Executing script from Alexa trigger

Question

I've been working with the example-minimal.py script from https://github.com/toddmedema/echo and need to alter it so that rather than printing the status changes to the terminal, it executes another script.

I'm a rank amateur but eager to learn and even more eager to get this project done.

Thanks in advance for any help you can provide!!

""" fauxmo_minimal.py - Fabricate.IO

    This is a demo python file showing what can be done with the debounce_handler.
    The handler prints True when you say "Alexa, device on" and False when you say
    "Alexa, device off".

    If you have two or more Echos, it only handles the one that hears you more clearly.
    You can have an Echo per room and not worry about your handlers triggering for
    those other rooms.

    The IP of the triggering Echo is also passed into the act() function, so you can
    do different things based on which Echo triggered the handler.
"""

import fauxmo
import logging
import time

from debounce_handler import debounce_handler

logging.basicConfig(level=logging.DEBUG)

class device_handler(debounce_handler):
    """Publishes the on/off state requested,
       and the IP address of the Echo making the request.
    """
    TRIGGERS = {"device": 52000}

    def act(self, client_address, state, name):
        print "State", state, "on ", name, "from client @", client_address
        return True

if __name__ == "__main__":
    # Startup the fauxmo server
    fauxmo.DEBUG = True
    p = fauxmo.poller()
    u = fauxmo.upnp_broadcast_responder()
    u.init_socket()
    p.add(u)

    # Register the device callback as a fauxmo handler
    d = device_handler()
    for trig, port in d.TRIGGERS.items():
        fauxmo.fauxmo(trig, u, p, None, port, d)

    # Loop and poll for incoming Echo requests
    logging.debug("Entering fauxmo polling loop")
    while True:
        try:
            # Allow time for a ctrl-c to stop the process
            p.poll(100)
            time.sleep(0.1)
        except Exception, e:
            logging.critical("Critical exception: " + str(e))
            break

What have you tried so far and what issues are you experiencing? — Ari Cooper-Davis, Mar 14 '17 at 23:25
I have no idea where to start, to be honest. So I've tried asking on the Instructables for it and searching the net for something similar that I can adapt... — Dave Zachritz, Mar 14 '17 at 23:53

score 0 · Answer 1 · edited May 23 '17 at 12:25

I'm going to try and be helpful by going through that script and explaining what each bit does. This should help you understand what it's doing, and therefore what you need to do to get it running something else:

import fauxmo

This is a library that allows whatever device is running the script to pretend to be a Belkin WeMo; a device that is triggerable by the Echo.

import logging
import time
from debounce_handler import debounce_handler

This is importing some more libraries that the script will need. Logging will be used for logging things, which is useful for debugging, time will be used to cause the script to pause so that you can quit it by typing ctrl-c, and the debounce_handler library will be used to keep multiple Echos from reacting to the same voice command (which would cause a software bounce).

logging.basicConfig(level=logging.DEBUG)

Configures a logger that will allow events to be logged to assist in debugging.

class device_handler(debounce_handler):
    """Publishes the on/off state requested,
       and the IP address of the Echo making the request.
    """

    TRIGGERS = {"device": 52000}

    def act(self, client_address, state, name):
        print "State", state, "on ", name, "from client @", client_address
        return True

We've created a class called device_handler which contains a dictionary called TRIGGERS and a function called act.

act takes a number of variables as input; self (any data structures in the class, such as our TRIGGERS dictionary), client_address, state, and name. We don't know what these are yet, but the names are quite self explanatory, so we can guess that client_address is probably going to be the IP address of the Echo, *state" that it is in, and name will be its name. This is the function that you're going to want to edit, since it is the final function triggered by the Echo. You can probably just stick whatever you function you want after the print statement. The act function returns True when called.

if __name__ == "__main__":

This will execute everything indented below it if you're running the script directly. More detail about that here if you want it.

    # Startup the fauxmo server
    fauxmo.DEBUG = True
    p = fauxmo.poller()
    u = fauxmo.upnp_broadcast_responder()
    u.init_socket()
    p.add(u)

As the comment suggests, this starts the fake WeMo server. We enable debugging, which just prints any debug messages to the command line, create a poller, p, which can process incoming messages, and create a upnp broadcast responder, u, which can handle UPnP device registration. We then tell u to initialise a socket, setting itself up on the network listening for UPnP devices, and add u to p so that we can respond when a broadcast is received.

    # Register the device callback as a fauxmo handler
    d = device_handler()
    for trig, port in d.TRIGGERS.items():
        fauxmo.fauxmo(trig, u, p, None, port, d)

As the comment says, this sets up an instance of the device handler class that we made earlier. Now we for-loop through the items in our TRIGGERS dictionary in our device handler d and calls fauxmo.fauxmo using the information it has found in the dictionary. If we look at the dictionary definition in the class definition we can see that there's only one entry, a trig device on port 52000. This essentially does the bulk of the work, making the actual fake WeMo device talk to the Echo. If we look at that fauxmo.fauxmo function we see that, when it receives a suitable trigger it calls the act function in the device_handler class we defined before.

    # Loop and poll for incoming Echo requests
    logging.debug("Entering fauxmo polling loop")
    while True:
        try:
            # Allow time for a ctrl-c to stop the process
            p.poll(100)
            time.sleep(0.1)
        except Exception, e:
            logging.critical("Critical exception: " + str(e))
            break

And here we enter the fauxmo polling loop. This indefinitely loops through the following code, checking to see if we've received a message. The code below it tries to poll for messages, to see if its received anything, then wait for a bit, then poll again. Except, if it can't do that for some reason, then the script will break and the error will be logged so you can see what went wrong.

score -1 · Answer 2 · answered Aug 10 '18 at 13:18

Just to clarify; If the Fauxmo loop is running then the script is fine, right?

I think the TO is not getting any connection between the Echo and the WeMo fake device. It can help if you install the WeMo skill first. You may require an original WeMo device initially though.

I know these are old threads but it might help someone still.

Executing script from Alexa trigger

2 Answers2