我正在创建一个MySQL数据库。我有两个不同的表:销售(id, idoftheproduct, quantity
)和供应(id, idoftheproduct, quantity
)。我想要一个触发器,无论何时我们添加新的销售或新的供应,它都会增加或减少另一个名为Stock(idoftheproduct, quantity
)的表的值
答案 0 :(得分:1)
总之,不要。而是创建一个view:
CREATE VIEW Stock
AS
SELECT IDofTheProduct, SUM(Quantity) AS Quantity
FROM ( SELECT IDofTheProduct, Quantity
FROM Supplies
UNION ALL
SELECT IdOfTheProduct, -Quantity
FROM Sales
) t
GROUP BY IDofTheProduct;
这样,当您的基础表更改视图时,您的视图将会发生变化。这在更新/插入期间的开销较小,并且总是准确的,即使触发器由于某种原因没有触发也是如此。
修改强>
道歉,我忘了MySQL不允许在视图中使用子查询。一种解决方案是创建一个新视图来代替子查询:
CREATE VIEW SalesAndSupplies
AS
SELECT 'Supplies' AS `Type`,
IDofTheProduct,
Quantity
FROM Supplies
UNION ALL
SELECT 'Sales' AS `Type`,
IDofTheProduct,
-Quantity
FROM Sales;
CREATE VIEW Stock
AS
SELECT IDofTheProduct, SUM(Quantity) AS Quantity
FROM SalesAndSupplies
GROUP BY IDofTheProduct;
<强> Example on SQL Fiddle 强>
答案 1 :(得分:0)
带有视图的解决方案非常棒,但随着时间的推移,它会变慢,特别是如果您有大量的销售和耗材数据,因为您总是计算它。
因此,如果您决定通过触发器实现它,那么您必须至少有四个触发器(两个表上都插入和删除)。由于所有这些都更新了库存,因此可以更好地将部分代码分解为存储过程。
更新库存的存储过程
CREATE PROCEDURE sp_update_stock(IN pid INT, IN qty DECIMAL(11, 3))
INSERT INTO stock (idoftheproduct, quantity)
VALUES (pid, qty)
ON DUPLICATE KEY UPDATE quantity = quantity + qty;
如果执行时不存在idoftheproduct
的行,则会创建该行。否则,它将被更新以反映更改。
现在触发
CREATE TRIGGER tg_sales_insert
AFTER INSERT ON sales
FOR EACH ROW
CALL sp_update_stock(NEW.idoftheproduct, -1 * NEW.quantity);
CREATE TRIGGER tg_supplies_insert
AFTER INSERT ON supplies
FOR EACH ROW
CALL sp_update_stock(NEW.idoftheproduct, NEW.quantity);
CREATE TRIGGER tg_sales_delete
AFTER DELETE ON sales
FOR EACH ROW
CALL sp_update_stock(OLD.idoftheproduct, OLD.quantity);
CREATE TRIGGER tg_supplies_delete
AFTER DELETE ON supplies
FOR EACH ROW
CALL sp_update_stock(OLD.idoftheproduct, -1 * OLD.quantity);
现在您可以在销售和供应中插入和删除。
INSERT INTO Supplies VALUES (NULL, 1, 100), (NULL, 1, 50), (NULL, 1, 75);
INSERT INTO Sales VALUES (NULL, 1, 2),(NULL, 1, 10), (NULL, 1, 5);
DELETE FROM Sales WHERE id = 1;
DELETE FROM Supplies WHERE id = 3;
如果看看stock
会看到
| IDOFTHEPRODUCT | QUANTITY | ----------------------------- | 1 | 135 |
这是 SQLFiddle 演示