相关子查询的外部查询有多个结果。

Question

我想知道是否可以让内层查询返回多个结果，再由外层查询来查询？

下面是一个可复制的例子。

在这个例子中，我想知道每个部门工资最高的员工的名字和工资。

CREATE TABLE Worker (
    WORKER_ID INT NOT NULL PRIMARY KEY AUTO_INCREMENT,
    FIRST_NAME CHAR(25),
    LAST_NAME CHAR(25),
    SALARY INT(15),
    JOINING_DATE DATETIME,
    DEPARTMENT CHAR(25)
);

INSERT INTO Worker 
    (WORKER_ID, FIRST_NAME, LAST_NAME, SALARY, JOINING_DATE, DEPARTMENT) VALUES
        (001, 'Monika', 'Arora', 100000, '14-02-20 09.00.00', 'HR'),
        (002, 'Niharika', 'Verma', 80000, '14-06-11 09.00.00', 'Admin'),
        (003, 'Vishal', 'Singhal', 300000, '14-02-20 09.00.00', 'HR'),
        (004, 'Amitabh', 'Singh', 500000, '14-02-20 09.00.00', 'Admin'),
        (005, 'Vivek', 'Bhati', 500000, '14-06-11 09.00.00', 'Admin'),
        (006, 'Vipul', 'Diwan', 200000, '14-06-11 09.00.00', 'Account'),
        (007, 'Satish', 'Kumar', 75000, '14-01-20 09.00.00', 'Account'),
        (008, 'Geetika', 'Chauhan', 90000, '14-04-11 09.00.00', 'Admin');

找出每个部门的最高工资是很容易的。

SELECT MAX(SALARY), DEPARTMENT
FROM Worker
GROUP BY DEPARTMENT;

但我如何把赚取这些薪水的员工的名字也添加到结果中？

非常感谢

Answer 1

你可以在查询中使用操作符 IN:

SELECT * FROM Worker
WHERE (DEPARTMENT, SALARY) IN (
  SELECT DEPARTMENT, MAX(SALARY) 
  FROM Worker
  GROUP BY DEPARTMENT 
);

或者，如果你使用MySql 8.0+，并带有 RANK() 窗口功能。

SELECT w.* 
FROM (
  SELECT *, RANK() OVER (PARTITION BY DEPARTMENT ORDER BY SALARY DESC) rnk
  FROM Worker
) w
WHERE w.rnk = 1

参见 演示. 结果。

> WORKER_ID | FIRST_NAME | LAST_NAME | SALARY | JOINING_DATE        | DEPARTMENT
> --------: | :--------- | :-------- | -----: | :------------------ | :---------
>         3 | Vishal     | Singhal   | 300000 | 2014-02-20 09:00:00 | HR        
>         4 | Amitabh    | Singh     | 500000 | 2014-02-20 09:00:00 | Admin     
>         5 | Vivek      | Bhati     | 500000 | 2014-06-11 09:00:00 | Admin     
>         6 | Vipul      | Diwan     | 200000 | 2014-06-11 09:00:00 | Account