SQL删除重复数据只保留一条

在几千条记录里,存在着些相同的记录,如何能用SQL语句,删除掉重复的呢
1、查找表中多余的重复记录,重复记录是根据单个字段(peopleId)来判断 
select * from people 
where peopleId in (select peopleId from people group by peopleId having count(peopleId) > 1) 

2、删除表中多余的重复记录,重复记录是根据单个字段(peopleId)来判断,只留有rowid最小的记录 
delete from people 
where   peopleName in (select peopleName    from people group by peopleName      having count(peopleName) > 1) 
and   peopleId not in (select min(peopleId) from people group by peopleName     having count(peopleName)>1) 

3、查找表中多余的重复记录(多个字段) 
select * from vitae a 
where (a.peopleId,a.seq) in (select peopleId,seq from vitae group by peopleId,seq having count(*) > 1) 

4、删除表中多余的重复记录(多个字段),只留有rowid最小的记录 
delete from vitae a 
where (a.peopleId,a.seq) in (select peopleId,seq from vitae group by peopleId,seq having count(*) > 1) 
and rowid not in (select min(rowid) from vitae group by peopleId,seq having count(*)>1) 

5、查找表中多余的重复记录(多个字段),不包含rowid最小的记录 
select * from vitae a 
where (a.peopleId,a.seq) in (select peopleId,seq from vitae group by peopleId,seq having count(*) > 1) 
and rowid not in (select min(rowid) from vitae group by peopleId,seq having count(*)>1)   

6.消除一个字段的左边的第一位:

update tableName set [Title]=Right([Title],(len([Title])-1)) where Title like ‘村%‘

7.消除一个字段的右边的第一位:

update tableName set [Title]=left([Title],(len([Title])-1)) where Title like ‘%村‘

8.假删除表中多余的重复记录(多个字段),不包含rowid最小的记录 
update vitae set ispass=-1
where peopleId in (select peopleId from vitae group by peopleId

郑重声明:本站内容如果来自互联网及其他传播媒体,其版权均属原媒体及文章作者所有。转载目的在于传递更多信息及用于网络分享,并不代表本站赞同其观点和对其真实性负责,也不构成任何其他建议。