多表查询，也称为关联查询，指两个或更多个表一起完成查询操作。

• 如果查询语句中出现了多个表中都存在的字段，则必须指明此字段所在的表。
  • SELECT tb_a.id, tb_b.id FROM tb_a,tb_b WHERE tb_a.id=tb_b.id;
• 可以给表起别名，在SELECT和WHERE中使用表的别名
  • SELECT t1.id, t2.id FROM tb_a as t1,tb_b as t2 WHERE t1.id=t2.id;
• 如果给表起了别名，一旦在SELECT或WHERE中使用表名的话，则必须使用表的别名，而不是使用表的原名。
• 如果有n个表实现多表的查询，则需要至少n-1个连接条件

1. 多表查询示例

新建表：

SELECT e.employee_id as '员工id',e.last_name as '员工名称',d.department_name as '员工所在部门',l.city as '部门所在城市'
FROM employees e, departments d, locations l
WHERE e.department_id=d.department_idAND d.location_id=l.location_id;

结果：

员工id	员工名称	员工所在部门	部门所在城市
200	Whalen	Administration	Seattle
201	Hartstein	Marketing	Toronto
202	Fay	Marketing	Toronto
114	Raphaely	Purchasing	Seattle
115	Khoo	Purchasing	Seattle

2. 多表查询分类

2.1 等/非等值连接

2.1.1 等值连接

SELECT employees.employee_id as '员工id',employees.last_name as '员工名称',employees.department_id as '部门id',departments.department_id as '部门id',departments.department_name as '部门名称',departments.location_id as '位置id'
FROM employees, departments
WHERE employees.department_id = departments.department_id;

员工id	员工名称	部门id	部门id	部门名称	位置id
200	Whalen	10	10	Administration	1700
201	Hartstein	20	20	Marketing	1800
202	Fay	20	20	Marketing	1800
114	Raphaely	30	30	Purchasing	1700
…

2.1.2非等值连接

SELECT e.last_name as '员工名称',e.salary as '员工薪资',j.grade_level as '员工级别'
FROM employees e, job_grades j
WHERE e.salary BETWEEN j.lowest_sal AND j.highest_sal;
-- 或者
WHERE e.salary >= j.lowest_sal AND e.salary <= j.highest_sal;

员工名称	员工工资	员工级别
King	24000	E
Kochhar	17000	E
De Haan	17000	E
Hunold	9000	C
Ernst	6000	C
Austin	4800	B

2.2 自然/非自然连接

SELECT emp.employee_id as '员工id', emp.last_name as '员工名称', mgr.employee_id as '管理者id', mgr.last_name as '管理者名称'
FROM employees emp, employees mgr
WHERE emp.manager_id = mgr.employee_id;
或者
SELECT CONCAT(worker.last_name ,' works for ' , manager.last_name)
FROM employees worker, employees manager
WHERE worker.manager_id = manager.employee_id ;

员工id	员工名称	管理者id	管理者名称
101	Kochhar	100	King
102	De Haan	100	King
103	Hunold	102	De Haan
104	Ernst	103	Hunold
105	Austin	103	Hunold
106	Pataballa	103	Hunold

2.3 内/外连接

• 内连接: 合并具有同一列的两个以上的表的行, 结果集中不包含一个表与另一个表不匹配的行
• 外连接: 两个表在连接过程中除了返回满足连接条件的行以外**还返回左（或右）表中不满足条件的行，这种连接称为左（或右）外连接。**没有匹配的行时, 结果表中相应的列为空(NULL)。
• 如果是左外连接，则连接条件中左边的表也称为主表，右边的表称为从表。
• 如果是右外连接，则连接条件中右边的表也称为主表，左边的表称为从表。

2.3.1 内连接

SELECT e.employee_id, d.department_id, l.city
FROM employees eINNER JOIN departments dON e.department_id = d.department_idINNER JOIN locations lON d.location_id = l.location_id;-- 返回 employees_id和department_id都不为NULL的数据

2.3.2 外连接

-- 左外连接
SELECT e.employee_id, d.department_id
FROM employees eLEFT OUTER JOIN departments dON e.department_id = d.department_id;-- 右外连接SELECT e.employee_id, d.department_id
FROM employees eRIGHT OUTER JOIN departments dON e.department_id = d.department_id;

