select * from hengtu_demandpush a
where (a.did,a.mid) in (select did,mid from hengtu_demandpush group by did,mid having count(*) > 1)
或
select * from hengtu_demandpush group by did,mid having count(*)>1
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 peopleId in (select peopleId from people group by peopleId having count(peopleId) > 1)
- and rowid not in (select min(rowid) from people group by peopleId having count(peopleId )>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)