Altering Tables ( Modifying Columns )

One method to alter a column definition is to use a  MODIFY  clause.  The name of the column to be changed must be specified, followed by its new definition. Assume that the ID column's data type must be changed from  INT to  BIGINT, to allow the table to accommodate larger identification numbers.  In addition, the column needs to be changed to  UNSIGNED to disallow negative values.

The following statement accomplishes this task: 


ALTER TABLE eu_countries MODIFY new_population BIGINT(12) NOT NULL;

That ALTER TABLE statement changes the table structure as follows: 


+----------------------+-------------------+-------+------+----------+--------+ 
| Field                    | Type                | Null  | Key  | Default | Extra | 
+----------------------+-------------------+-------+------+----------+--------+ 
| name                   | char(52)          | NO   |         | NULL    |          | 
| new_population | decimal(12,0) | NO   |         | NULL    |          | 
+----------------------+-------------------+-------+------+-----------+------+ 


To disallow  NULL in the column, the column definition provided for  MODIFY must include the  NOT NULL attribute, even if the column was originally defined with NOT NULL.  This is true for other column attributes as well.  If they are not specified explicitly, the new definition won't carry them over from the old definition. 


The above ALTER TABLE/MODIFY statement changes the table row contents as follows: 


+----------------------------------+----------------------+ 
| name                                  | new_population | 
+----------------------------------+----------------------+ 
| Albania                              |        5101800      | 
| Andorra                             |         117000       | 
| Austria                               |       12137700    | 
| Belgium                             |       15358500    | 
| Bulgaria                             |       12286350    | 
| Bosnia and Herzegovina |        5958000     | 
:                        :                : 
+----------------------------------+----------------------+ 


Note:  MySQL will sometimes try to fit existing data in to the new datatype by applying data type conversions.