Question

想象一下这样的表：

RecordID  Key1  Key2
--------------------
1         111   222  
2         222   111  
3         555   444  
4         000   444  
5         444   000

由于重复项被视为在列Key1和Key2上具有相同集的记录。

例如，记录：（1和2）和（4和5）被视为重复记录，因为它们具有相同的 set 条目。

因此，重复数据删除的一个示例输出可能是：

1         111   222
3         555   444
4         000   444

如何在SQL中最好地实现这一点？

Answer 1

查询：

<强> SQLFIDDLExample

SELECT
t1.*
FROM MyTable t1
  LEFT JOIN MyTable t2
    ON t1.Key1 = t2.Key2
      AND t1.Key2 = t2.Key1
      AND t1.RecordID>t2.RecordID
WHERE t2.RecordID is null

结果：

| RECORDID | KEY1 | KEY2 |
--------------------------
|        1 |  111 |  222 |
|        3 |  555 |  444 |
|        4 |    0 |  444 |

Answer 2

这应该得到dupe'd记录：

SELECT alt.RecordID
FROM yourtable AS main
INNER JOIN yourtable AS alt ON (main.Key1 = alt.Key2) AND (main.Key2 = alt.Key1)

Answer 3

要获得您描述的结果，您可以尝试：

SELECT M1.RecordId
FROM MyTable M1
WHERE Key1 IN (
    SELECT Key2 
    FROM MyTable
    WHERE Key2 = M1.Key1 AND RecordID <> M1.RecordId
    )
    AND
    Key2 IN (
    SELECT Key1
    FROM MyTable
    WHERE Key1 = M1.Key2 AND RecordID <> M1.RecordId
    )

要删除记录，只需使用RecordId ...

类似的东西：

DELETE MyTable WHERE RecordId IN ([Previous Query])

获得DEDUPLICATE：

SELECT  M1.RecordID
FROM    MyTable AS M1 
        LEFT JOIN MyTable AS MK1 ON M1.Key1 = MK1.Key2 AND M1.RecordID <> MK1.RecordID AND MK1.RecordId > M1.RecordId
        LEFT JOIN MyTable AS MK2 ON M1.Key2 = MK2.Key1 AND M1.RecordID <> MK2.RecordID AND MK2.RecordId > M1.RecordId
WHERE MK2.Key1 IS NOT NULL

:)感到误解了你

SQL将重复项删除为多列的集合

3 个答案: