Experimental Python implementation of an Alexa Voice Service client.
I use this as one component of a home-grown Raspberry Pi radio, which has the following features:
- Flic support
- PiGlow support
- Podcast playing
- Streaming music playing
- Alexa Voice Service integration
- Text to speech from IVONA
The Alexa Voice Service component is provided by the python libraries in this repository.
It's certainly not ready for anyone else to just pick up and use, but I wanted to get it out there as I didn't find any other implementations.
Very very early - mostly this is experiments to try to understand the documentation.
- Streaming voice to AVS
- Supports most AVS features, including alarms, alerts, listen directives and so on
See the bottom of alexa.py - you can try running alexa.py for a
simple demo, which uses a local text-to-speech system rather than your
microphone to generate voice requests.
You'll need a tokens.json, which you can probably get with the help
of get-aws-token.py - edit and fill in the FILLMEIN parts. (Or
better, make it ask for those elements and send in a pull request!)
You'll need to configure the redirect_uri in your app in the AVS dev
site to http://localhost:3000/authresponse.
This list is more than what is needed - it's from my Ansible playbook for installing my overall Raspberry Pi radio system.
Raspbian packages:
- git
- python3-smbus
- python3-requests
- python3-gst-1.0
- python3-numpy
- python3-flask
- python3-sh
- python3-requests
- python3-gi
- python3-pip
- python3-feedparser
- python3-rpi.gpio
- gir1.2-gst-plugins-base-1.0
- gir1.2-gstreamer-1.0
- gstreamer1.0-tools
- gstreamer1.0-plugins-ugly
- gstreamer1.0-plugins-good
- gstreamer1.0-alsa
- gstreamer1.0-libav
- libasound2-dev
- i2c-tools
- python3-dateutil
- python3-tz
- python3-sqlalchemy
From PyPi (e.g., install with pip3):
- flask_restful
- colormath
- evdev
- pyalsaaudio
- pyudev
- requests_toolbelt
- hyper
- shove
- setuptools