选择不同的行

Question

我不知道如何去寻找这个，于是我开始了一个新的问题。

我有一张表，结构如下

----------------------------------------
| TALKS            |    PERSON
----------------------------------------
| Networks         |  John Doe   
| Steganography    |  Alex
| Assembly         |  Mark
| Networks         |  Mark
| Steganography    |  John Doe
| Networks         |  Mark
----------------------------------------

如何执行查询，以便

----------------------------------------
| TALKS            |    PERSON
----------------------------------------
| Networks         |  John Doe   
| Steganography    |
| Assembly         |  Mark
| Networks         |  
| Steganography    |  Alex
----------------------------------------

请注意，“Mark”有两次关于“网络”的讨论，但它只需要返回一次。

Answer 1

从桌子上选择不同的谈话

Answer 2

这就是你想要达到的目标。

(
  select talks,name 
  from table1
  group by name  
  order by talks
)
UNION ALL
(
  select distinct talks as talks, 
  case 
    when name is not null then ''
  end
  name
  from table1
)

演示

Answer 3

在您的隐含规范中有两件事正在进行。一种是显示不同的对话，第二种是格式化结果集，以避免在连续的行中重复单个人的名称。

此查询(http://sqlfiddle.com/#!2/dc2d3/4/0)同时执行两项操作。

select talks,
       if (@prev = person,'',@prev:=person) person
  from (
         select distinct talks, person
           from talk
          order by person, talks
       ) AS a,
  (select @prev := '') AS init

避免@变量猴子业务的替代查询(http://sqlfiddle.com/#!2/dc2d3/9/0)是

select group_concat(distinct talks order by talks) AS talks,
       person 
  from talk
 group by person