I think the database constraints are anyway necessary unless your database is set up for table level locking or with isolation level serializable as it would be possible for another txn to insert the record after your first txn's select. Since these constraints exist in the database, the corresponding exceptions need to be caught and handled, I guess.
Cheers, Krishna -----Original Message----- From: Alireza Fattahi [mailto:[EMAIL PROTECTED]] Sent: Thursday, January 09, 2003 4:41 PM To: 'Struts Users Mailing List' Subject: Error Handling Database constraints Hi, As we know there are two types of errors. 1) The user input error, for example and email without @. Well this can be handled easily and *separately from code* by commons validation frame work 2) The errors that violate the database consistency. For example in the user table the user_name filed must be unique. The thing we do is: first we add a constraint to database for user_name. Then when an insert requests first we select user_name from database to see if there is any duplicate. Well it means that we handle the constraint checking task our selves with java code. Why do we do this? Why did we just let the insert happens and work with SQLException? Because the returned error from database can not be handled or modified nicely. These errors are changed for each database. Are we doing the right way? Is there any better way available for error handling Database Constraints? Alireza. -- To unsubscribe, e-mail: <mailto:[EMAIL PROTECTED]> For additional commands, e-mail: <mailto:[EMAIL PROTECTED]> -- To unsubscribe, e-mail: <mailto:[EMAIL PROTECTED]> For additional commands, e-mail: <mailto:[EMAIL PROTECTED]>
