Question

我有这个查询：

SELECT 
    `t2`.`upc`,
    FORMAT(SUM(`t2`.`qty`), 0) AS <COLUMN NAME>
FROM
    (`orders` `t1`
    JOIN `sales` `t2` ON ((`t2`.`order_id` = `t1`.`id`)))
WHERE <CONDITION>
GROUP BY `t2`.`upc`

与以下CONDITIONS配合使用的情况：

t1.`date_placed` > (CURDATE() - INTERVAL 42 DAY)
t1.order_status = 1

我试图合并这些单独的查询，如果不满足条件，则将其对应列的结果设为0。

这是我的尝试

SELECT 
    `t2`.`upc`,
    CASE 
    WHEN (`t1`.`date_placed` > (CURDATE() - INTERVAL 42 DAY)) 
    THEN FORMAT(SUM(`t2`.`qty`), 0)
    ELSE 0
    END AS `qty_sold`,
    CASE
    WHEN t1.order_status = 1
    THEN FORMAT(SUM(`t2`.`qty`, 0))
    ELSE 0
    END AS `order_qty`
FROM
    `orders` `t1`
    JOIN `sales` `t2` ON (`t2`.`order_id` = `t1`.`id`)
WHERE `upc`=1001 OR `upc`=1002 OR `upc`=1003
GROUP BY `t2`.`upc`;

订单表：

    +-----------+--------------+------------+
    |    id     | date_placed  |order_status|   
    +-----------+--------------+------------+
    |     1     |  2019-02-22  |      1     |
    +-----------+--------------+------------+
    |     2     |  2019-02-22  |      1     |
    +-----------+--------------+------------+
    |     3     |  2018-02-22  |      2     |
    +-----------+--------------+------------+

销售表：

    +-----------+--------------+------+-----+
    |    id     |   order_id   | upc  | qty |
    +-----------+--------------+------+-----+
    |     1     |      1       | 1001 |  1  |
    +-----------+--------------+------+-----+
    |     2     |      1       | 1002 |  1  |
    +-----------+--------------+------+-----+
    |     3     |      1       | 1003 |  1  |
    +-----------+--------------+------+-----+
    |     4     |      2       | 1001 |  3  |
    +-----------+--------------+------+-----+
    |     5     |      3       | 1001 |  2  |
    +-----------+--------------+------+-----+
    |     6     |      3       | 1002 |  1  |
    +-----------+--------------+------+-----+

预期结果：

    +-----------+--------------+------------+
    |    upc    |   qty_sold   | order_qty  |   
    +-----------+--------------+------------+
    |   1001    |      4       |      4     |
    +-----------+--------------+------------+
    |   1002    |      1       |      1     |
    +-----------+--------------+------------+
    |   1003    |      1       |      1     |
    +-----------+--------------+------------+

我在做什么错？如果无法尝试使用CASE语句，是否有另一种方式组合这些查询？

任何建议将不胜感激：）

Answer 1

case应该是聚合函数的参数：

FORMAT(SUM(CASE WHEN (`t1`.`date_placed` > (CURDATE() - INTERVAL 42 DAY)) THEN `t2`.`qty`ELSE 0
           END), 0) AS `qty_sold`,
FORMAT(SUM(CASE WHEN t1.order_status = 2  THEN t2`.`qty` ELSE 0 
           END), 0) `order_qty`

Answer 2

在案例表达式之前使用总和，并使用in

进行替换或条件化

SELECT 
        `t2`.`upc`,
        sum(CASE 
        WHEN (`t1`.`date_placed` > (CURDATE() - INTERVAL 42 DAY)) 
        THEN FORMAT((`t2`.`qty`), 0)
        ELSE 0
        END ) AS `qty_sold`,
        sum(CASE
        WHEN t1.order_status = 2 
        THEN FORMAT((`t2`.`qty`, 0))
        ELSE 0
        END) AS `order_qty`
    FROM
        `orders` `t1`
        JOIN `sales` `t2` ON (`t2`.`order_id` = `t1`.`id`)
    WHERE t2.upc in (1001,1002,1003) 
    GROUP BY `t2`.`upc`;

MySQL错误代码：1064将SELECT CASE与聚合函数一起使用时

2 个答案: