SQL删除重复记录

2015-01-01 walter lee 更多博文 » 博客 » GitHub »

sql

原文链接 http://www.xiangguo.li/sql_and_nosql/2015/01/01/sql
注：以下为加速网络访问所做的原文缓存，经过重新格式化，可能存在格式方面的问题，或偶有遗漏信息，请以原文为准。

{% include JB/setup %} 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 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)