Simple python event signaling

python, event, signal, signals, signaling
pip install smokesignal==0.7.0


smokesignal - simple python signaling

Build Status

smokesignal is a simple python library for sending and receiving signals. It draws some inspiration from the django signal framework but is meant as a general purpose variant.

Requirements & Compatibility

smokesignal requires no dependencies outside of the python standard library. It has been tested on and is compatible with python versions 2.6, 2.7, 3.2, and 3.3.

How To Use

Most uses of smokesignal involve registering a single callable to respond to a signal using on and sending signals using emit.

Registering Callbacks

Most callback registration happens using on, which can be used either as a decorator or a direct function call. This method also accepts an optional argument max_calls which indicates the maximum number of times a callback should respond to an emitted signal:

import smokesignal

def my_callback():

@smokesignal.on('foo', max_calls=2)
def my_callback():

smokesignal.on('foo', my_callback, max_calls=5)

If you intend a callback to respond to a signal at most one time, rather than indicating max_calls=1, you can use once as a convenience:

import smokesignal

def my_callback():

Sending Signals

Signals are sent to all registered callbacks using emit. This method optionally accepts argument and keyword argument lists that are passed directly to each callback:

import smokesignal

# Each callback responding to 'foo' is called

# Each callback responding to 'foo' is called with arguments
smokesignal.emit('foo', 1, 2, 3, four=4)

You can also send signals with the included context manager emitting. By default, this context manager accepts one argument, which is a signal to send once the context manager exits. However, you can supply keyword arguments for enter and exit that will be sent at those points of the context manager:

import smokesignal

# 'foo' emitted on exit
with smokesignal.emitting('foo'):

# 'foo' emitted on enter, 'bar' emitted on exit
with smokesignal.emitting(enter='foo', exit='bar'):

Disconnecting Callbacks

If you no longer wish for a callback to respond to any signals, you can use either disconnect_from, if you intend on removing specific signals, or disconnect if you intend to remove all of them:

import smokesignal

# my_callback will no longer respond to signals

# my_callback will no longer respond to 'foo', but may repond to others
smokesignal.disconnect_from(my_callback, 'foo')

Twisted Support

When Twisted is installed (14.0 or greater recommended), emit will return a deferred that resolves to a list of results.

import smokesignal

def f1(): return 'f1'

def f2(): return 'f2'

d = smokesignal.emit('tx')

def results(rr):
    print rr

# => ['f1', 'f2']

You can also pass in an explicit errback argument. This will get called for each Failure caused by one of your receivers.

import smokesignal

def err(): return {}['ono!']

def handleError(f):
    return f.type.__name__

# pass in the errback argument to handle failures
d = smokesignal.emit('tx', errback=handleError)

def results(rr):
    print rr

# => ['KeyError']

Errback handling works the way you would expect in Twisted: If your errback handler returns the failure (or raises an exception), the operation will fail, but if it returns anything else (including None), it will be treated as a successful result.

Other Batteries Included

You can clear large swaths of callbacks using either clear or clear_all. If you call clear without any arguments, it effectively works like clear_all:

# Remove all callbacks responding to a specific signal

# Remove all callbacks responding to all signals

Sometimes you may want to get a list of signals a callback responds to or quickly check if a callback will respond to a certain signal. The signals and responds_to are available for this purpose. Note that registering a callback to respond to a signal will also create callable attributes of the callback for easier interaction with these methods:

# Get a tuple of all signals a callback responds to

# Check if a callback responds to a signal
smokesignal.responds_to(my_callback, 'foo')

# Or as attributes of the callback


What would be great is if you could decorate instance methods using on or once. However, that doesn't work because there is no knowledge of the class instance at the time the callback is registered to respond to signals:

import smokesignal

class Foo(object):

    def callback(self):

There is a workaround if you would like instance methods to respond to callbacks:

import smokesignal

class Foo(object):
    def __init__(self, *args, **kwargs):
        smokesignal.on('bar', self.callback)
        super(Foo, self).__init__(*args, **kwargs)

    def callback(self):

The above will register the callback without any argument requirements but will also ensure that the intended callback method is called correctly.



  • Callbacks now have callable attributes responds_to and signals
  • Calling clear with no arguments will clear all callbacks for all signals


  • Added emitting context manager
  • Updated internals no longer require decorator magic for enforcing maximum call counts


  • Initial Version

Contribution and License

Developed by Shaun Duncan and is licensed under the terms of a MIT license.