Catching only certain exceptions

Let’s go back to our input validation example.

What happens if you push ^C (i.e., control-c) at the prompt in IDLE? Here’s what we get:

Please enter a number: ^CTraceback (most recent call last):
    File "/home/main.py", line 2, in <module>
        s = input('Please enter a number: ')
KeyboardInterrupt

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
    File "/home/main.py", line 6, in <module>
        print("Uh oh: '" + s + "' isn't a valid number.")
NameError: name 's' is not defined

Notice that two exceptions happened. (Yikes! 😬) First, we got a KeyboardInterrupt. That’s normal: that’s what ^C does. That brought us to the exception handler, where we got another exception, this time because s isn’t defined.

What’s going on here? Why wasn’t s defined? We were in a try/ except… why did we get a second exception, and why didn’t it get handled? Here’s what happened in the code: while input was waiting for user input, pushing ^C caused a KeyboardInterrupt. That exception bubbled up out of input and into our code, causing us to run the except block. But that means input never returned anything to s, so the except block couldn’t even talk about the input… because it never happened!

In general, you don’t want to catch every exception, just those that you could reasonably respond to. A ^C is a user saying, “Help, get me out of here!”, and it’s best to let that happen. But the ValueError from a bad input to int is eminently handle-able. In order to catch only certain exceptions, you can name them after the except. Here’s a revision of the code:

Notice that when using ^C in IDLE it works fine, but giving a non-number gives the nice output.

If you care about multiple exceptions, you can write more than one except clause. Here’s an example:

When an exception happens, Python will try each except in turn, matching the exception thrown against the type named there. It’s possible to name more than one:

The program before this was probably better: it gives more detailed feedback to the user. In any case

The program before this was probably better: it gives more detailed feedback to the user. In any case, Python has many possible exceptions. It’s worth getting vaguely acquainted with them!

Finally, Python’s try/ except have more features than we’ve shown you: else blocks and finally blocks. Read the Python tutorial on Errors and Exceptions to learn more.