MySQL 数据库中删除重复记录的方法总结
演示数据
表结构：
mysql> desc demo;
+-------+------------------+------+-----+---------+----------------+
| Field | Type | Null | Key | Default | Extra |
+-------+------------------+------+-----+---------+----------------+
| id | int(11) unsigned | NO | PRI | NULL | auto_increment |
| site | varchar(100) | NO | MUL | | |
+-------+------------------+------+-----+---------+----------------+
2 rows in set (0.00 sec)
数据：
mysql> select * from demo order by id;
+----+------------------------+
| id | site |
+----+------------------------+
| 1 | http://www.CodeBit.cn |
| 2 | http://YITU.org |
| 3 | http://www.ShuoWen.org |
| 4 | http://www.CodeBit.cn |
| 5 | http://www.ShuoWen.org |
+----+------------------------+
5 rows in set (0.00 sec)
当没有创建表或创建索引权限的时候，可以用下面的方法：
如果你要删除较旧的重复记录，可以使用下面的语句：
mysql> delete from a
-> using demo as a, demo as b
-> where (a.id > b.id)
-> and (a.site = b.site);
Query OK, 2 rows affected (0.12 sec)
mysql> select * from demo order by id;
+----+------------------------+
| id | site |
+----+------------------------+
| 1 | http://www.CodeBit.cn |
| 2 | http://YITU.org |
| 3 | http://www.ShuoWen.org |
+----+------------------------+
3 rows in set (0.00 sec)
如果你要删除较新的重复记录，可以使用下面的语句：
mysql> delete from a
-> using demo as a, demo as b
-> where (a.id < b.id)
-> and (a.site = b.site);
Query OK, 2 rows affected (0.12 sec)
mysql> select * from demo order by id;
+----+------------------------+
| id | site |
+----+------------------------+
| 2 | http://YITU.org |
| 4 | http://www.CodeBit.cn |
| 5 | http://www.ShuoWen.org |
+----+------------------------+
3 rows in set (0.00 sec)
你可以用下面的语句先确认将被删除的重复记录：
mysql> SELECT a.*
-> FROM demo a, demo b
-> WHERE a.id > b.id
-> AND (a.site = b.site);
+----+------------------------+
| id | site |
+----+------------------------+
| 1 | http://www.CodeBit.cn |
| 3 | http://www.ShuoWen.org |
+----+------------------------+
2 rows in set (0.00 sec)
如果有创建索引的权限，可以用下面的方法：
在表上创建唯一键索引：
mysql> alter ignore table demo add unique index ukey (site);
Query OK, 5 rows affected (0.46 sec)
Records: 5 Duplicates: 2 Warnings: 0
mysql> select * from demo order by id;
+----+------------------------+
| id | site |
+----+------------------------+
| 1 | http://www.CodeBit.cn |
| 2 | http://YITU.org |
| 3 | http://www.ShuoWen.org |
+----+------------------------+
3 rows in set (0.00 sec)
重复记录被删除后，如果需要，可以删除索引：
mysql> alter table demo drop index ukey;
Query OK, 3 rows affected (0.37 sec)
Records: 3 Duplicates: 0 Warnings: 0
如果有创建表的权限，可以用下面的方法：
创建一个新表，然后将原表中不重复的数据插入新表：
mysql> create table demo_new as select * from demo group by site;
Query OK, 3 rows affected (0.19 sec)
Records: 3 Duplicates: 0 Warnings: 0
mysql> show tables;
+----------------+
| Tables_in_test |
+----------------+
| demo |
| demo_new |
+----------------+
2 rows in set (0.00 sec)
mysql> select * from demo order by id;
+----+------------------------+
| id | site |
+----+------------------------+
| 1 | http://www.CodeBit.cn |
| 2 | http://YITU.org |
| 3 | http://www.ShuoWen.org |
| 4 | http://www.CodeBit.cn |
| 5 | http://www.ShuoWen.org |
+----+------------------------+
5 rows in set (0.00 sec)
mysql> select * from demo_new order by id;
+----+------------------------+
| id | site |
+----+------------------------+
| 1 | http://www.CodeBit.cn |
| 2 | http://YITU.org |
| 3 | http://www.ShuoWen.org |
+----+------------------------+
3 rows in set (0.00 sec)
然后将原表备份，将新表重命名为当前表：
mysql> rename table demo to demo_old, demo_new to demo;
Query OK, 0 rows affected (0.04 sec)
mysql> show tables;
+----------------+
| Tables_in_test |
+----------------+
| demo |
| demo_old |
+----------------+
2 rows in set (0.00 sec)
mysql> select * from demo order by id;
+----+------------------------+
| id | site |
+----+------------------------+
| 1 | http://www.CodeBit.cn |
| 2 | http://YITU.org |
| 3 | http://www.ShuoWen.org |
+----+------------------------+
3 rows in set (0.00 sec)
注意：使用这种方式创建的表会丢失原表的索引信息！
mysql> desc demo;
+-------+------------------+------+-----+---------+-------+
| Field | Type | Null | Key | Default | Extra |
+-------+------------------+------+-----+---------+-------+
| id | int(11) unsigned | NO | | 0 | |
| site | varchar(100) | NO | | | |
+-------+------------------+------+-----+---------+-------+
2 rows in set (0.00 sec)
如果要保持和原表信息一致，你可以使用 show create table demo; 来查看原表的创建语句，然后使用原表的创建语句创建新表，接着使用 insert … select 语句插入数据，再重命名表即可。
