At 15:00 -0500 2/22/05, Gustafson, Tim wrote:
>Hi there!
>
>I have a table, defined as follows:
>
>CREATE TABLE `WebSiteDomainNames` (
> `ID` int(10) unsigned NOT NULL auto_increment,
> `WebSite` int(10) unsigned NOT NULL default '0',
> `DomainName` int(10) unsigned NOT NULL default '0',
> `Alias` char(16) default NULL,
> PRIMARY KEY (`ID`),
> UNIQUE KEY `DomainName` (`DomainName`,`Alias`),
>) ENGINE=InnoDB DEFAULT CHARSET=latin1
>
>The way I read this definition, it should be impossible for someone to
>put in two rows with the same DomainName and Alias, however, right now I
>have the following rows in the table:
>
>+-----+---------+------------+-------+
>| ID | WebSite | DomainName | Alias |
>+-----+---------+------------+-------+
>| 543 | 1086 | 1334 | NULL |
>| 545 | 1086 | 1334 | NULL |
>| 509 | 1086 | 1334 | * |
>+-----+---------+------------+-------+
>
>And I can insert even more NULL rows if I want to. Shouldn't the UNIQUE
>key prevent this from happening?
Not for NULL values, no. See the description for UNIQUE indexes here:
http://dev.mysql.com/doc/mysql/en/create-table.html
--
Paul DuBois, MySQL Documentation Team
Madison, Wisconsin, USA
MySQL AB, www.mysql.com
