mysql group by组内排序方法

痛定思痛。 2022-05-29 10:28 234阅读 0赞

#                                mysql group by组内排序方法 #

mysql的group by语法可以根据指定的规则对数据进行分组，分组就是将一个数据集划分成若干个小区域，然后再针对若干个小区域进行数据处理。本文将介绍mysql使用group by分组时，实现组内排序的方法。

mysql的group by语法可以对数据进行分组，但是分组后的数据并不能进行组内排序。  
  
  
例如一个评论表有多个用户评论，需要获取每个用户最后评论的内容。

创建测试数据表及数据

CREATE TABLE `comment` ( `id` int(10) unsigned NOT NULL AUTO_INCREMENT, `user_id` int(10) unsigned NOT NULL, `content` varchar(200) NOT NULL, `addtime` datetime NOT NULL, `lastmodify` datetime NOT NULL, PRIMARY KEY (`id`), KEY `user_id` (`user_id`), KEY `addtime` (`addtime`), KEY `uid_addtime` (`user_id`,`addtime`) ) ENGINE=InnoDB DEFAULT CHARSET=utf8;
    
    INSERT INTO `comment` (`id`, `user_id`, `content`, `addtime`, `lastmodify`) VALUES (1, 1, '评论1', '2017-05-17 00:00:00', '2017-05-17 00:00:00'), (2, 1, '评论2', '2017-05-17 00:00:01', '2017-05-17 00:00:01'), (3, 2, '评论1', '2017-05-17 00:00:02', '2017-05-17 00:00:02'), (4, 2, '评论2', '2017-05-17 00:00:03', '2017-05-17 00:00:03'), (5, 3, '评论1', '2017-05-17 00:00:04', '2017-05-17 00:00:04'), (6, 1, '评论3', '2017-05-17 00:00:05', '2017-05-17 00:00:05'), (7, 4, '评论1', '2017-05-17 00:00:06', '2017-05-17 00:00:06'), (8, 4, '评论2', '2017-05-17 00:00:07', '2017-05-17 00:00:07'), (9, 4, '评论3', '2017-05-17 00:00:08', '2017-05-17 00:00:08'), (10, 4, '评论4', '2017-05-17 00:00:09', '2017-05-17 00:00:09'), (11, 3, '评论2', '2017-05-17 00:00:10', '2017-05-17 00:00:10');
    
    select * from comment;
    +----+---------+---------+---------------------+---------------------+
    | id | user_id | content | addtime             | lastmodify          |
    +----+---------+---------+---------------------+---------------------+
    |  1 |       1 | 评论1   | 2017-05-17 00:00:00 | 2017-05-17 00:00:00 |
    |  2 |       1 | 评论2   | 2017-05-17 00:00:01 | 2017-05-17 00:00:01 |
    |  3 |       2 | 评论1   | 2017-05-17 00:00:02 | 2017-05-17 00:00:02 |
    |  4 |       2 | 评论2   | 2017-05-17 00:00:03 | 2017-05-17 00:00:03 |
    |  5 |       3 | 评论1   | 2017-05-17 00:00:04 | 2017-05-17 00:00:04 |
    |  6 |       1 | 评论3   | 2017-05-17 00:00:05 | 2017-05-17 00:00:05 |
    |  7 |       4 | 评论1   | 2017-05-17 00:00:06 | 2017-05-17 00:00:06 |
    |  8 |       4 | 评论2   | 2017-05-17 00:00:07 | 2017-05-17 00:00:07 |
    |  9 |       4 | 评论3   | 2017-05-17 00:00:08 | 2017-05-17 00:00:08 |
    | 10 |       4 | 评论4   | 2017-05-17 00:00:09 | 2017-05-17 00:00:09 |
    | 11 |       3 | 评论2   | 2017-05-17 00:00:10 | 2017-05-17 00:00:10 |
    +----+---------+---------+---------------------+---------------------+
      
       
        1
        2
        3
        4
        五
        6
        7
        8
        9
        10
        11
        12
        13
        14
        15
        16
        17
        18
        19
        20
        21
        22
        23
        24
        25
        26
        27
        28
        29
        三十
        31
        32
        33
        34
        35
        36
        37
        38
        39
        40
        41

在发表意见表中，每个用户最后评论的内容就是ID为6,4,11,10的记录。

通过查询使用组

select * from comment group by user_id; +----+---------+---------+---------------------+---------------------+
    | id | user_id | content | addtime | lastmodify | +----+---------+---------+---------------------+---------------------+
    |  1 |       1 | 评论1   | 2017-05-17 00:00:00 | 2017-05-17 00:00:00 |
    |  3 |       2 | 评论1   | 2017-05-17 00:00:02 | 2017-05-17 00:00:02 |
    |  5 |       3 | 评论1   | 2017-05-17 00:00:04 | 2017-05-17 00:00:04 |
    | 7 | 4 | 评论1 | 2017-05-17 00:00:06 | 2017-05-17 00:00:06 | +----+---------+---------+---------------------+---------------------+
      
       
        1
        2
        3
        4
        五
        6
        7
        8
        9

