相信這個問題也會困擾很多人,尤爲是使用過其它數據庫(如Oracle)的人,以前我也沒有太在乎這個問題,再加上一些書籍和網上的文章講的不夠細緻,又沒測試過,致使我一直理解錯誤。下面經過實例來解釋,在開始以前先簡單介紹下字符和字節的區別。
字符 人們使用的記號,抽象意義上的一個符號。一個漢字和英文就是一個字符,如'1', '中', 'a', '$', '¥',……
字節 計算機中存儲數據的單元,一個8位的二進制數,是一個很具體的存儲空間,如0x01, 0x45, 0xFA,……
詳細的能夠參考http://www.regexlab.com/zh/encoding.htm,正確理解相關概念仍是很重要的。
好了,若是你已經理解了字符和字節的區別,下面就容易多了。
環境:CentOS5.x,MySQL5.1.x
UTF8字符集下:
SQL>create table test(id int auto_increment,name varchar(10),primary key(id));
SQL>insert into test values(null,'1234567890');
Query OK, 1 row affected (0.00 sec)
SQL>insert into test values(null,'一二三四五六七八九十');
Query OK, 1 row affected (0.00 sec)
SQL>insert into test values(null,'abcdefghig');
Query OK, 1 row affected (0.01 sec)
SQL>insert into test values(null,12345678901);
ERROR 1406 (22001): Data too long for column 'name' at row 1
SQL>insert into test values(null,'一二三四五六七八九十1');
ERROR 1406 (22001): Data too long for column 'name' at row 1
SQL>insert into test values(null,'一二三四五六七八九十一');
ERROR 1406 (22001): Data too long for column 'name' at row 1
SQL>select id,name,length(name),char_length(name) from test;數據庫
+----+--------------------------------+--------------+-------------------+
| id | name | length(name) | char_length(name) |
+----+--------------------------------+--------------+-------------------+
| 1 | 1234567890 | 10 | 10 |
| 2 | 一二三四五六七八九十 | 30 | 10 |
| 3 | abcdefghig | 10 | 10 |
+----+--------------------------------+--------------+-------------------+
3 rows in set (0.00 sec)
GBK字符集下:
SQL>create table test(id int auto_increment,name varchar(10),primary key(id));
SQL>insert into test values(null,'1234567890');
Query OK, 1 row affected (0.00 sec)
SQL>insert into test values(null,'一二三四五六七八九十');
Query OK, 1 row affected (0.00 sec)
SQL>insert into test values(null,'abcdefghig');
Query OK, 1 row affected (0.01 sec)
SQL>insert into test values(null,12345678901);
ERROR 1406 (22001): Data too long for column 'name' at row 1
SQL>insert into test values(null,'一二三四五六七八九十1');
ERROR 1406 (22001): Data too long for column 'name' at row 1
SQL>insert into test values(null,'一二三四五六七八九十一');
ERROR 1406 (22001): Data too long for column 'name' at row 1
SQL>select id,name,length(name),char_length(name) from test;測試
+----+----------------------+--------------+-------------------+
| id | name | length(name) | char_length(name) |
+----+----------------------+--------------+-------------------+
| 1 | 1234567890 | 10 | 10 |
| 2 | 一二三四五六七八九十 | 20 | 10 |
| 3 | abcdefghig | 10 | 10 |
+----+----------------------+--------------+-------------------+
3 rows in set (0.00 sec)
因而可知,varchar定義的長度的單位是字符,哪怕是1個多字節字符也是1個字符,如中文和英文字母都被看成1個字符來對待。
那麼varchar可以定義的最大長度是多少呢?這個和你當前所使用的字符集有關。拋開字符,其最大長度爲65535字節(這是最大行大小,由全部列共享),而放在不一樣的字符集下,可以定義的最大長度就會有所不一樣,如UTF8下是21845。聽說MySQL5中varchar的長度也爲字符,而MySQL4中的則爲字節,未經證明,感興趣的有環境能夠本身測下。
順便補充一下,char數據類型定義的長度也爲字符,其最大長度爲255。
SQL>create table test(id int auto_increment,name char(5),primary key(id));
Query OK, 0 rows affected (0.09 sec)
SQL>insert into test values(null,'123');
Query OK, 1 row affected (0.00 sec)
SQL>insert into test values(null,'12345');
Query OK, 1 row affected (0.00 sec)
SQL>insert into test values(null,'一二三');
Query OK, 1 row affected (0.00 sec)
SQL>insert into test values(null,'一二三四五');
Query OK, 1 row affected (0.00 sec)
SQL>insert into test values(null,123456);
ERROR 1406 (22001): Data too long for column 'name' at row 1
SQL>insert into test values(null,'一二三四五1');
ERROR 1406 (22001): Data too long for column 'name' at row 1
SQL>select id,name,length(name),char_length(name) from test;spa
+----+-----------------+--------------+-------------------+
| id | name | length(name) | char_length(name) |
+----+-----------------+--------------+-------------------+
| 1 | 123 | 3 | 3 |
| 2 | 12345 | 5 | 5 |
| 3 | 一二三 | 9 | 3 |
| 4 | 一二三四五 | 15 | 5 |
+----+-----------------+--------------+-------------------+
4 rows in set (0.00 sec)