按指定列去重並求各值的最後一條記錄

按指定列去重並求各值的最後一條記錄

轉自:http://wubx.net/mutli-rows-get-last-one/

在工作中經常遇到這樣的需求，一張表中某些列值是不唯一的，但是想拿到該列每個唯一記錄的最行的記錄。
這麼辦？分組？中間表？關聯查詢？還是看看下面這個簡單的方法吧。

1、創建測試表

mysql> CREATE TABLE `test` (
    -> `userid` int(11) DEFAULT NULL,
    -> `atime` datetime DEFAULT NULL,
    -> KEY `idx_userid` (`userid`)
    -> ) ENGINE=InnoDB DEFAULT CHARSET=utf8;
Query OK, 0 rows affected (0.01 sec)

2、插入測試數據

mysql> insert into test (userid,atime) values
    -> (1,'2013-08-12 11:05:25'),
    -> (2,'2013-08-12 11:05:29'),
    -> (3,'2013-08-12 11:05:32'),
    -> (5,'2013-08-12 11:05:34'),
    -> (1,'2013-08-12 12:05:40'),
    -> (2,'2013-08-12 12:05:43'),
    -> (3,'2013-08-12 12:05:48'),
    -> (5,'2013-08-12 12:06:03');
Query OK, 8 rows affected (0.00 sec)
Records: 8  Duplicates: 0  Warnings: 0

3、查看數據

mysql> select * from test ;
+--------+---------------------+
| userid | atime               |
+--------+---------------------+
|      1 | 2013-08-12 11:05:25 |
|      2 | 2013-08-12 11:05:29 |
|      3 | 2013-08-12 11:05:32 |
|      5 | 2013-08-12 11:05:34 |
|      1 | 2013-08-12 12:05:40 |
|      2 | 2013-08-12 12:05:43 |
|      3 | 2013-08-12 12:05:48 |
|      5 | 2013-08-12 12:06:03 |
+--------+---------------------+
8 rows in set (0.00 sec)

可以看到表中userid不唯一，現在要取每個userid對應的最後一條atime記錄。
給一個簡方法：

mysql> select userid,substring_index(group_concat(atime order by atime desc),",",1) as atime from test group by userid;
+--------+---------------------+
| userid | atime               |
+--------+---------------------+
|      1 | 2013-08-12 12:05:40 |
|      2 | 2013-08-12 12:05:43 |
|      3 | 2013-08-12 12:05:48 |
|      5 | 2013-08-12 12:06:03 |
+--------+---------------------+
4 rows in set (0.00 sec)

4、結果分析

上面的結果是不是很神奇？
這裏巧用了substring_index字符串截取函數和group_concat字符串分組拼接函數；分步驟查看就一目瞭然了:

1）group_concat函數作用

mysql> select userid,group_concat(atime order by atime desc) as atime from  test group by userid;
+--------+-----------------------------------------+
| userid | atime                                   |
+--------+-----------------------------------------+
|      1 | 2013-08-12 12:05:40,2013-08-12 11:05:25 |
|      2 | 2013-08-12 12:05:43,2013-08-12 11:05:29 |
|      3 | 2013-08-12 12:05:48,2013-08-12 11:05:32 |
|      5 | 2013-08-12 12:06:03,2013-08-12 11:05:34 |
+--------+-----------------------------------------+
4 rows in set (0.00 sec)

2）substring_index函數作用

mysql> select userid,substring_index(group_concat(atime order by atime desc),",",1) as atime from test group by userid;
+--------+---------------------+
| userid | atime               |
+--------+---------------------+
|      1 | 2013-08-12 12:05:40 |
|      2 | 2013-08-12 12:05:43 |
|      3 | 2013-08-12 12:05:48 |
|      5 | 2013-08-12 12:06:03 |
+--------+---------------------+
4 rows in set (0.00 sec)

完畢!