Fax Robot API

A Python API to send faxes. Now, with chicken!

Introduction

Fax Robot is an easy-to-use API for sending faxes through your own faxmodem(s). It's based on a Flask web API and one or more worker processes. Faxes are sent into a queue, so you can send tons of them in, and they'll be processed one by one. Or, if you have a bunch of faxmodems and phone lines, you can run multiple workers and power through them really quickly.

Sending faxes to long distance numbers is not free, so Fax Robot can optionally collect payments from API users via Stripe. Users purchase stored credit, and faxes are charged against this credit based on a fixed cost per page. Or, if you're feeling generous, you can make it free for anyone with network access.

Fax Robot was designed to be easy to set up and configure by anyone with basic Python devops skills. It's licensed under the GNU General Public License, Version 3.

Also check out faxrobot-www, which is a frontend webapp for the Fax Robot API.

Dependencies

• One or more physical faxmodems hooked up to land lines
• Python 2.7
• PostgreSQL
• Redis
• ImageMagick (particularly the convert and composite commands)
• paps
• efax (the linux command to send faxes)
• (Optional) Stripe API, only needed if you're collecting payments.
• (Optional) SparkPost API, for sending account-related emails to API users.
• (Optional) Amazon S3 storage. Only needed if you have multiple servers for your workers (basically stores your temp files in the cloud)

Installation

There are two pieces to Fax Robot: a Python Flask API and the worker process that actually runs through the queue and sends the faxes. How you run these in production will vary depending on your hosting infrastructure (production deployment is left as an exercise to the reader), but local testing is easy.

Set up your Virtualenv and packages

You should use [Virtualenv] to isolate Fax Robot's required package dependencies away from your system Python installation. You don't have to, but come on. Assuming you have virtualenv installed, do the following commands from within your Fax Robot project directory to install Fax Robot's required packages:

virtualenv venv
source venv/bin/activate
pip install -r requirements.txt

Set up your environment variables

Fax Robot's configuration is stored in system environment variables. You should create a file called .env in your Fax Robot project directory (or wherever you like). The file should look something like this:

export DEBUG=true
export SECRET_KEY=asdf1234
export DATABASE_URI=postgres://username:password@localhost/faxrobot
export REDIS_URI=redis://127.0.0.1:6379?db=1
export REQUIRE_PAYMENTS=on
export DEFAULT_COST_PER_PAGE=0.06
export AWS_STORAGE=off
export AWS_ACCESS_KEY=AKSPLGHLOL
export AWS_SECRET_KEY='trolol'
export AWS_S3_BUCKET='s3.faxrobot.io'
export FAXROBOT_DIR=/htdocs/faxrobot
export GLOBAL_SALT=omgtrolol
export STRIPE_SECRET_KEY=sk_test_12345678
export SERVEOFFICE2PDF_URL=http://convert.dev/convert
export PROJECT_NAME='Fax Robot'
export SPARKPOST_API_KEY=asdf
export EMAIL_FROM='[email protected]'
export EMAIL_FROM_NAME='Fax Robot'
export EMAIL_SUPPORT='[email protected]'
export EMAIL_FEEDBACK='[email protected]'
export FAXROBOT_URL='https://faxrobot.io'

Here are the specific environment variables, and what they do:

• DEBUG: Whether to run the API in debug mode. Useful for local testing.

• SECRET_KEY: Secret key used by Flask for sessions. It's probably not used for anything, but good practice to have it just in case.

• DATABASE_URI (required): A PostgreSQL connection string to use for database access. Database tables are created automatically with SQLAlchemy, so you don't have to set the schema up manually.

• REDIS_URI (required): A Redis connection string. Required for the rq worker to operate.

• DEFAULT_COST_PER_PAGE: The default cost per page in USD to charge for outgoing faxes. The default is 0.06.

• REQUIRE_PAYMENTS: If set to "on" then API users will need to purchase credit via Stripe API endpoints in order to send faxes. This also means you will need a valid STRIPE_SECRET_KEY in the configuration.

• AWS_STORAGE: If set to "on" then temporary data for jobs will be uploaded to Amazon S3. This is only useful if you have more than one physical server running the Fax Robot worker process. Turning this on also means you'll have to set up the other AWS_ prefixed configuration variables.

• AWS_ACCESS_KEY: Your Amazon AWS Access Key ID.

• AWS_SECRET_KEY: Your Amazon AWS Secret Key.

• AWS_S3_BUCKET: The Amazon S3 bucket to use for data storage.

• FAXROBOT_DIR (required): The directory you have the Fax Robot project stored in.

• GLOBAL_SALT (required): A global salt value used for hashing passwords and other secret values. This should be an arbitrary string (longer is better)

• STRIPE_SECRET_KEY: Only required if REQUIRE_PAYMENTS is turned on. This is your Stripe secret key used for processing payments.

• SERVEOFFICE2PDF_URL: ServeOffice2PDF is a separate web application that converts Microsoft Office file formats (like .doc and .docx) to PDF. This is optional, and only used to support faxing these formats via the Fax Robot API. By default Microsoft Office formats are unsupported.

• PROJECT_NAME: The name you're giving to your instance of the project.

• SPARKPOST_API_KEY: Used for sending transactional emails to Fax Robot API accounts. If not present, then emails are disabled.

• EMAIL_FROM: The from email address for transactional emails.

• EMAIL_FROM_NAME: The from email name for transactional emails.

• EMAIL_SUPPORT: A support email address to include in relevent transactional emails.

• EMAIL_FEEDBACK: A feedback email address to include in relevent transactional emails.

• FAXROBOT_URL: The URL to a front-end client webapp for your Fax Robot instance.

Once your environment variables are configured, load them into your shell using the source command:

source .env

Running the Fax Robot API

The API listens for incoming fax jobs (and other CRUD type API requests) and pushes faxes onto the worker's queue. To start it, you'll simply run python api.py after following the steps above. To recap:

source venv/bin/activate
source .env
python api.py

When running in Debug mode, the API will listen on localhost:9001.

Running the Worker

The worker listens for fax jobs on the rq queue and processes them. After loading your environment up, you'll run python worker.py with your faxmodem device specified as an argument. For example:

source venv/bin/activate
source .env
python worker.py --device /dev/ttyUSB0 --callerid '+1 202 555 1212'

If you have more than one faxmodem hooked up, you can run multiple instances of worker.py for maximum pwnage.

worker.py command line arguments

• --help: show help message

• --device: the modem device to send faxes through, eg /dev/ttyUSB0. The device for your faxmodem must be accessible by the user account you're running the worker.py process under. (hint: add your user account to the dialout group)

• --callerid: the caller ID string to display on outgoing faxes, in international format. Some fax machines may not accept characters other than numbers, space, and '+'.

• --listen: which queue(s) the worker should listen to. Can be high, default, or low, or a comma-separated list thereof, where the first items in the list are prioritized more highly. See the section below on queue priorities.

Worker queue priorities

Fax Robot uses rq to process jobs asynchronously from incoming API requests. These jobs are tagged by priority: high, default, and low. You should always have at least one worker.py instance dedicated to high priority queue items.

• high: The job which does the initial cost processing and image conversion for new faxes is high priority, because users will often poll the fax status endpoint to see cost information before sending faxes.

• default: Sending faxes is a default priority task. Each faxmodem should be given its own worker process listening on the default queue.

• low: You can optionally flag certain accounts as low priority by editing them in the database. This allows you to prioritize outgoing faxes from other accounts before sending any from the low priority accounts (if both are pending at the same time). Be sure to set your fax worker instance to favor default jobs above low jobs like so:

  python worker.py --listen 'default,low'
  

API Documentation

Read the full API documentation at https://faxrobot.io/api