SQL:如何选择具有重复项的唯一行

Question

我们的客户向我们提供重复项目的契约(Contract)。我们宣传独特的产品。如何在不进行许多子查询的情况下选择一组带有描述的不同 upcs？可怕的例子:

/* 
eventID     int
groupID     int            // This field is different between fields
upc_ean     numeric(18,0)
description varchar(512)   // Description is slightly different but same info
size        varchar(512)   // Size is slightly different but same info
*/

select A.eventid, A.upc_ean,

       ( select top 1 description 
         from myTable B 
         where B.eventid = A.eventid and B.upc_ean = A.upc_ean) as description,

       ( select top 1 size
         from myTable B
         where B.eventid = A.eventid and B.upc_ean = A.upc_ean) as size

from ( select distinct eventid, upc_ean from myTable) A

有没有办法在没有子查询的情况下做同样的事情，以某种方式将两者结合在一起，不会产生记录或使用 eventid 和 upc_ean 作为 PK 显示重复项？

Answer 1

如有必要，您可以在以下的 OVER 部分中添加 ORDER BY 子句。

使用 CTE:

WITH example AS (
  SELECT a.eventid, 
         a.upc_ean,
         a.description,
         a.size,
         ROW_NUMBER() OVER(PARTITION BY a.eventi, a.upc_ean) AS rank
    FROM YOUR_TABLE a)
SELECT x.eventid,
       x.upc_ean,
       x.description, 
       x.size
  FROM example x
 WHERE x.rank = 1

没有 CTE:

SELECT x.eventid,
       x.upc_ean,
       x.description, 
       x.size
  FROM (SELECT a.eventid, 
               a.upc_ean,
               a.description,
               a.size,
               ROW_NUMBER() OVER(PARTITION BY a.eventi, a.upc_ean) AS rank
          FROM YOUR_TABLE a) x
 WHERE x.rank = 1