xaal.lib 0.7.17

Official Python stack for xAAL protocol

xaal.lib is the official Python stack to develop home-automation devices and gateways with the xAAL protocol. For a full description of the protocol check out http://recherche.imt-atlantique.fr/xaal/

Dependencies

xaal.lib depends on :

• cbor2

• pysodium

• configobj

• coloredlogs

• decorator

• tabulate

• aioconsole

Install

Please refer to the official full documentation to install the lib in a virtualenv

Usage

The main goal of xaal.lib is to provide an API to easily develop devices & gateways. xaal.lib.Engine send / receive / parse to|from xAAL Bus.

To receive / parse / display incoming xAAL messages, you can simply try something like this:

from xaal.lib import Engine

def display(msg):
    print(msg)

eng = Engine()
eng.subscribe(display)
eng.run()

The Engine will call the display function every time it receive a xAAL message.

Let’s take a look at a simple lamp device :

from xaal.lib import Device,Engine,tools

# create and configure the lamp device, with a random address
dev = Device("lamp.basic", tools.get_random_uuid())
dev.product_id = 'Dummy Lamp'
dev.url  = 'http://www.acme.org'
dev.info = 'My fake lamp'

# add an xAAL attribute 'light'
light = dev.new_attribute('light')

# declare two device methods ON & OFF
def on():
    light.value = True

def off():
    light.value = False

dev.add_method('turn_on',on)
dev.add_method('turn_off',off)

# last step, create an engine and register the lamp
eng = Engine()
eng.add_device(dev)
eng.run()

To avoid to rewrite the same code for each device, you can use the xaal.schemas package. This package provides a set of predefined devices, you can use them as a template to create your own device.

from xaal.schemas import devices
from xaal.lib import Engine

# create and configure the lamp device
dev = devices.lamp()

# last step, create an engine and register the lamp
eng = Engine()
eng.add_device(dev)
eng.run()

FAQ

The core engine run forever so how can I use it in webserver, GUI or to develop device with IO. The whole API is absolutely not thread safe, so don’t use threads unless you exactly know what’s going on. Anyways, you have several options to fix this issue:

• You can use you own loop and periodically call eng.loop() for example, you can do something like this:

  while 1:
      do_some_stuff()
      eng.loop()

• You can use a engine timer, to perform some stuff.

  def read_io():
      pass
  
  # call the read_io function every 10 sec
  eng.add_timer(read_io,10)
  eng.run()

• Use the AsyncEngine. Python version > 3.8 provides async programming with asyncio package. AsyncEngine use the same API as Engine, but it is a asynchronous engine. You can use coroutines in device methods, timers functions and callbacks. It provides additionals features like the on_start and on_stop callbacks.

• Use an alternate coroutine lib, you can use gevent or greenlet for example. Look at apps/rest for a simple greenlet example.