Oracle、SQLServer 刪除表中的重復數據，只保留一條記錄

原文地址：

https://blog.csdn.net/yangwenxue_admin/article/details/51742426

https://www.cnblogs.com/springsnow/p/10334469.html

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

****************Oracle 刪除表中的重復數據****************

一、重復記錄根據單個字段來判斷

1、首先，查找表中多余的重復記錄，重復記錄是根據單個字段（FIELD_CODE）來判斷

select * from R_RESOURCE_DETAILS where FIELD_CODE in
(select FIELD_CODE from R_RESOURCE_DETAILS group by FIELD_CODE having count(FIELD_CODE) >1)

2、刪除表中多余的重復記錄，重復記錄是根據單個字段（FIELD_CODE）來判斷，只留有rowid最小的記錄

delete from R_RESOURCE_DETAILS where (FIELD_CODE) in 
(select FIELD_CODE from R_RESOURCE_DETAILS group by FIELD_CODE having count(FIELD_CODE) >1) 
and rowid not in (select min(rowid) from R_RESOURCE_DETAILS group by FIELD_CODE having count(*)>1)

二、重復記錄根據多個字段來判斷

1、查找表中多余的重復記錄（多個字段）

select * from R_RESOURCE_DETAILS a where (a.FIELD_CODE,a.DTA_ITEM_NAME) in
(select FIELD_CODE,DTA_ITEM_NAME from R_RESOURCE_DETAILS group by FIELD_CODE,DTA_ITEM_NAME having count(*) > 1)

2、刪除表中多余的重復記錄（多個字段），只留有rowid最小的記錄

delete from R_RESOURCE_DETAILS a where (a.FIELD_CODE,a.DTA_ITEM_NAME) in 
(select FIELD_CODE,DTA_ITEM_NAME from R_RESOURCE_DETAILS group by FIELD_CODE,DTA_ITEM_NAME having count(*) > 1) 
and 
rowid not in (select min(rowid) from R_RESOURCE_DETAILS group by FIELD_CODE,DTA_ITEM_NAME having count(*)>1)

3、查找表中多余的重復記錄（多個字段），不包含rowid最小的記錄

select * from R_RESOURCE_DETAILS a 
where (a.FIELD_CODE,a.DTA_ITEM_NAME) in 
(select FIELD_CODE,DTA_ITEM_NAME from R_RESOURCE_DETAILS group by FIELD_CODE,DTA_ITEM_NAME having count(*) > 1) 
and 
rowid not in (select min(rowid) from R_RESOURCE_DETAILS group by FIELD_CODE,DTA_ITEM_NAME having count(*)>1)

****************SQLServer 刪除表中的重復數據****************

1、查詢表中Name 重復的數據
2、有唯一列，通過唯一列最大或最小方式刪除重復記錄
3、無唯一列使用ROW_NUMBER()函數刪除重復記錄

1、查詢表中Name 重復的數據

select Name from Student group by  Name having count(Name) > 1

2、有唯一列，通過唯一列最大或最小方式刪除重復記錄

檢查表中是否有主鍵或者唯一值的列，當前可以數據看到ID是唯一的，可以通過Name分組排除掉ID最大或最小的行

delete from Student
  where Name in( select Name from Student group by  Name having count(Name) > 1) and 
 ID not in(select  max(ID) from Student group by  Name having count(Name) > 1 )

執行刪除腳本后查詢

3、無唯一列使用ROW_NUMBER()函數刪除重復記錄

如果表中沒有唯一值的列，可以通過 來刪除重復數據 
重復執行插入腳本，查看表數據，表中沒有唯一列值

Delete T From
 (Select Row_Number() Over(Partition By [Name] order By [ID]) As RowNumber,* From Student)T
 Where T.RowNumber > 1

小知識點

語法：ROW_NUMBER() OVER(PARTITION BY COLUMN ORDER BY COLUMN)
         表示根據COLUMN分組，在分組內部根據 COLUMN排序，而此函數計算的值就表示每組內部排序后的順序編號（組內連續的唯一的)
         函數“Row_Number”必須有 OVER 子句。OVER 子句必須有包含 ORDER BY
         Row_Number() Over(Partition By [Name] order By [ID])  表示已name列分組，在每組內以ID列進行升序排序，每組內返回一個唯一的序號

執行刪除腳本后查詢表數據