mysql高级

mysql支持三种类型的连接查询，分别为：

内连接，左连接，右连接

内连接：   where /     inner join         /   join        

(1)select s.name,c.name from mingxingstudents as s ,classes as c where s.cls_id=c.id;

(2)select s.name,c.name from mingxingstudents as s inner join classes as c on s.cls_id=c.id;

(3)select s.name,c.name from mingxingstudents as s join classes as c on s.cls_id = c.id;

(4)select s.name,c.name from mingxingstudents as s cross join classes as c on s.cls_id = c.id;

左右连接查询

左连接查询（左边的表是主表，全部显示，右边的是从表）

select s.name,c.name from mingxingstudents as s left outer join classes as c on s.cls_id = c.id;

(上面这个查询语句中的outer可以省略select s.name,c.name from mingxingstudents as s left join classes as c on s.cls_id = c.id;

)

右连接查询（刚好与左连接相反）

select s.name,c.name from mingxingstudents as s right outer join classes as c on s.cls_id = c.id;

交叉查询（容易产生笛卡儿积现象)

select * from mingxingstudents cross join classes;

自关联查询的基本操作

（自连接，自身关联自身，自连接必然用到子查询）

select p.name as省,c.name as 市 from province as p inner join province as c on c.pid = p.id where p.name ="河北省"

子查询

在一个 select 语句中,嵌入了另外一个 select 语句, 那么被嵌入的 select 语句称之为子查询语句

主查询:主要查询的对象,第一条 select 语句

主查询和子查询的关系:子查询是嵌入到主查询中;子查询是辅助主查询的,要么充当条件,要么充当数据源;子查询是可以独立存在的语句,是一条完整的 select 语句

子查询分类:

(1)标量子查询: 子查询返回的结果是一个数据(一行一列)  select avg(height) from mingxingstudents;

(2)列级子查询: 返回的结果是一列(一列多行)select classes.name from classes inner join mingxingstudents on classes.id = mingxingstudents.cls_id;

(3)行级子查询: 返回的结果是一行(一行多列)select max(age),max(height) from mingxingstudents;

查询班级中年龄最小并且身高也最小的人：

select  * from mingxingstudents where (age,height) = (select min(age),min(height)from mingxingstudents);

(4)表级子查询: 返回的结果是多行多列(一般不用，查询效率相对费时间。可以使用连接查询实现的)

select t.sname,t.cname from (select s.name as sname,c.name as cname from mingxingstudents as s inner join classes as c on s.cls_id = c.id) as t;

子查询中出现的关键字

any与some,any与in等价

select classes.name from classes where id = any(select cls_id from mingxingstudents);

select classes.name from classes where id in(select cls_id from mingxingstudents);

where 列 = all(列子查询)：等于里面所有

where 列 <> all(列子查询)：不等于其中所有

select classes.name from classes where id = all(select cls_id from mingxingstudents);

select classes.name from classes where id <> all(select cls_id from mingxingstudents);

in范围

select classes.name from classes where id not in(select cls_id from mingxingstudents);

select * from goods where price >(select avg(price) from goods) order by price desc;

select price from goods where cate = '超级本' order by price;

select * from goods where price >= any(select price from goods where cate ='超级本') order by price desc;

---

将一个表进行分表处理

判断不存在创建一个空表create table if not exists goods_cates(id int unsigned primary key auto_increment,cate_name varchar(40));

将查询到的数据插入到表中insert into goods_cates(cate_name) select cate from goods group by cate;

将原表中的字段设置成对应的新表里的id

update goods as g inner join goods_cates as c on g.cate = c.cate_name set g.cate = c.id;

再创建一个新表并插入数据

create table goods_brands(id int unsigned primary key auto_increment not null,brand_name varchar(40))select brand_name from goods group by brand_name;

再次更新原表（将brand_name = 设置为brand_id）

update goods as g inner join goods_brands as b on g.brand_name = b.brand_name set g.brand_name = b.id;

---

修改表结构:(修改了两个部分，cate的类型被修改成int，brand_name 修改成了brand_id，类型为int)

alter table goods change cate cate_id int unsigned,change brand_name brand_id int unsigned;

---

多张表的连接查询

(两个左连接的方式)select * from goods as g left join goods_cates as c on g.cate_id = c.id left join goods_brands as b on g.brand_id = b.id;

(两个右连接的方式)select * from goods as g right join goods_cates as c on g.cate_id =c.id right join goods_brands as b on g.brand_id = b.id;

(两个内连接)select * from goods as g inner join goods_cates as c on g.cate_id =c.id inner join goods_brands as b on g.brand_id = b.id;

外键约束的创建和删除

alter table goods add foreign key (brand_id) references goods_brands(id);

alter table goods add foreign key (cate_id) references goods_brands(id);

删除外键：alter table goods drop  foreign key goods_ibfk_1;

删除key：alter table goods drop key catte_id;

备份和还原

备份一个表：(注意：退出mysql后使用)

mysqldump -uroot -p python students > D:\ps.sql

备份一个数据库：

mysqldump -uroot -p python > D:\ss.sql

备份多个数据库

视图

视图的创建：create view v_mstu as select * from mingxingstudents;

视图的查询：select * from v_mstu

删除视图：drop view v_mstu

索引

是一种特殊的文件(InnoDB数据表上的索引是表空间的一个组成部分)，它们包含着对数据表所有记录的引用指针。

查看索引：show index from 表名；

创建索引：create index 索引名 on 表名

删除索引：drop index 索引名 on 表名

开启运行时间：set profiling=1

显示查看所需时间： show profiling；

当数据表中创建主键和外键的时候，就会自动的创建索引。

建立太多的索引将会影响更新和插入的速度，因为它需要同样更新每个索引文件。对于一个经常需要更新和插入的表格，就没有必要为一个很少使用的where字句单独建立索引了，对于比较小的表，排序的开销不会很大，也没有必要建立另外的索引。建立索引会占用磁盘空间。

数据库的交互

connect对象：host,port,database,user,password,charset

对象的方法：

close

commit，

rollback，

cursor（）返回cursor对象，执行sql语句并获得结果

execute()执行语句

fetchone（）

fetchall（）

调用connection对象的cursor（）方法：cursor=conn.cursor()

对象的属性：

rowcount只读属性，表示最近一次execute（）执行后受影响的行数

connection获得当前连接对象