整理删除重复数据的语句

--前期准备
SELECT ProviderGUID,bankaccounts,count(1) as bum INTO #temp1 from p_ProviderBanks  group by ProviderGUID,bankaccounts 
having count(1)>1

SELECT rowid=IDENTITY(INT,1,1),flag=0, providerguid,a.ProviderBankGUID,a.BankAccounts INTO #temp2 FROM p_ProviderBanks a WHERE EXISTS (SELECT * FROM #temp1 WHERE a.ProviderGUID=ProviderGUID AND  a.BankAccounts=BankAccounts)

SELECT TOP (10-1) * FROM #temp1
SELECT rowid=IDENTITY(INT,1,1),flag=0,providerguid,BankAccounts,bum INTO #temp3  FROM #temp1
SELECT * FROM #temp2
SELECT * FROM #temp3
DROP TABLE #temp3


--删除重复数据
BEGIN 
SET NOCOUNT ON 
DECLARE @num INT 
DECLARE @rowid INT 
DECLARE @providerguid UNIQUEIDENTIFIER
DECLARE @bankaccounts VARCHAR(400) 
SELECT @rowid=MIN(rowid) FROM #temp3 WHERE flag=0
DECLARE @strsql NVARCHAR(200)
--CREATE TABLE erpbak.dbo.tempdelete 
--(
--providerbankguid UNIQUEIDENTIFIER
--)
PRINT @rowid
WHILE @rowid is not NULL
BEGIN
    SELECT @providerguid=ProviderGUID,@bankaccounts=BankAccounts,@num=bum FROM #temp3 WHERE  rowid=@rowid
    SET @num=@num-1
    --PRINT @bankaccounts
    --PRINT @num
    SET @strsql=‘
    insert into erpbak.dbo.tempdelete 
    SELECT TOP ‘+CAST( @num AS VARCHAR(5) )+‘  ProviderBankGUID FROM #temp2 WHERE BankAccounts=‘‘‘+CAST(@bankaccounts AS VARCHAR(100))+‘‘‘ AND ProviderGUID=‘‘‘+CAST(@providerguid AS VARCHAR(100))+‘‘‘‘
    --PRINT @strsql
    EXEC sp_executesql @strsql
    UPDATE #temp3 SET flag=1 WHERE rowid=@rowid
    SELECT @rowid=MIN(rowid) FROM #temp3 WHERE flag=0
END
SET NOCOUNT OFF 
END