3.UNION的使用

3.1 合并查询结果

利用UNION关键字，可以给出多条SELECT语句，并将它们的结果组合成单个结果集。合并时两个表对应的列数和数据类型必须相同，并且相互对应，各个SELECT语句之间使用UNION或UNION ALL关键字分隔。
语法格式：

SELECT column,... FROM table1
UNION [ALL]
SELECT column,... FROM table2

3.1.1 UNION操作符

UNION 操作符返回两个查询的结果集的并集，去除重复记录。

3.1.2 UNION ALL操作符

UNION ALL操作符返回两个查询的结果集的并集。对于两个结果集的重复部分，不去重。

select 'newMember' as "memberType", count(distinct full_union_id) as "memberCount" 
from XXX.XXX  
where XX_id = cast(66666666 as varchar) and XX_time >= '2021-06-16 00:00:00' and XX_time <= '2022-06-16 00:00:00' union allselect 'oldMember' as "memberType", count(distinct full_union_id) as "memberCount" 
from XXX.XXX  
where XX_id = cast(66666666 as varchar) and XX_time <'2021-06-16 00:00:00' 

memberType	memberCount
oldMember	2,666,888
newMember	1,888,666

注意：执行UNION ALL语句时所需要的资源比UNION语句少。如果明确知道合并数据后的结果数据不存在重复数据，或者不需要去除重复的数据，则尽量使用UNION ALL语句，以提高数据查询的效率。

4. 7种JOIN操作

中图：内连接

SELECT employee_id, last_name, department_name
FROM employees eJOIN departments dON e.department_id = d.department_id;

左上图：左外连接

SELECT employee_id, last_name, department_name
FROM employees eLEFT JOIN departments dON e.department_id = d.department_id;

右上图：右外连接

SELECT employee_id,last_name,department_name
FROM employees eRIGHT JOIN departments dON e.department_id = d.department_id;

左中图

SELECT employee_id, last_name, department_name
FROM employees eLEFT JOIN departments dON e.department_id = d.department_id
WHERE d.department_id IS NULL;

右中图

SELECT employee_id, last_name, department_name
FROM employees eRIGHT JOIN departments dON e.department_id = d.department_id
WHERE e.department_id IS NULL;

左下图： 左中图 + 右上图

SELECT employee_id,last_name,department_name
FROM employees eLEFT JOIN departments dON e.department_id = d.department_id
WHERE d.department_id IS NULL
UNION ALL #没有去重操作，效率高
SELECT employee_id,last_name,department_name
FROM employees eRIGHT JOIN departments dON e.department_id = d.department_id;

右下图：左中图 + 右中图

SELECT employee_id,last_name,department_name
FROM employees eLEFT JOIN departments dON e.department_id = d.department_id
WHERE d.department_id IS NULL
UNION ALL
SELECT employee_id,last_name,department_name
FROM employees eRIGHT JOIN departments dON e.department_id = d.department_id
WHERE e.department_id IS NULL

5. join 多张表

SELECT B.id, A.sid, C.c_type, C.c_name, D.sku_id
FROM db_name.table_name_a ARIGHT JOIN db_name.table_name_b B ON A.id = B.idRIGHT JOIN db_name.table_name_c C ON  B.channel_id  = C.channel_idRIGHT JOIN db_name.table_name_d D ON A.id = D.oid
WHERE A.code = 16AND A.order_status != 4AND A.create_time > '2023-03-08 21:00:00'AND A.tid = '1234567890'
ORDER BY A.create_time DESC 
LIMIT 200;

注意：

我们要控制连接表的数量。多表连接就相当于嵌套 for 循环一样，非常消耗资源，会让 SQL 查询性能下降得很严重，因此不要连接不必要的表。在许多 DBMS 中，也都会有最大连接表的限制。

【强制】超过三个表禁止 join。需要 join 的字段，数据类型保持绝对一致；多表关联查询时，保证被关联的字段需要有索引。

说明：即使双表 join 也要注意表索引、SQL 性能。

来源：阿里巴巴《Java开发手册》