面试题: mysql 数据库去重

时间：2022-03-06 11:53

去重面试总结

前言：题目大概是这样的。

建表：

CREATE TABLE `test2` (
  `id` int(11) NOT NULL AUTO_INCREMENT,
  `peopleId` int(11) DEFAULT NULL,
  `name` varchar(255) DEFAULT NULL,
  PRIMARY KEY (`id`)
) ENGINE=InnoDB AUTO_INCREMENT=7 DEFAULT CHARSET=utf8;
 
-- ----------------------------
-- Records of test2
-- ----------------------------
INSERT INTO `test2` VALUES (‘1‘, ‘1‘, ‘倒一‘);
INSERT INTO `test2` VALUES (‘2‘, ‘1‘, ‘倒一‘);
INSERT INTO `test2` VALUES (‘3‘, ‘3‘, ‘等等‘);
INSERT INTO `test2` VALUES (‘4‘, ‘2‘, ‘421‘);
INSERT INTO `test2` VALUES (‘5‘, ‘2‘, ‘421‘);
INSERT INTO `test2` VALUES (‘6‘, ‘2‘, ‘421‘);

1、查找表中多余的重复记录，重复记录是根据单个字段（Id）来判断

网上答案：select * from 表 where Id in (select Id from 表 group byId having count(Id) > 1)

mysql：select * from test2 where id in (select id from test2 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（Id）来判断，只留有rowid最小的记录

网上答案：DELETE from 表 WHERE (id) IN ( SELECT id FROM 表 GROUP BY id HAVING COUNT(id) > 1) AND ROWID NOT IN (SELECT MIN(ROWID) FROM 表 GROUP BY id HAVING COUNT(*) > 1);

mysql：

DELETE FROM test2 WHERE peopleId IN
(  
   select a.peopleId FROM
     (
            SELECT * FROM test2
   GROUP BY peopleId  
   HAVING count(peopleId) > 1
        )a 
)
AND id NOT IN (  
        select b.id FROM
        (
            SELECT * FROM test2 
    GROUP BY name 
    HAVING count(name) > 1  
        )b
 
)

3、查找表中多余的重复记录（多个字段）

select * from 表 a where (a.Id,a.seq) in(select Id,seq from 表 group by Id,seq having count(*) > 1)

SELECT
    *
FROM
    test3 a
WHERE
    (a.id, a.seq) IN (
        SELECT
            id,
            seq
        FROM
            test3
        GROUP BY
            id,
            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 表 a where (a.Id,a.seq) in (select Id,seq from 表 group by Id,seq having count() > 1) and rowid not in (select min(rowid) from 表 group by Id,seq having count()>1)

DELETE FROM
    test3 
WHERE
    (id, seq) IN (
        SELECT
            a.id,
            a.seq
        FROM
            (
                SELECT
                    id,
                    seq
                FROM
                    test3
                GROUP BY
                    id,
                    seq
                HAVING
                    count(*) > 1
            ) a
    )
AND (id, seq, `name`) NOT IN (
    SELECT
        b.*
    FROM
        (
            SELECT
                *
            FROM
                test3
            GROUP BY
                id,
                seq
            HAVING
                count(*) > 1
        ) b
)

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 表 a where (a.Id,a.seq) in (select Id,seq from 表 group by Id,seq having count() > 1) and rowid not in (select min(rowid) from 表 group by Id,seq having count()>1)

select * FROM
    test3 
WHERE
    (id, seq) IN (
        SELECT
            a.id,
            a.seq
        FROM
            (
                SELECT
                    id,
                    seq
                FROM
                    test3
                GROUP BY
                    id,
                    seq
                HAVING
                    count(*) > 1
            ) a
    )
AND (id, seq, `name`) NOT IN (
    SELECT
        b.*
    FROM
        (
            SELECT
                *
            FROM
                test3
            GROUP BY
                id,
                seq
            HAVING
                count(*) > 1
        ) b
)

胜负查询：

CREATE TABLE `t_game` (
  `game_date` varchar(255) DEFAULT NULL,
  `game_res` varchar(255) DEFAULT NULL
) ENGINE=InnoDB DEFAULT CHARSET=utf8;
 
-- ----------------------------
-- Records of t_game
-- ----------------------------
INSERT INTO `t_game` VALUES (‘2018-03-20‘, ‘胜‘);
INSERT INTO `t_game` VALUES (‘2018-03-20‘, ‘胜‘);
INSERT INTO `t_game` VALUES (‘2018-03-20‘, ‘负‘);
INSERT INTO `t_game` VALUES (‘2018-03-21‘, ‘负‘);
INSERT INTO `t_game` VALUES (‘2018-03-21‘, ‘胜‘);
INSERT INTO `t_game` VALUES (‘2018-03-21‘, ‘负‘);

select game_date,(select count(*) from t_game where game_date = t.game_date and game_res = ‘胜‘) as ‘胜‘
,(select count(*) from t_game where game_date = t.game_date and game_res = ‘负‘) as ‘负‘
 from t_game as t group by game_date;