文章目录
- 0. 交集、并集、差集含义说明
- 1. 简单演示上图七种情况
- 0. A、B表数据准备
- 1. left outer join 简称 left join 左表所有数据,右表关联数据,没有的以null填充
- 2. right outer join 简称 right join,右表所有数据,左表关联数据,没有的以null填充
- 3. inner join 简称 join 交集
- 4. A left join B where B.key is null,A、B的差集是 A-B = 1
- 5. B right join where A.key is null,B、A的差集是 B-A = 4
- 6、7. full outer join 简写为full join(mysql不支持,oracle可以)
- 2. 笛卡尔积 A * B
- 3. inner/left join重复数据,查询会扩散,造成笛卡尔积现象
- 4. a inner/left join b,没有on关键字会报错
- 5. left join on 后 and中主表的过滤条件不会起作用
![在这里插入图片描述](https://i-blog.csdnimg.cn/blog_migrate/fff4844378b1fa3bd346865831227dd8.png)
0. 交集、并集、差集含义说明
A 是1、2、3
B是2、3、4
A、B的交集是A∩B = 2、3
A、B的并集是 AUB = 1、2、3、4
A、B的差集是 A-B = 1
B、A的差集是 B-A = 4
1. 简单演示上图七种情况
0. A、B表数据准备
CREATE TABLE `xin_stu_t` (
`id` bigint NOT NULL AUTO_INCREMENT COMMENT '主键',
`relation_id` bigint DEFAULT NULL COMMENT '外键, 记录教师id',
`student_name` varchar(30) DEFAULT NULL COMMENT '姓名',
`student_age` bigint DEFAULT NULL COMMENT '年龄',
`school` varchar(300) DEFAULT NULL COMMENT '学校',
PRIMARY KEY (`id`) USING BTREE,
KEY `xin_s_relation_id` (`relation_id`),
KEY `xin_s_student_name` (`student_name`)
) ENGINE=InnoDB AUTO_INCREMENT=18 DEFAULT CHARSET=utf8mb4 COLLATE=utf8mb4_0900_ai_ci COMMENT='学生表';
CREATE TABLE `xin_teach_t` (
`id` bigint NOT NULL AUTO_INCREMENT COMMENT '主键',
`teacher_name` varchar(30) DEFAULT NULL COMMENT '教师姓名',
`teacher_age` bigint DEFAULT NULL COMMENT '教师年龄',
`school` varchar(300) DEFAULT NULL COMMENT '学校',
PRIMARY KEY (`id`) USING BTREE
) ENGINE=InnoDB AUTO_INCREMENT=9 DEFAULT CHARSET=utf8mb4 COLLATE=utf8mb4_0900_ai_ci COMMENT='教师表';
INSERT INTO lelele.xin_stu_t (relation_id,student_name,student_age,school) VALUES
(NULL,'尤仁义1',11,'徐州中学'),
(1,'尤仁义2',12,'徐州中学'),
(NULL,'朱有理1',11,'徐州中学'),
(2,'朱有理2',12,'徐州中学'),
(2,'朱有理3',13,'徐州中学'),
(3,'宋昆明1',11,'徐州中学'),
(3,'宋昆明2',12,'徐州中学'),
(9,'宋昆明3',13,'徐州中学');
INSERT INTO lelele.xin_teach_t (teacher_name,teacher_age,school) VALUES
('王翠花1',31,'徐州中学'),
('王翠花2',31,'徐州中学'),
('王翠花3',33,'徐州中学'),
('王翠花4',34,'徐州中学'),
('王翠花5',35,'徐州中学');
1. left outer join 简称 left join 左表所有数据,右表关联数据,没有的以null填充
select A.*,B.* from xin_stu_t A left join xin_teach_t B on A.relation_id = B.id
select A.*,B.* from xin_stu_t A left outer join xin_teach_t B on A.relation_id = B.id
2. right outer join 简称 right join,右表所有数据,左表关联数据,没有的以null填充
select A.*,B.* from xin_stu_t A right join xin_teach_t B on A.relation_id = B.id
select A.*,B.* from xin_stu_t A right outer join xin_teach_t B on A.relation_id = B.id
3. inner join 简称 join 交集
select A.*,B.* from xin_stu_t A inner join xin_teach_t B on A.relation_id = B.id
select A.*,B.* from xin_stu_t A join xin_teach_t B on A.relation_id = B.id
4. A left join B where B.key is null,A、B的差集是 A-B = 1
select A.*,B.* from xin_stu_t A left join xin_teach_t B on A.relation_id = B.id where B.id is null
5. B right join where A.key is null,B、A的差集是 B-A = 4
select A.*,B.* from xin_stu_t A right join xin_teach_t B on A.relation_id = B.id where A.id is null
6、7. full outer join 简写为full join(mysql不支持,oracle可以)
2. 笛卡尔积 A * B
1. 不带连接条件的笛卡尔积
select A.*,B.* from xin_stu_t A, xin_teach_t B
2. 带连接条件的笛卡尔积变inner join
select A.*,B.* from xin_stu_t A, xin_teach_t B where A.relation_id = B.id
3. inner/left join重复数据,查询会扩散,造成笛卡尔积现象
造数据
drop table if exists xin_stu_t_bak;
drop table if exists xin_teach_t_bak;
create table `xin_stu_t_bak` (
`id` bigint not null COMMENT '主键',
`relation_id` bigint default null COMMENT '外键, 记录教师id',
`student_name` varchar(30) default null COMMENT '姓名',
`student_age` bigint default null COMMENT '年龄',
`school` varchar(300) default null COMMENT '学校'
);
CREATE TABLE `xin_teach_t_bak` (
`id` bigint NOT NULL COMMENT '主键',
`teacher_name` varchar(30) DEFAULT NULL COMMENT '教师姓名',
`teacher_age` bigint DEFAULT NULL COMMENT '教师年龄',
`school` varchar(300) DEFAULT NULL COMMENT '学校'
);
insert into xin_stu_t_bak select * from xin_stu_t;
insert into xin_teach_t_bak select * from xin_teach_t;
INSERT INTO lelele.xin_teach_t_bak (id,teacher_name,teacher_age,school) values (1,'王翠花1',31,'徐州中学');
INSERT INTO lelele.xin_teach_t_bak (id,teacher_name,teacher_age,school) VALUES (1,'王翠花1',31,'徐州中学');
INSERT INTO lelele.xin_teach_t_bak (id,teacher_name,teacher_age,school) VALUES (2,'王翠花2',31,'徐州中学');
3.1 inner join 重复数据 结果集会扩散
select A.,B. from xin_stu_t_bak A inner join xin_teach_t_bak B on A.relation_id = B.id order by A.id;
3.2 inner join 重复数据,结果集去重后同inner join不重复数据
select distinct A.,B. from xin_stu_t_bak A inner join xin_teach_t_bak B on A.relation_id = B.id order by A.id;
3.3 left join 重复数据 结果集会扩散
select A.,B. from xin_stu_t_bak A left join xin_teach_t_bak B on A.relation_id = B.id order by A.id;
3.4 left join 重复数据,结果集去重后同left join不重复数据
select distinct A.,B. from xin_stu_t_bak A left join xin_teach_t_bak B on A.relation_id = B.id order by A.id;
4. a inner/left join b,没有on关键字会报错
- on是放a、b表的连接条件,只会查出满足条件的数据
- 若b表数据重复,会发生笛卡尔积现象
- 若on后的关联条件不是正常的连接条件A.relation_id = B.id,而是A.relation_id=‘111’,虽然不会报错,但它也变成笛卡尔积语句了
- from a,b where a.id = b.relation_id,相当于from a inner join b on a.id = b.relation_id
5. left join on 后 and中主表的过滤条件不会起作用
A left join B on A.relation_id = B.id and A.列名 = ‘’ and B.列名 = ‘’ 左表,A.列名 = ''这个条件不起作用
说明:
left join on 后的 and 条件中 主表的条件不生效,从表的条件生效,并且从表先进行筛选后数据量可能变少,更便于与主表关联,有利于提高查询效率。
建议:
主表的筛选条件放在 where 中,从表的 筛选条件放在 on 后的 and 中。
- on条件是在生成临时表时使用的条件,它不管and中的条件是否为真,都会返回左边表中的所有记录。 所以说 and后 加 左表的条件对
左表无影响。and后的条件只对右表产生影响 (right join翻过来) - where条件是在临时表生成好后,再对临时表进行过滤的条件。这时已经没有left
join的含义(必须返回左表的记录)了,条件不为真的就全部过滤掉。即对 join后的数据再进行过滤,过滤出只符合where后的条件。 - 在匹配阶段 where子句的条件都不会被使用。仅在匹配阶段完成以后,wehre子句条件才会被使用。它将从匹配阶段产生的数据中检索过滤
总结:
-
当条件加在 left(right) join on之后,则只会影响右(左)边数据,不会影响左(右)表数据,不管条件是否成立,左(右))边数据都能正常返回;
-
inner join and 和 inner join where and 其查询结果基本一致;
-
当条件加在full on之后,则会根据and之后的条件分别影响左表或右表数据;
-
当条件加在where 之后则表示对关联结果再进行筛选,此时的结果将会依赖于where后边条件的真假