android中 *** 作数据库sql语句，联合唯一索引，重复则替换，怎么修改这句sq

不会是id都相同吧

select A from 表 group by A having count(A)>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)

1、使用唯一性约束，不过如果是在事务中批量提交时，一个失败将导致整个事务rollback。

2、先用select查询数据是否重复，再决定是否要插入此条数据，实现起来比较麻烦，特别是字段较多时，效率方面未对比测试。

3、使用语句insert or ignore into table (fields) values (values)或replace into table (fields) values (values)此法比较简洁。需要注意的是，当表有一个PRIMARY KEY或UNIQUE索引才有意义。

可以使用 UNIQUE 约束确保在非主键列中不输入重复的值.尽管 UNIQUE 约束和 PRIMARY KEY 约束都强制唯一性,但想要强制一列或多列组合（不是主键）的唯一性时应使用 UNIQUE 约束而不是 PRIMARY KEY 约束.

可以对一个表定义多个 UNIQUE 约束,但只能定义一个 PRIMARY KEY 约束.

而且,UNIQUE 约束允许 NULL 值,这一点与 PRIMARY KEY 约束不同.不过,当与参与 UNIQUE 约束的任何值一起使用时,每列只允许一个空值.

FOREIGN KEY 约束可以引用 UNIQUE 约束.

---