可以看到结果，分组后面只会返回分组内部的第一条数据。因为group by语法没有进行组内的排序的功能，只会按mysql默认的排序显示。  
  
  
如何才能对分组的内部数据进行排序了，这个需要根据不同的需求处理   
  
  
。1.id最大的，评论时间肯定最新

这种情况我们可以使用ID代替时间去搜寻并组内排序，使用MAX（ID），就可以获取到每个分组中最大的评论的ID（即最新的评论）

select * from comment where id in(select max(id) from comment group by user_id) order by user_id; +----+---------+---------+---------------------+---------------------+
    | id | user_id | content | addtime | lastmodify | +----+---------+---------+---------------------+---------------------+
    |  6 |       1 | 评论3   | 2017-05-17 00:00:05 | 2017-05-17 00:00:05 |
    |  4 |       2 | 评论2   | 2017-05-17 00:00:03 | 2017-05-17 00:00:03 |
    | 11 |       3 | 评论2   | 2017-05-17 00:00:10 | 2017-05-17 00:00:10 |
    | 10 | 4 | 评论4 | 2017-05-17 00:00:09 | 2017-05-17 00:00:09 | +----+---------+---------+---------------------+---------------------+
      
       
        1
        2
        3
        4
        五
        6
        7
        8
        9

2.id与评论时间没有关系，ID大的评论时间可能不是最新

这种情况我们就需要使用最大值（添加时间）来获取最新的评论，但因为不同用户的评论时间有可能相同，因此还需要加多USER\_ID这个条件去查询。

重新创建测试数据

truncate table comment;
    
    INSERT INTO `comment` (`id`, `user_id`, `content`, `addtime`, `lastmodify`) VALUES (1, 1, '评论1', '2017-05-17 00:00:00', '2017-05-17 00:00:00'), (2, 1, '评论2', '2017-05-17 00:10:01', '2017-05-17 00:10:01'), (3, 2, '评论1', '2017-05-17 00:10:02', '2017-05-17 00:10:02'), (4, 2, '评论2', '2017-05-17 00:00:03', '2017-05-17 00:00:03'), (5, 3, '评论1', '2017-05-17 00:10:04', '2017-05-17 00:10:04'), (6, 1, '评论3', '2017-05-17 00:00:05', '2017-05-17 00:00:05'), (7, 4, '评论1', '2017-05-17 00:00:06', '2017-05-17 00:00:06'), (8, 4, '评论2', '2017-05-17 00:10:07', '2017-05-17 00:10:07'), (9, 4, '评论3', '2017-05-17 00:00:08', '2017-05-17 00:00:08'), (10, 4, '评论4', '2017-05-17 00:00:09', '2017-05-17 00:00:09'), (11, 3, '评论2', '2017-05-17 00:00:10', '2017-05-17 00:00:10');
    
    select * from comment;
    +----+---------+---------+---------------------+---------------------+
    | id | user_id | content | addtime             | lastmodify          |
    +----+---------+---------+---------------------+---------------------+
    |  1 |       1 | 评论1   | 2017-05-17 00:00:00 | 2017-05-17 00:00:00 |
    |  2 |       1 | 评论2   | 2017-05-17 00:10:01 | 2017-05-17 00:10:01 |
    |  3 |       2 | 评论1   | 2017-05-17 00:10:02 | 2017-05-17 00:10:02 |
    |  4 |       2 | 评论2   | 2017-05-17 00:00:03 | 2017-05-17 00:00:03 |
    |  5 |       3 | 评论1   | 2017-05-17 00:10:04 | 2017-05-17 00:10:04 |
    |  6 |       1 | 评论3   | 2017-05-17 00:00:05 | 2017-05-17 00:00:05 |
    |  7 |       4 | 评论1   | 2017-05-17 00:00:06 | 2017-05-17 00:00:06 |
    |  8 |       4 | 评论2   | 2017-05-17 00:10:07 | 2017-05-17 00:10:07 |
    |  9 |       4 | 评论3   | 2017-05-17 00:00:08 | 2017-05-17 00:00:08 |
    | 10 |       4 | 评论4   | 2017-05-17 00:00:09 | 2017-05-17 00:00:09 |
    | 11 |       3 | 评论2   | 2017-05-17 00:00:10 | 2017-05-17 00:00:10 |
    +----+---------+---------+---------------------+---------------------+
      
       
        1
        2
        3
        4
        五
        6
        7
        8
        9
        10
        11
        12
        13
        14
        15
        16
        17
        18
        19
        20
        21
        22
        23
        24
        25
        26
        27
        28
        29
        三十
        31

符合条件的应该是ID为2,3,5,8的记录

select a.* from comment as a right join (select user_id, max(addtime) as maxtime from comment where user_id is not null group by user_id) as b on a.user_id=b.user_id and a.addtime=b.maxtime order by a.user_id asc;
    
    +------+---------+---------+---------------------+---------------------+
    | id   | user_id | content | addtime             | lastmodify          |
    +------+---------+---------+---------------------+---------------------+
    |    2 |       1 | 评论2   | 2017-05-17 00:10:01 | 2017-05-17 00:10:01 |
    |    3 |       2 | 评论1   | 2017-05-17 00:10:02 | 2017-05-17 00:10:02 |
    |    5 |       3 | 评论1   | 2017-05-17 00:10:04 | 2017-05-17 00:10:04 |
    |    8 |       4 | 评论2   | 2017-05-17 00:10:07 | 2017-05-17 00:10:07 |
    +------+---------+---------+---------------------+---------------------+
      
       
        1
        2
        3
        4
        五
        6
        7
        8
        9
        10
        11
        12

使用正确的加入可以减少。外层的数据集  
，其中user\_id是不是空可以使组由USER\_ID时使用索引。