1. 专业术语

RDBMS(Relaction DataBase Mangement System):MySQL、SQL Server、IBM DB2、Oracle

SQL(Struted Query Language): 结构化查询语言,是用来连接和操作RDBMS的标准计算机语言

ER(Entity RelationShip):实体关系图,用来描述业务实体数据之间的关系。

三个图形:矩形(实体类型)、椭圆形(属性名)、菱形(关系)、连线(几对几的关系)

SQL语言分类:

DDL:数据定义语言,一般用来对数据库表进行结构调整的,比如Create、Drop、Alter

DML:数据操作语言,对数据进行增删改查操作,比如Insert、Delete、Update、Select

DCL: 数据控制语言, grant、commit、rollback

ACID:

A (Atomictiy): 原子性

C(Consistency): 一致性

I (ISolation): 隔离性, 由事务隔离级别决定隔离性影响的大小

D(Durability): 持久性

PRIMARY KEY: 主键ID

FOREIGN KEY:互联网公司极少用物理外键,用逻辑外键

INDEX:索引 (主键索引、普通索引、唯一索引)

2. SQL之DML语句

2.1 增删改

  • 插入一条: 插入一条频道

    insert into wm_channel values (9,"Scala","新型编程语言",0,1,9,"2022-06-27 12:00:00");
  • 插入多条: 插入多条频道

    insert into wm_channel values (10,".NET","微软件编程语言",0,0,10,"2022-06-27 13:00:00"),(11,"C++","很好的编程语言",0,1,11,"2022-06-27 14:00:00");
  • 删除ID为11的频道

    delete from wm_channel where id=11;
  • 修改ID为10的频道名称为c++,描述改为不错的编程语言

    update wm_channel set name="c++",description="不错的编程语言" where id=10;

2.2 单表查询

  • 条件查询:查询频道名为java的频道

    select * from wm_channel where name = "java";
  • 逻辑查询

    • 与查询:查询名称为java且状态为1的频道

      select * from wm_channel where name = "java" and status = 1;
    • 或查询:查询名称为java或状态为0的频道

      select * from wm_channel where name = "java" or status = 0;
    • 非查询:查询状态不为1的频道

      select * from wm_channel where status != 1;
  • 模糊查询:查询描述包含框架的频道

    select  * from wm_channel where wm_channel.description like  "%框架%";
  • 区间查询

    • 区间查询1:查询序号为6和7的频道

      select * from wm_channel where ord in (6,7);
    • 区间查询2:查询序号大于5的频道

      select * from wm_channel where ord > 5;
    • 区间查询3:查询时间大于2022-06-24 00:00:00的频道

       select * from wm_channel where created_time > "2022-06-24 00:00:00";
    • 区间查询4:查询时间大于等于2022-06-22 00:00:00且小于等于2022-06-24 00:00:00的频道

      select  * from wm_channel where created_time between "2021-06-22 00:00:00" and "2021-06-24 00:00:00";

2.3 多表关联查询

查询文章标题、所属频道名、文章状态、文章创建时间

  • 写法一- 直接FROM两张表

    select wn.title,wc.name channelName,wn.status,wn.created_time from wm_news wn,wm_channel wc where wn.channel_id = wc.id;
  • 写法二- JOIN两张表

    select wn.title,wc.name channelName,wn.status,wn.created_time from wm_news wn left join wm_channel wc on  wn.channel_id = wc.id;

2.4 子查询

查询频道名称为java和Python的所有文章(仅需查询文章标题、频道id、创建时间)

select title,channel_id,created_time from wm_news where channel_id in (select id from wm_channel where name in ( 'java','Python'))

2.5 分页排序

  • 分页查询

    • 分页查询1:从第2条开始查询3条频道

      select * from wm_channel limit 1,3;
    • 分页查询2:查询前4条频道

      select * from wm_channel limit 4;
  • 排序查询:查询文章标题、发布时间按照发布时间倒排序

    select  title,publish_time from wm_news order by publish_time desc ;

2.6 聚合查询

  • 查询频道序号最小值

    select min(ord) from wm_channel;
  • 查询频道序号最大值

    select max(ord) from wm_channel;
  • 查询频道序号平均值

    select  avg(ord) from wm_channel;
  • 查询频道序号之和

    select sum(ord) from wm_channel;
  • 查询每天对应的已发布的文章数量,只查询发布数量大于2的记录,按数量倒排序,取前5条

    select date_format(publish_time,"%Y-%m-%d") dateStr,count(*) dateCount from wm_news group by dateStr having dateCount>2 order by dateCount desc limit 5;
最后修改:2023 年 07 月 22 日
如果觉得我的文章对你有用,请随意赞赏