Alexa Browser Client

Alexa client in your browser. Django app. Talk to Alexa from your desktop, phone, or tablet browser.

---

Demo

The demo should really be heard, so click the gif below to view it in youtube.

Run the demo

First follow these steps:

1. Configure your Amazon oauth configuration
2. Set your environment variables
3. Install:

$ git clone [email protected]:richtier/alexa-browser-client.git
$ cd alexa-browser-client
$ virtualenv .venv -p python3.6 && source .venv/bin/activate && make test_requirements

4. Compile snowboy
5. Install and run redis
6. $ make demo
7. Go to http://localhost:8000 for basic demo, or http://localhost:8000/mixer/ to play with the response audio

Installation

pip install alexa_browser_client

Make sure your settings INSTALLED_APPS contains at least these entries:

INSTALLED_APPS = [
    'django.contrib.staticfiles',
    'channels',
    'alexa_browser_client',
]

Dependencies

Snowboy detects when the wakeword "Alexa" is uttered.

You must compile Snowboy manually. Copy the compiled snowboy folder to the top level of you project. By default, the folder structure should be:

.
├── ...
├── snowboy
|   ├── snowboy-detect-swig.cc
|   ├── snowboydetect.py
|   └── resources
|       ├── alexa.umdl
|       └── common.res
└── ...

If the default folder structure does not suit your needs you can customize the wakeword detector.

You must have redis running too, as Django Channels relies on redis.

Routing and urls

Add url(r'^', include('alexa_browser_client.config.urls')), to urls.py url_patterns.

Add include('alexa_browser_client.config.routing.channel_routing') to your routing.py channel_routing.

Authentication

This app uses Alexa Voice Service. To use AVS you must first have a developer account. Then register your product here. Choose "Application" under "Is your product an app or a device"?

Ensure you update your settings.py:

Setting	Notes
ALEXA_BROWSER_CLIENT_AVS_CLIENT_ID	Retrieve by clicking on the your product listed here
ALEXA_BROWSER_CLIENT_AVS_CLIENT_SECRET	Retrieve by clicking on the your product listed here
ALEXA_BROWSER_CLIENT_AVS_DEVICE_TYPE_ID	Retrieve by reading "Product ID" here

Refresh token

You will need to login to Amazon via a web browser to get your refresh token.

To enable this first go here and click on your product to set some security settings under Security Profile and, assuming you're running on localhost:8000, set the following:

setting	value
Allowed Origins	https://localhost:8000/refreshtoken/
Allowed Return URLs	https://localhost:8000/refreshtoken/callback/

Usage

Once you have all the settings configured:

• Run django: ./manage.py runserver
• Go to http://localhost:8000 and start talking to Alexa.

Customization

Wakeword

The default wakeword is "Alexa". You can change this by customizing the lifecycle's audio_detector_class:

# my_project/consumers.py

import alexa_browser_client
import command_lifecycle


class CustomAudioDetector(command_lifecycle.wakeword.SnowboyWakewordDetector):
    wakeword_library_import_path = 'dotted.import.path.to.wakeword.Detector'
    resource_file = b'path/to/resource_file.res'
    decoder_model = b'path/to/model_file.umdl'


class CustomAudioLifecycle(alexa_browser_client.AudioLifecycle):
    audio_detector_class = CustomAudioDetector


class CustomAlexaConsumer(alexa_browser_client.AlexaConsumer):
    audio_lifecycle_class = CustomAudioLifecycle

Then in your routing.py:

import alexa_browser_client.consumers
from channels.routing import ProtocolTypeRouter, URLRouter
from channels.sessions import SessionMiddlewareStack

from django.conf.urls import url


application = ProtocolTypeRouter({
    'websocket': SessionMiddlewareStack(
        URLRouter([
            url(r"^ws/$", alexa_browser_client.consumers.AlexaConsumer),
        ])
    ),
})

Versioning

We use SemVer for versioning. For the versions available, see the PyPI.

Other projects

This project uses Voice Command Lifecycle and Alexa Voice Service Client.