MODIFY COLUMN
ALTER TABLE .. MODIFY COLUMN
语句用于修改已有表上的列,包括列的数据类型和属性。若要同时重命名,可改用 CHANGE COLUMN
语句。
从 v5.1.0 版本起,TiDB 开始支持 Reorg 类型变更,包括但不限于:
- 从
VARCHAR
转换为BIGINT
DECIMAL
精度修改- 从
VARCHAR(10)
到VARCHAR(5)
的长度压缩
语法图
- AlterTableStmt
- ModifyColumnSpec
- ColumnType
- ColumnOption
- ColumnName
AlterTableStmt
::= 'ALTER' 'IGNORE'? 'TABLE' TableName ModifyColumnSpec ( ',' ModifyColumnSpec )*
ModifyColumnSpec
::= 'MODIFY' ColumnKeywordOpt 'IF EXISTS' ColumnName ColumnType ColumnOption* ( 'FIRST' | 'AFTER' ColumnName )?
ColumnType
::= NumericType
| StringType
| DateAndTimeType
| 'SERIAL'
ColumnOption
::= 'NOT'? 'NULL'
| 'AUTO_INCREMENT'
| 'PRIMARY'? 'KEY' ( 'CLUSTERED' | 'NONCLUSTERED' )?
| 'UNIQUE' 'KEY'?
| 'DEFAULT' ( NowSymOptionFraction | SignedLiteral | NextValueForSequence )
| 'SERIAL' 'DEFAULT' 'VALUE'
| 'ON' 'UPDATE' NowSymOptionFraction
| 'COMMENT' stringLit
| ( 'CONSTRAINT' Identifier? )? 'CHECK' '(' Expression ')' ( 'NOT'? ( 'ENFORCED' | 'NULL' ) )?
| 'GENERATED' 'ALWAYS' 'AS' '(' Expression ')' ( 'VIRTUAL' | 'STORED' )?
| 'REFERENCES' TableName ( '(' IndexPartSpecificationList ')' )? Match? OnDeleteUpdateOpt
| 'COLLATE' CollationName
| 'COLUMN_FORMAT' ColumnFormat
| 'STORAGE' StorageMedia
| 'AUTO_RANDOM' ( '(' LengthNum ')' )?
ColumnName ::=
Identifier ( '.' Identifier ( '.' Identifier )? )?
示例
Meta-Only Change
CREATE TABLE t1 (id int not null primary key AUTO_INCREMENT, col1 INT);
Query OK, 0 rows affected (0.11 sec)
INSERT INTO t1 (col1) VALUES (1),(2),(3),(4),(5);
Query OK, 5 rows affected (0.02 sec)
Records: 5 Duplicates: 0 Warnings: 0
ALTER TABLE t1 MODIFY col1 BIGINT;
Query OK, 0 rows affected (0.09 sec)
SHOW CREATE TABLE t1\G;
*************************** 1. row ***************************
Table: t1
Create Table: CREATE TABLE `t1` (
`id` int(11) NOT NULL AUTO_INCREMENT,
`col1` bigint(20) DEFAULT NULL,
PRIMARY KEY (`id`)
) ENGINE=InnoDB DEFAULT CHARSET=utf8mb4 COLLATE=utf8mb4_bin AUTO_INCREMENT=30001
1 row in set (0.00 sec)
Reorg-Data Change
CREATE TABLE t1 (id int not null primary key AUTO_INCREMENT, col1 INT);
Query OK, 0 rows affected (0.11 sec)
INSERT INTO t1 (col1) VALUES (12345),(67890);
Query OK, 2 rows affected (0.00 sec)
Records: 2 Duplicates: 0 Warnings: 0
ALTER TABLE t1 MODIFY col1 VARCHAR(5);
Query OK, 0 rows affected (2.52 sec)
SHOW CREATE TABLE t1\G;
*************************** 1. row ***************************
Table: t1
CREATE TABLE `t1` (
`id` int(11) NOT NULL AUTO_INCREMENT,
`col1` varchar(5) DEFAULT NULL,
PRIMARY KEY (`id`) /*T![clustered_index] CLUSTERED */
) ENGINE=InnoDB DEFAULT CHARSET=utf8mb4 COLLATE=utf8mb4_bin AUTO_INCREMENT=30001
1 row in set (0.00 sec)
注意:
当所变更的类型与已经存在的数据行产生冲突时,TiDB 会进行报错处理。在上述例子中,TiDB 将进行如下报错:
alter table t1 modify column col1 varchar(4); ERROR 1406 (22001): Data Too Long, field len 4, data len 5由于和 Async Commit 功能兼容,DDL 在开始进入到 Reorg Data 前会有一定时间(约 2.5s)的等待处理:
Query OK, 0 rows affected (2.52 sec)
MySQL 兼容性
不支持修改主键列上需要 Reorg-Data 的类型,但是支持修改 Meta-Only 的类型。例如:
CREATE TABLE t (a int primary key); ALTER TABLE t MODIFY COLUMN a VARCHAR(10); ERROR 8200 (HY000): Unsupported modify column: column has primary key flagCREATE TABLE t (a int primary key); ALTER TABLE t MODIFY COLUMN a INT(10) UNSIGNED; ERROR 8200 (HY000): Unsupported modify column: column has primary key flagCREATE TABLE t (a int primary key); ALTER TABLE t MODIFY COLUMN a bigint; Query OK, 0 rows affected (0.01 sec)不支持修改生成列的类型。例如:
CREATE TABLE t (a INT, b INT as (a+1)); ALTER TABLE t MODIFY COLUMN b VARCHAR(10); ERROR 8200 (HY000): Unsupported modify column: column is generated不支持修改分区表上的列类型。例如:
CREATE TABLE t (c1 INT, c2 INT, c3 INT) partition by range columns(c1) ( partition p0 values less than (10), partition p1 values less than (maxvalue)); ALTER TABLE t MODIFY COLUMN c1 DATETIME; ERROR 8200 (HY000): Unsupported modify column: table is partition table不支持部分数据类型(例如,部分时间类型、Bit、Set、Enum、JSON 等)的变更,因为 TiDB cast 函数与 MySQL 的行为有一些兼容性问题。例如:
CREATE TABLE t (a DECIMAL(13, 7)); ALTER TABLE t MODIFY COLUMN a DATETIME; ERROR 8200 (HY000): Unsupported modify column: change from original type decimal(13,7) to datetime is currently unsupported yet