[转]MySQL 【去重留一】一条sql语句完成 思路总结

最后在一个技术群里得到了完美的答案,看这条sql语句:

DELETE consum_record
FROM
    consum_record, 
    (
        SELECT
            min(id) id,
            user_id,
            monetary,
            consume_time
        FROM
            consum_record
        GROUP BY
            user_id,
            monetary,
            consume_time
        HAVING
            count(*) > 1
    ) t2
WHERE
    consum_record.user_id = t2.user_id 
    and consum_record.monetary = t2.monetary
    and consum_record.consume_time  = t2.consume_time
AND consum_record.id > t2.id;

上面这条sql语句,仔细看一下,揣摩出思路也不难,大概也分为3步来理解:

  1. (SELECT min(id) id, user_id, monetary, consume_time FROM consum_record GROUP BY user_id, monetary, consume_time HAVING count(*) > 1 ) t2 查询出重复记录形成一个集合(临时表t2),集合里是每种重复记录的最小ID
  2. consum_record.user_id = t2.user_id and consum_record.monetary = t2.monetary and consum_record.consume_time = t2.consume_time 关联判断重复基准的字段
  3. 根据条件,删除原表中id大于t2中id的记录

看到这个语句的时候,心里想这也太厉害了。这么一个简单的sql语句,竟然可以解决这么复杂的问题,涨姿势了~
运行起来也超级快,原先的代码循环执行,需要116s左右,而这里0.3s就可以了,厉害了~

perfect_sql.png

perfect_sql.png

发表评论

电子邮件地址不会被公开。 必填项已用*标注


*