Skip to main content

ALTER TABLE COLUMN

Introduced or updated: v1.2.71
ENTERPRISE EDITION FEATURE
MASKING POLICY is an Enterprise Edition feature. Contact Databend Support for a license.

Modifies a table by adding, converting, renaming, changing, or removing a column.

Syntax

-- Add a column to the end of the table
ALTER TABLE [IF EXISTS] [database.]<table_name>
ADD COLUMN <column_name> <data_type> [NOT NULL | NULL] [DEFAULT <constant_value>];

-- Add a column to a specified position
ALTER TABLE [IF EXISTS] [database.]<table_name>
ADD COLUMN <column_name> <data_type> [NOT NULL | NULL] [DEFAULT <constant_value>] [FIRST | AFTER <column_name>]

-- Add a virtual computed column
ALTER TABLE [IF EXISTS] [database.]<table_name>
ADD COLUMN <column_name> <data_type> AS (<expr>) VIRTUAL;

-- Convert a stored computed column to a regular column
ALTER TABLE [IF EXISTS] [database.]<table_name>
MODIFY COLUMN <column_name> DROP STORED;

-- Rename a column
ALTER TABLE [IF EXISTS] [database.]<table_name>
RENAME COLUMN <column_name> TO <new_column_name>;

-- Change the data type of one or multiple columns
ALTER TABLE [IF EXISTS] [database.]<table_name>
MODIFY COLUMN <column_name> <new_data_type> [DEFAULT <constant_value>][, COLUMN <column_name> <new_data_type> [DEFAULT <constant_value>], ...]

-- Set / Unset masking policy for a column
ALTER TABLE [IF EXISTS] [database.]<table_name>
MODIFY COLUMN <column_name> SET MASKING POLICY <policy_name>

ALTER TABLE [IF EXISTS] [database.]<table_name>
MODIFY COLUMN <column_name> UNSET MASKING POLICY

-- Remove a column
ALTER TABLE [IF EXISTS] [database.]<table_name>
DROP COLUMN <column_name>;
note
  • Only a constant value can be accepted as a default value when adding or modifying a column. If a non-constant expression is used, an error will occur.
  • Adding a stored computed column with ALTER TABLE is not supported yet.
  • When you change the data type of a table's columns, there's a risk of conversion errors. For example, if you try to convert a column with text (String) to numbers (Float), it might cause problems.
  • When you set a masking policy for a column, make sure that the data type (refer to the parameter arg_type_to_mask in the syntax of CREATE MASKING POLICY) defined in the policy matches the column.

Examples

Example 1: Adding, Renaming, and Removing a Column

This example illustrates the creation of a table called "default.users" with columns 'username', 'email', and 'age'. It showcases the addition of columns 'id' and 'middle_name' with various constraints. The example also demonstrates the renaming and subsequent removal of the "age" column.

-- Create a table
CREATE TABLE default.users (
username VARCHAR(50) NOT NULL,
email VARCHAR(255),
age INT
);

-- Add a column to the end of the table
ALTER TABLE default.users
ADD COLUMN business_email VARCHAR(255) NOT NULL DEFAULT 'example@example.com';

DESC default.users;

Field |Type |Null|Default |Extra|
--------------+-------+----+---------------------+-----+
username |VARCHAR|NO |'' | |
email |VARCHAR|NO |'' | |
age |INT |NO |0 | |
business_email|VARCHAR|NO |'example@example.com'| |

-- Add a column to the beginning of the table
ALTER TABLE default.users
ADD COLUMN id int NOT NULL FIRST;

DESC default.users;

Field |Type |Null|Default |Extra|
--------------+-------+----+---------------------+-----+
id |INT |NO |0 | |
username |VARCHAR|NO |'' | |
email |VARCHAR|NO |'' | |
age |INT |NO |0 | |
business_email|VARCHAR|NO |'example@example.com'| |

-- Add a column after the column 'username'
ALTER TABLE default.users
ADD COLUMN middle_name VARCHAR(50) NULL AFTER username;

DESC default.users;

Field |Type |Null|Default |Extra|
--------------+-------+----+---------------------+-----+
id |INT |NO |0 | |
username |VARCHAR|NO |'' | |
middle_name |VARCHAR|YES |NULL | |
email |VARCHAR|NO |'' | |
age |INT |NO |0 | |
business_email|VARCHAR|NO |'example@example.com'| |

-- Rename a column
ALTER TABLE default.users
RENAME COLUMN age TO new_age;

DESC default.users;

Field |Type |Null|Default |Extra|
--------------+-------+----+---------------------+-----+
id |INT |NO |0 | |
username |VARCHAR|NO |'' | |
middle_name |VARCHAR|YES |NULL | |
email |VARCHAR|NO |'' | |
new_age |INT |NO |0 | |
business_email|VARCHAR|NO |'example@example.com'| |

-- Remove a column
ALTER TABLE default.users
DROP COLUMN new_age;

DESC default.users;

Field |Type |Null|Default |Extra|
--------------+-------+----+---------------------+-----+
id |INT |NO |0 | |
username |VARCHAR|NO |'' | |
middle_name |VARCHAR|YES |NULL | |
email |VARCHAR|NO |'' | |
business_email|VARCHAR|NO |'example@example.com'| |

Example 2: Adding a Computed Column

This example demonstrates creating a table for storing employee information, inserting data into the table, and adding a computed column to calculate the age of each employee based on their birth year.

-- Create a table
CREATE TABLE Employees (
ID INT,
Name VARCHAR(50),
BirthYear INT
);

-- Insert data
INSERT INTO Employees (ID, Name, BirthYear)
VALUES
(1, 'John Doe', 1990),
(2, 'Jane Smith', 1985),
(3, 'Robert Johnson', 1982);

-- Add a computed column named Age
ALTER TABLE Employees
ADD COLUMN Age INT64 AS (2023 - BirthYear) VIRTUAL;

SELECT * FROM Employees;

ID | Name | BirthYear | Age
------------------------------------
1 | John Doe | 1990 | 33
2 | Jane Smith | 1985 | 38
3 | Robert Johnson| 1982 | 41

Example 3: Converting a Computed Column

This example creates a table called "products" with columns for ID, price, quantity, and a computed column "total_price." The ALTER TABLE statement removes the computed functionality from the "total_price" column, converting it into a regular column.

CREATE TABLE IF NOT EXISTS products (
id INT,
price FLOAT64,
quantity INT,
total_price FLOAT64 AS (price * quantity) STORED
);

ALTER TABLE products
MODIFY COLUMN total_price DROP STORED;

Example 4: Changing Data Type of a Column

This example creates a table named "students_info" with columns for "id," "name," and "age," inserts some sample data, and then modifies the data type of the "age" column from INT to VARCHAR(10).

CREATE TABLE students_info (
id INT,
name VARCHAR(50),
age INT
);

INSERT INTO students_info VALUES
(1, 'John Doe', 25),
(2, 'Jane Smith', 28),
(3, 'Michael Johnson', 22);

ALTER TABLE students_info MODIFY COLUMN age VARCHAR(10) DEFAULT '0';
INSERT INTO students_info (id, name) VALUES (4, 'Eric McMond');

SELECT * FROM students_info;

id|name |age|
--+---------------+---+
4|Eric McMond |0 |
1|John Doe |25 |
2|Jane Smith |28 |
3|Michael Johnson|22 |

Example 5: Setting Masking Policy a Column

This example illustrates the process of setting up a masking policy to selectively reveal or mask sensitive data based on user roles.

-- Create a table and insert sample data
CREATE TABLE user_info (
id INT,
email STRING
);

INSERT INTO user_info (id, email) VALUES (1, 'sue@example.com');
INSERT INTO user_info (id, email) VALUES (2, 'eric@example.com');

-- Create a role
CREATE ROLE 'MANAGERS';
GRANT ALL ON *.* TO ROLE 'MANAGERS';

-- Create a user and grant the role to the user
CREATE USER manager_user IDENTIFIED BY 'databend';
GRANT ROLE 'MANAGERS' TO 'manager_user';

-- Create a masking policy
CREATE MASKING POLICY email_mask
AS
(val string)
RETURNS string ->
CASE
WHEN current_role() IN ('MANAGERS') THEN
val
ELSE
'*********'
END
COMMENT = 'hide_email';

-- Associate the masking policy with the 'email' column
ALTER TABLE user_info MODIFY COLUMN email SET MASKING POLICY email_mask;

-- Query with the Root user
SELECT * FROM user_info;

id|email |
--+---------+
2|*********|
1|*********|