Question

表格详细信息：

   CREATE TABLE `table1` (
    `user` VARCHAR(50) NULL DEFAULT NULL,
    `id` VARCHAR(50) NULL DEFAULT NULL
)
COLLATE='utf8mb4_0900_ai_ci'
ENGINE=InnoDB
;

INSERT INTO `table1` (`user`, `id`) VALUES ('bob', '1');
INSERT INTO `table1` (`user`, `id`) VALUES ('bob', '2');
INSERT INTO `table1` (`user`, `id`) VALUES ('kev', '3');
INSERT INTO `table1` (`user`, `id`) VALUES ('kev', '4');

CREATE TABLE `table2` (
    `id` VARCHAR(50) NULL DEFAULT NULL,
    `data1` VARCHAR(50) NULL DEFAULT NULL,
    `data2` VARCHAR(50) NULL DEFAULT NULL
)
COLLATE='utf8mb4_0900_ai_ci'
ENGINE=InnoDB
;

INSERT INTO `table2` (`id`, `data1`, `data2`) VALUES ('1', 'a', 'a1');
INSERT INTO `table2` (`id`, `data1`, `data2`) VALUES ('1', 'b', 'b1');
INSERT INTO `table2` (`id`, `data1`, `data2`) VALUES ('1', 'b', 'c1');
INSERT INTO `table2` (`id`, `data1`, `data2`) VALUES ('2', 'a', 'd1');
INSERT INTO `table2` (`id`, `data1`, `data2`) VALUES ('2', 'b', 'e1');
INSERT INTO `table2` (`id`, `data1`, `data2`) VALUES ('3', 'g', 'f1');

CREATE TABLE `table3` (
    `id` VARCHAR(50) NULL DEFAULT NULL,
    `data11` VARCHAR(50) NULL DEFAULT NULL,
    `data22` VARCHAR(50) NULL DEFAULT NULL
)
COLLATE='utf8mb4_0900_ai_ci'
ENGINE=InnoDB
;

INSERT INTO `table3` (`id`, `data11`, `data22`) VALUES ('1', 'a10', 'a11');
INSERT INTO `table3` (`id`, `data11`, `data22`) VALUES ('1', 'b10', 'b11');
INSERT INTO `table3` (`id`, `data11`, `data22`) VALUES ('1', 'c10', 'c11');
INSERT INTO `table3` (`id`, `data11`, `data22`) VALUES ('2', 'd10', 'd11');
INSERT INTO `table3` (`id`, `data11`, `data22`) VALUES ('2', 'e10', 'e11');
INSERT INTO `table3` (`id`, `data11`, `data22`) VALUES ('3', 'f10', 'f11');

我尝试过的查询是：

    select 
            t1.user,
            t1.id,
            group_concat(t2.data1 ) d1,
            group_concat( t2.data2) d2,
            group_concat( t3.data11 ) d11,
            group_concat(t3.data22 ) d22
        from table1 t1
        left join table2 t2 
                on   t2.id = t1.id
        left join table3 t3
                on t3.id = t1.id
                where t1.user='bob' group by t1.id;

但是我在d1，d2，d11，d22中获得了多个数据，并且由于第二次左连接

预期输出：

+-------+------+------------+-----------+---------------+---------------+
|  user |   id |     d1     |     d2    |       d11     |     d22       |
+-------+------+------------+-----------+---------------+---------------+
|  bob  |   1  |   a,b,b    | a1,b1,c1  |  a10,b10,c10  |  a11,b11,c11  |
|  bob  |   2  |   a,b      | d1,e1     |  d10,e10      |  d11,e11      |
+-------+------+------------+-----------+---------------+---------------+

实际上我是使用jdbc将数组数据从api存储到mysql，我已插入数据现在我必须在三个表中检索这些数据并形成数组格式并将其发送到Web api，所以这里有人知道如何执行此操作吗？我需要做一个嵌套查询吗？

Answer 1

您可以将派生表联接到首先进行聚合的子查询中。

SELECT t1.user,
       t1.id,
       t2.d1,
       t2.d2,
       t3.d11,
       t3.d22
       FROM table1 t1
            LEFT JOIN (SELECT t2.id,
                              group_concat(t2.data1) d1,
                              group_concat(t2.data2) d2
                              FROM table2 t2
                              GROUP BY t2.id) t2
                      ON t2.id = t1.id
            LEFT JOIN (SELECT t3.id,
                              group_concat(t3.data11) d11,
                              group_concat(t3.data22) d22
                              FROM table3 t3
                             GROUP BY t3.id) t3
                      ON t3.id = t1.id
      WHERE t1.user = 'bob';

通过三个表加入和组合concat

1 个答案: