SQL语句 删除表user 中字段name 内容重复的记录，A表 ，B表 字段相同 有关联关系的表 对表两个表中的数据

今天去面试！我给大意了，直接写了个删除语句！又中招了。。。

在网上找了半天也没找到合适的代码！

回来想了想才知道自已太大意了！想了良久，想出了这么个办法，

insert into fuck values(user_id.nextval,'liang','123');

select * from fuck

delete from fuck
 where id in (select id
                from (select t.*, rownum r from fuck t where name = 'liang') 
               where r > 1);

今天又想了想，还是不行呀！只能指定值！一不做二不休，又是几个小时的研究,他奶奶的，功夫不负有心上，终于思考出一个思路，并成功！也许有更简单的写法，

就是用第三张临时表，没有测试过，还有一种是很复杂的查询语句，方法很多种吧

declare 
v_sql varchar2(500);
v_tempName varchar2(30);

cursor t_name is/*提取有重得的名字*/
   select name from fuck  group by name having count(name) >1;
   
begin
v_sql:='delete from fuck  
                where id in (select id  
                from (select t.*, rownum r from fuck t where name = :1)   
                where r > 1) ';
 
  for t_names in t_name loop /*循环读取并删除*/
   execute immediate v_sql using t_names.name;
   dbms_output.put_line('成功删除重复数据: '||t_names.name);
   end loop;
end;

方式二：

insert into fucks(id,name) values(1,'aa');
insert into fucks(id,name) values(1,'aa');
insert into fucks(id,name) values(2,'aa');
insert into fucks(id,name) values(2,'aa');
insert into fucks(id,name) values(3,'vv');
insert into fucks(id,name) values(3,'vv');
insert into fucks(id,name) values(8,'bbb');




/***********创建临时表 并查询出数据不重复的字段***************/
create global temporary table tempTable on commit preserve rows 
as select distinct id,name from fucks;
/***********删除原有的表***************/
drop table fucks;
/***********把监时表插入到  新建原有表***************/
create  table fucks as select * from tempTable;


drop table tempTable;--需要重新登陆才能删除临时表


select * from fucks;
select * from tempTable;


--方式三
delete from fucks f1
 where  rowid not in (select min(rowid)
                   from fucks f2
                  where f1.id = f2.id
                    and f1.name = f2.name)


/***********************************字段相同  有关联关系的表   对表两个表中的数据***********************/
create table newfucks as select *  from fucks


select * from newfucks;


select * from newfucks a join fucks b on a.id=b.id;


select * from newfucks a join fucks b on a.id=b.id and a.name=b.name;


insert into newfucks(id,name) values(8,'hdddd');


/***********************************查询a表 条件 a和b表不相同的数据***********************/
select *
  from newfucks a
 where a.id not in (select id  from fucks)
   or a.name not in (select name from fucks);
   
 --方试二 进行优化  
select *  from newfucks a
 where not exists (select 1  from fucks b where a.name=b.name and a.id=b.id)


   
   
   
   

--去掉重复数据 保留最后插入id 方法很多，还得学习呀...

--按名称分组 之后 按id 排序  再以每个组生成序列号
 select id,name,row_number() over(partition  by name order by id desc) fn from fucks
 
 --取第一个序号 最后生成的id 
 select * from (
  select id,name,row_number() over(partition  by name order by id desc) fn from fucks
 ) where fn=1;
 
 --savepoint a1;
 
 --delete
 delete from fucks where id not in(
  select id from (
  select id,name,row_number() over(partition  by name order by id desc) fn from fucks
 ) where fn=1
 );