Question

例如，如果我们有值：

username             Category
-------------------- ----------
Brian                cs
Tom                  cs
Joe                  cs
Allen                cs
Bill                 ts
Steven               ts
Fred                 ts
Ted                  ts

我需要输出

username             Category
-------------------- ----------
Brian                cs
Tom                  
Joe                  
Allen                
Bill                 ts
Steven               
Fred                 
Ted

我试过了

1）T-SQL How to select rows without duplicate values from one column?

2）http://www.rajapet.com/2010/04/suppressing-repeated-column-value-in.html

3）How to select records without duplicate on just one field in SQL?

什么都不好（我所期待的）

还有其他办法吗？

Answer 1

您可以使用window function：

select username, 
       case when rn = 1 then cat else '' end as category
from (
   select username, 
          category as cat, 
          row_number() over (partition by category order by username) as rn
   from the_table
) t
order by cat, username;

SQLFiddle：http://sqlfiddle.com/#!15/423ed/4

Answer 2

这是使用子查询的另一种方法。请注意，两个答案都会按字母顺序将类别值提供给第一个用户名，即Allen将获得cs和Bill ts。

select a.username,b.category from
foo a
left join
(select category,min(username) as username from foo group by category) b
on b.username = a.username
order by a.category,a.username

[编辑]如果你想分配到你表格中的第一个实例（即返回你在问题中提出的内容），你可以这样做：

    with foobar as 
        (select username,category, row_number() over() as rownum from foo )

   select a.username,coalesce(b.category,'') as category from
    foobar a
    left join
    (select category,min(rownum) as rownum from foobar group by category) b
    on b.rownum = a.rownum
    order by a.rownum,a.username

在PostgreSQL中抑制重复的列值？

2 个答案: