EventEmitter

The EventEmitter class can be used directly or as a mix-in to provide the ability to publish and subscribe to events.

import logging
from eventemitter import EventEmitter

log = logging.getLogger(__name__)
emitter = EventEmitter()

def log_event(event, *args, **kwargs):
    log.debug('%s %s %s', event, args, kwargs)

emitter.on('an-event', log_event)
emitter.emit('an-event', 1, 2, keyword=3)
await asyncio.sleep(0)  # 'an-even (1, 2), {keyword: 3}' gets logged.

Listener functions can be defined using def or async def. All listeners are executed in a deferred way. The coro that calls emit must yield for the event to propagate.

EventIterable

If the callback-style model of listening for events is undesirable, an async iterable is provided to offer a second model for handling events.

import logging
from eventemitter import EventEmitter
from eventemitter import EventIterable

log = logging.getLogger(__name__)
emitter = EventEmitter()
iterable = EventIterable(emitter, 'an-event')

async for args, kwargs in iterable:

    log.debug('%s %s %s', event, args, kwargs)

The EventIterable implements the async iterable interface and can be used in conjunction with any of the tools in aitertools.

Style

As long as the code passes the PEP8 and PyFlakes gates then the style is acceptable.

Docs

The PEP257 gate will check that all public methods have docstrings. If you’re adding something new, like a helper function, try out the napoleon style of docstrings.

Tests

Make sure the patch passes all the tests. If you’re adding a new feature don’t forget to throw in a test or two. If you’re fixing a bug then definitely add at least one test to prevent regressions.