In this chapter we examined the MySQL error handlers that allow you to catch error conditions and take appropriate corrective actions. Without error handlers, your stored programs will abort whenever they encounter SQL errors, returning control to the calling program. While this might be acceptable for some simple stored programs, it is more likely that you will want to trap and handle errors within the stored program environment, especially if you plan to call one stored program from another. In addition, you need to declare handlers for cursor loops so that an error is not thrown when the last row is retrieved from the cursor.
Handlers can be constructed to catch all errors, although this is currently not best practice in MySQL, since you do not have access to an error code variable that would allow you to differentiate between possible error conditions or to report an appropriate diagnostic to the calling program. Instead, you should declare individual handlers for error conditions that can reasonably be anticipated. When an unexpected error occurs, it is best to let the stored program abort so that the calling program has access to the error codes and messages.
Handlers can be constructed that catch either ANSI-standard SQLSTATE codes or MySQL-specific error codes. Using the SQLSTATE codes leads to more portable code, but because specific SQLSTATE codes are not available for all MySQL error conditions, you should feel free to construct handlers against MySQL-specific error conditions.
To improve the readability of your code, you will normally want to declare named conditions against the error codes you are handling, so that the intention of your handlers is clear. It is far easier to understand a handler that traps DUPLICATE_KEY_VALUE than one that checks for MySQL error code 1062.
At the time of writing, some critical SQL:2003 error-handling functionality has yet to be implemented in MySQL, most notably the ability to directly access the SQLSTATE or SQLSTATE variables, as well as the ability to raise an error condition using the SIGNAL statement. In the absence of a SQLSTATE or SQLCODE variable, it is good practice for you to define handlers against all error conditions that can reasonably be anticipated that populate a SQLCODE-like variable that you can use within your program code to detect errors and take appropriate action. We expect MySQL to add these "missing" features in version 5.2you should check to see if they have been implemented in the time since this book was written (see the book's web site for details). Note also that it is currently possible to provide a workaround (though a somewhat awkward one) for the missing SIGNAL statement if you find that it is absolutely necessary in your programs.
Part I: Stored Programming Fundamentals
Introduction to MySQL Stored Programs
MySQL Stored Programming Tutorial
Language Fundamentals
Blocks, Conditional Statements, and Iterative Programming
Using SQL in Stored Programming
Error Handling
Part II: Stored Program Construction
Creating and Maintaining Stored Programs
Transaction Management
MySQL Built-in Functions
Stored Functions
Triggers
Part III: Using MySQL Stored Programs in Applications
Using MySQL Stored Programs in Applications
Using MySQL Stored Programs with PHP
Using MySQL Stored Programs with Java
Using MySQL Stored Programs with Perl
Using MySQL Stored Programs with Python
Using MySQL Stored Programs with .NET
Part IV: Optimizing Stored Programs
Stored Program Security
Tuning Stored Programs and Their SQL
Basic SQL Tuning
Advanced SQL Tuning
Optimizing Stored Program Code
Best Practices in MySQL Stored Program Development