Mysql数据库中查询重复数据和去重数据 , 删除重复数据的sql及分析

数据库中有重复数据时,用到哪些sql语句?
建表:

CREATE TABLE `user` (
  `id` bigint(255) NOT NULL AUTO_INCREMENT,
  `name` varchar(20) COLLATE utf8mb4_general_ci NOT NULL DEFAULT '' COMMENT '名称',
  `age` int(2) NOT NULL DEFAULT '0',
  PRIMARY KEY (`id`)
) ENGINE=InnoDB AUTO_INCREMENT=5 DEFAULT CHARSET=utf8mb4 COLLATE=utf8mb4_general_ci;

这里有若干数据,并掺杂了重复数据

1. 查看过滤重复后的数据

思路, group by 分组可以对多个列进行分组, 分组后可以过滤掉重复的数据
这里在mysql5.7以上版本会报错,因为不支持select那些group by和聚合函数之外的字段
sql语句:

SELECT id,`name`,age,count(1)
	FROM user GROUP BY `name`,age

这里要么把id去掉,要么选择临时方案:

set @@global.sql_mode ='STRICT_TRANS_TABLES,NO_ZERO_IN_DATE,NO_ZERO_DATE,ERROR_FOR_DIVISION_BY_ZERO,NO_ENGINE_SUBSTITUTION';
set @@SESSION.sql_mode ='STRICT_TRANS_TABLES,NO_ZERO_IN_DATE,NO_ZERO_DATE,ERROR_FOR_DIVISION_BY_ZERO,NO_ENGINE_SUBSTITUTION';

SELECT id,`name`,age,count(1)
	FROM user GROUP BY `name`,age;

2. 查看重复的数据

刚刚的语句已经把每个组对应的count数查询出来了,那么count>1的自然是重复的数据

SELECT id,`name`,age,count(1) as c
	FROM user GROUP BY `name`,age having c > 1

3. 删除重复的数据留下一条

思路: 刚刚已经把重复的数据查询出来了,包括id, 那么查询出每个重复组中的唯一一个id,也就是x,就可以delete … id not in (x)

上面说虽然5.7以上版本默认不支持查询group by 以外的字段,比如id,但是聚合函数还是可以的
子语句1:

SELECT MIN(id) FROM user
			GROUP BY name,age 

查询出来的id就是我们需要留下的不重复的数据的id

按理来说只要:
delete from user where id not in 子语句1

DELETE FROM user
	WHERE id NOT IN (
		SELECT MIN(id) FROM user
			GROUP BY name,age 
	)

但是报错了

DELETE FROM user
	WHERE id NOT IN (
		SELECT MIN(id) FROM user
			GROUP BY name,age 
	)
> 1093 - You can't specify target table 'user' for update in FROM clause
> 时间: 0.007s

因为在mysql中,不能在一条Sql语句中,即查询这些数据,同时修改这些数据

解决方法：select的结果再通过一个中间表temp进行select多一次，就可以避免这个错误

DELETE FROM user
	WHERE id NOT IN (
		SELECT temp.min_id FROM (
			SELECT MIN(id) min_id FROM user
				GROUP BY name,age
			)AS temp
	);
select * from user;
	

删除成功: