计算给定列

时间:2018-04-13 18:49:32

标签: mysql sql

如果数据库表的列的值类似于' A.B',' C.D',' A.C',我该怎么办?使用sql查询来查找具有' A',' B',' C'或者' D' (按字符分组)?

我期待结果是:

 _________________________
|  Characters| COUNT(*)   |
|*************************|
|    A       |    10      |
|*************************|
|    B       |    15      |
|************|************|
|    C       |     8      |
|************|************|
|    D       |     17     |
|____________|____________|

包含' A'的列的条目和' B' (比如' A.B')应计入A' A'和' B'。

2 个答案:

答案 0 :(得分:4)

您可以使用模式匹配来完成此操作。这是一种方法:

select pattern.name, count(t.column)
from (select '%A%' as pattern, 'A' as name union all
      select '%B%' as pattern, 'B' as name union all
      select '%C%' as pattern, 'C' as name union all
      select '%D%' as pattern, 'D' as name
     ) patterns left join
     t
     on t.column like patterns.pattern
group by pattern.name;

请注意,这使用子查询为模式定义“派生表”。这种确切的语法可能不适用于所有数据库,但类似的东西应该有效。

答案 1 :(得分:1)

你走了:

create table my_table (
  characters varchar(100)
);

insert into my_table (characters) values ('A.B');
insert into my_table (characters) values ('C.D');
insert into my_table (characters) values ('A.C');

select 'A' as letter, count(*) from my_table where characters like '%A%'
union select 'B', count(*) from my_table where characters like '%B%'
union select 'C', count(*) from my_table where characters like '%C%'
union select 'D', count(*) from my_table where characters like '%D%';

结果:

letter  count(*)
------  --------
A              2
B              1
C              2
D              1