除了基于列的重复项中的重复项

Question

这是我正在尝试做的事情，

• 获取基于 2 列的重复项（假设返回 500 行）

• 根据以上 2 列 + 另一列获取重复项（假设返回 100 行）

现在我想获得剩余的 400 行。简而言之，我想要所有由于 columnC 而没有重复项的重复项...

-- get duplicates based on ColumnA, ColumnB
SELECT '-'
    ,ColumnA
    ,ColumnB
    ,ColumnC
    ,COUNT(*)
FROM MainTable
     ...SOME joins(INNER AND left)
WHERE ColumnA IS NOT NULL
GROUP BY ColumnA
    ,ColumnB
    ,ColumnC
HAVING COUNT(*) > 1

EXCEPT

-- get duplicates based on ColumnA, ColumnB, ColumnC
SELECT '-'
    ,ColumnA
    ,ColumnB
    ,ColumnC
    ,COUNT(*)
FROM MainTable
     ...SOME joins(INNER AND left)
WHERE ColumnA IS NOT NULL
GROUP BY ColumnA
    ,ColumnB
    ,ColumnC
HAVING COUNT(*) > 1
ORDER BY COUNT(*) DESC

我只是无法完成此查询:(

Answer 1

在这里使用窗口函数可能会更简单：

数据库<>小提琴

WITH cte AS
( 
    SELECT 
        ColumnA,
        ColumnB,
        ColumnC,
        COUNT(*) OVER (PARTITION BY ColumnA, ColumnB)
          AS count_ab,
        COUNT(*) OVER (PARTITION BY ColumnA, ColumnB, ColumnC)
          AS count_abc
    FROM MainTable
         ...SOME joins(INNER AND left)
    WHERE ColumnA IS NOT NULL
)
SELECT
    ColumnA,
    ColumnB,
    ColumnC,
    count_ab
FROM 
    cte
WHERE
    count_ab > 1
  AND
    count_abc = 1 ;