Question

我有一个字符串资源表：

;WITH cte AS 
(
    SELECT 1 AS id , 'john' AS name, 10 AS age
    UNION   
    SELECT 2 AS id , 'john' AS name, 10 AS age
    UNION   
    SELECT 3 AS id , 'john' AS name, 12 AS age
    UNION   
    SELECT 4 AS id , 'paul' AS name, 6 AS age
    UNION   
    SELECT 5 AS id , 'paul ' AS name, 6 AS age
    UNION   
    SELECT 6 AS id , 'paul different' AS name, 7 AS age
    UNION   
    SELECT 7 AS id , 'ringo' AS name, 2 AS age

)

所以这个名字＆＃34; john＆＃34;年龄10岁。

后来，其他人（不是我）也添加了＃34; john＆＃34; 10岁。

所以我想清理所有副本。

但那不是问题。在我删除之前，我希望看到所有重复项。

所以我这样做了：

SELECT *
FROM   (
           SELECT ID,
                  name,
                  age,
                  ROW_NUMBER() OVER(PARTITION BY name, age ORDER BY id) AS rn
           FROM   cte

       ) a  WHERE a.rn>1
ORDER BY
       name,
       age,
       a.rn

结果：

enter image description here

这基本上显示了我的重复。但是，只有在当前值有更多版本的情况下，我才想看到rn=1的位置。

问题

换句话说：如何增强我的查询：

仅当有此记录的版本时，才显示所有版本的记录（所有行号，rn）

期望的结果：

ID  name    age rn
1   john    10  1
2   john    10  2
4   paul    6   1
5   paul    6   2

Sql online - demo

注意我知道我可以通过重新扫描同一name和age的表格来实现这一目标。 but我想是否有更优雅的方式。

Answer 1

使用exists运算符查找duplicated的名称。试试这个。

;WITH cte AS 
(
    SELECT 1 AS id , 'john' AS name, 10 AS age
    UNION   
    SELECT 2 AS id , 'john' AS name, 10 AS age
    UNION   
    SELECT 3 AS id , 'john' AS name, 12 AS age
    UNION   
    SELECT 4 AS id , 'paul' AS name, 6 AS age
    UNION   
    SELECT 5 AS id , 'paul ' AS name, 6 AS age
    UNION   
    SELECT 6 AS id , 'paul different' AS name, 7 AS age
    UNION   
    SELECT 7 AS id , 'ringo' AS name, 2 AS age

)
, cte1
         AS (SELECT ID,
                    name,
                    age,
                    Row_number() OVER(PARTITION BY name, age ORDER BY id) AS rn
             FROM   cte)
    SELECT *
    FROM   cte1 a
    WHERE  EXISTS (SELECT 1
                   FROM   cte1 b
                   WHERE  a.name = b.name and a.age=b.age
                          AND b.rn > 1)
    ORDER  BY name, age, a.rn

或使用Inner Join

    SELECT a.id,a.name,a.age
    FROM   cte1 a
           JOIN cte1 b
             ON a.name = b.name
                AND a.age = b.age
                AND b.rn > 1
    ORDER  BY a.name, a.age, a.rn

或者要在单表扫描中使用Dense_Rank加window function

    ;WITH cte AS 
    (
        SELECT 1 AS id , 'john' AS name, 10 AS age
        UNION   
        SELECT 2 AS id , 'john' AS name, 10 AS age
        UNION   
        SELECT 3 AS id , 'john' AS name, 12 AS age
        UNION   
        SELECT 4 AS id , 'paul' AS name, 6 AS age
        UNION   
        SELECT 5 AS id , 'paul ' AS name, 6 AS age
        UNION   
        SELECT 6 AS id , 'paul different' AS name, 7 AS age
        UNION   
        SELECT 7 AS id , 'ringo' AS name, 2 AS age

    )
    , cte1
             AS (SELECT ID,
                        name,
                        age,
                        count(age) over (partition by name,age) cnt,
                        dense_rank() OVER(PARTITION BY name ORDER BY age) AS rn
                 FROM   cte)
                 SELECT *
                 FROM   cte1
                 WHERE  rn = 1
                        AND cnt > 1

显示重复项的所有RowNumber记录？

1 个答案: