sql建立唯一的非聚集索引时由于数据重复建立失败，快速删除重复数据的 巧妙 sql

 

巧妙的想法

 

 

select * from
(
select max(weizi) as weizi,max(id) as id,count(1) as num from

(select  concat(biz,'_',appmsgid,'_',position) as weizi,id  from article) dd  group by weizi 

) fff

where num>1