Python has a signal module that is used to respond to signals generated by the operating system. Signals are a very low-level form of interprocess communication, but there are some cases where a program may wish to respond to a signal. For example, it may be useful to watch for program signals when writing developer toolkits.

This post demonstrates how to respond to an alarm signal. The example is borrowed from Programming Python: Powerful Object-Oriented Programming. I added my own comments to help explain the workings of the program.

Code

import sys, signal, time


# Function that returns the time
def now():
    return time.asctime()


# Function that handles the signal
def onSignal(signum, stackframe):
    print('Got alarm', signum, 'at', now())


while True:
    print('Setting at', now())
    
    # This tells the program to respond to the alarm signal
    # by calling the onSignal function
    signal.signal(signal.SIGALRM, onSignal)
    
    # Raise SIGALRM (Note this can be done by other processes also)
    signal.alarm(5)
    
    signal.pause()

Explanation

The code defines an onSignal function that works as a handler to operating system signals found on lines 10-11. All it does is prints text to the console. On line 19, we register onSignal as a handler for the SIGALRM os signal. Line 22 shows how to raise an os signal, which then invokes onSignal. Note that we don’t have to have our programs actually raise signals. We can also simply listen for other os signals raised by other programs (for example, the kill signal which is raised by executing killall in a unix shell).

References

Lutz, Mark. Programming Python. Beijing, OReilly, 2013.