPUT
LOT = 1EP17171590
输入输入ITEM_ID(1765716)并输入LOT(1EP17171590) 子查询 RECEIVING =>没有行返回(如果返回,结束SQL)
帕特里克,
我做了这个查询,它的工作但速度太慢,不知道如何优化它?
select
BATCH_ID,
BATCH_NO,
PRODUCT_ITEM_ID,
PRODUCT_ITEM
PRODUCT_LOT,
INGREDIENT_ITEM_ID,
INGREDIENT_ITEM
INGREDIENT_LOT,
LEVEL
FROM
(
SELECT
b.batch_id,
b.batch_no,
b.inventory_item_id as PRODUCT_ITEM_ID,
b.item as PRODUCT_ITEM,
b.lot as PRODUCT_LOT,
i.inventory_item_id AS INGREDIENT_ITEM_ID,
i.item as INGREDIENT_ITEM,
i.lot as INGREDIENT_LOT
FROM batch b
JOIN ingredients i
ON i.batch_id = b.batch_id
)
CONNECT BY NOCYCLE PRIOR
PRODUCT_ITEM_ID = INGREDIENT_ITEM_ID
AND PRODUCT_LOT = INGREDIENT_LOT
START WITH
PRODUCT_ITEM_ID = 1765716
AND PRODUCT_LOT = '1EP17171590'
;
答案 0 :(得分:0)
你可以这样做而不需要递归:
我使用您在屏幕截图中提供的示例数据创建了用于此目的的表:
create table RECIEVING
(
ITEM_ID number,
LOT varchar2(50)
);
insert into RECIEVING
values(1736957,'1FP17068674');
create table BATCH
(
BATCH number,
ITEM_ID number,
LOT varchar2(50)
);
insert into BATCH
values(351908,1765716,'1EP17171590');
create table INGREDIENTS
(
BATCH number,
ITEM_ID number,
LOT varchar2(50)
);
insert into INGREDIENTS
values(351908,1736957,'1FP17068674');
这是我使用的查询:
select RECIEVING.ITEM_ID,
RECIEVING.LOT
from BATCH
join INGREDIENTS
on BATCH.BATCH = INGREDIENTS.BATCH
join RECIEVING
on RECIEVING.ITEM_ID = INGREDIENTS.ITEM_ID
and RECIEVING.LOT = INGREDIENTS.LOT
where BATCH.ITEM_ID = 1765716
and BATCH.LOT = '1EP17171590';
如果一个批次有多种成分,那么只要它在RECIEVING中存在就会返回超过1行。我不确定这是否符合您的要求。
您提供的示例的输出如下所示:
ITEM_ID LOT
1736957 1FP17068674