首页 > 代码库 > oracle笔记
oracle笔记
1.获取唯一记录:distinct
select distinct e.employee_name,s.salary
from employees e,salary s
where e.employee_id=s.salary;
2.分组:group by
select e.employee_id, e.employee_name, sum(s.salary) total_salary
from employees e, salary s
where e.employee_id=s.employee_id
group by e.employee_id,e.employee_name;
3.过滤分组:having
where 子句可以过滤from 子句所指定的数据源,但是对于group by 子句所产生的分组无效。为了半分组按一定条件进行过滤。having 子句是依附于group by 子句存在而存在的。
select e.employee_id, e.employee_name, sum(s.salary) total_salary
from employees e, salary s
where e.employee_id=s.employee_id
group by e.employee_id,e.employee_name;
having (sum(s.salary)) > 10000;
4.排序:order by (asc 同时也是默认排序方式,降序 desc)
select distinct e.employee_name, s.salary
from employees e, salary s
where e.employee_id=s.employee_id
order by s.salary desc;
5.order by 与 group by
select e.employee_name, sum(s.salary) total_salary
from employees e, salary s
where e.employee_id=s.employee_id
group by e.employee_id,e.employee_name
order by total_salary desc;
注:order by 后面不能接除 e.employee_name, total_salary之外的列进行排序。
6.order by 与 distinct
order by 与 distinct 关键字同时使用时,也必须遵循一个规则:order by 子句所指定的排序列,必须出现在
select 表达式中。
select e.employee_name
from employees e,salary s
where e.employee_id=s.employee_id
order by s.salary;-------成功
但:
select distinct e.employee_name
from employee e,salary s
where e.employee_id=s.employee_id
order by s.salary;-------失败
oracle笔记
声明:以上内容来自用户投稿及互联网公开渠道收集整理发布,本网站不拥有所有权,未作人工编辑处理,也不承担相关法律责任,若内容有误或涉及侵权可进行投诉: 投诉/举报 工作人员会在5个工作日内联系你,一经查实,本站将立刻删除涉嫌侵权内容。