Everybody knows that errors can occur even in simple programs. But what really matters is what happens after the error occurs. How is the error handled? Can we control its? Can our program recover, or should it just die?

The aim of exception handling is to separate run-time error reporting and error-handling. Errors that cannot be resolved locally may be reported somewhere in a program while they are handled somewhere else. The errorhandling mechanism can be seen as a run-time analogue to the compile-time type checking. The code developed using it should have a better chance of running as expected and producing more reliable results.

Exception handling requires the use of keywords: try, catch and throw. Programs prepare to catch exceptions by trying statements that might generate a special condition. When a C++ program throws an exception, you can transfer or throw control to another part of the program called the exception handler that handles that type of exception. The handler is said to catch the exception. If no exception is encountered during the execution of the code in the try-block, the catch-block will be ignored. Otherwise, the flow of control will go to the catch-block and it can return the message. A catch-block is called an exception handler. It can be placed only immediately after a try-block or after another exception handler.

For example, let us write the program Calculator to calculate z = x @ y, where @ is one of +, ?, *, or /.