最近总感觉sql语句不对劲，所以就看了一些官方文档发现了一些以前没有注意的函数：感觉在查询的时候可以用得上，毕竟是内置函数，用起来效率应该会好一些的：

FIND_IN_SET(str,strlist)

mysql> SELECT FIND_IN_SET('b','a,b,c,d');

        -> 2

如：SELECT FIND_IN_SET(a.roleId,r.roleName), a.* from t_admin a ,t_role r where a.roleId = r.roleId AND FIND_IN_SET(a.roleId,r.roleId);

代替in、not in

INSTR(str,substr)

mysql> SELECT INSTR('foobarbar', 'bar');

        -> 4

如：select a.* FROM  t_admin a where INSTR(a.account,'发达');

可以用于模糊查询 代替like  提高效率

LOCATE(substr,str), LOCATE(substr,str,pos)

mysql> SELECT LOCATE('xbar', 'foobar');

        -> 0

mysql> SELECT LOCATE('bar', 'foobarbar', 5);

        -> 7

select a.* FROM  t_admin a where LOCATE('发达',a.account);

可以用于模糊查询 代替like  提高效率

• select  
      substring_index(email,'@',-1) as domain,  
      count(distinct email) as count,  
      count(distinct email)  
          / (select count(distinct email) from t_user where email like '%@%')  
          * 100 as percent  
  from t_user  
  where email like '%@%'   
  group by 1   
  order by 3 ASC   
  limit 10;

select

    substring_index(email,'@',-1) as domain,

    count(distinct email) as count,

    count(distinct email)

        / (select count(distinct email) from t_user where email like '%@%')

        * 100 as percent

from t_user

where email like '%@%'

group by 1

order by 3 ASC

limit 10;