SQL SERVER按多字段查找重复的数据并删除只保留一条

由于一次操作失误，给表中插入了多条重复的数据，所以就需要删除重复的数据只保留一条，一时想不到好方法，各种查资料，终于找到了，特意写到这里，方便以后自己用~

查询：

select A.n_PatentID,A.s_OtherFileName,A.s_FileName FROM TB_OtherFiles A,
(SELECT n_PatentID,s_OtherFileName,s_FileName FROM TB_OtherFiles
GROUP BY n_PatentID,s_OtherFileName,s_FileName HAVING COUNT(1)>=2
) AS B

WHERE A.s_OtherFileName=B.s_OtherFileName AND A.n_PatentID=B.n_PatentID AND A.s_FileName=B.s_FileName AND A.n_PatentID >0
order by n_PatentID

删除：

DELETE T FROM (SELECT n_PatentID,s_OtherFileName,s_FileName,s_recycle, ROW_NUMBER()OVER(PARTITION BY n_PatentID,s_OtherFileName,s_FileName ORDER BY RAND()) AS RN FROM TB_OtherFiles where n_PatentID>0) AS t WHERE RN>1