The santosbot from andreffs18

View Code? Open in Web Editor NEW

🤖Bot that spits out quotes from Gustavo Santos

License: MIT License

Python 38.95% Jupyter Notebook 42.39% Dockerfile 1.48% JavaScript 1.10% Vue 5.92% Shell 0.72% HTML 1.85% SCSS 7.56% Procfile 0.03%

santosbot's Introduction

Slack Bot that spits out quotes from Gustavo Santos whenever a Trigger Word is written.

Installation | Usage | Deployment

_{Built with ❤︎ by André Silva}

Installation

To run this locally you just need to have docker installed.

Make sure you export your SLACK_API_TOKEN. More on that can be found on the /bot README file.

$ export SLACK_API_TOKEN=xoxb-*****

Usage

This project is divided into 3 parts:

Backend API: Python ASGI web server that exposes an web API for getting quotes.
Frontend App: VueJS app that helps to showcase bot functionality.
Slack Bot: Actual slack bot that you can run to interact with.

By running with $ docker-compose up you instantiate all 3 services, which then can be accessed on a browser (http://localhost:8080).

If you configured the SLACK_API_TOKEN then you should also be able to use the bot on your Slack team account, out of the box! 📦

Instructions on how to run each service separately can be found in each service folder.

Deployment

This project is hosted on Heroku, on just one dyno!. Yup, 3 apps, different languages, one dyno. You can check it on https://santosbot.herokuapp.com.

The structure is pretty simple:

There is an nginx app, working as reverse proxy, that redirects all requests to the frontend app;
All "/api" request are proxied to the backend server;
The actual bot is just running on a different process, and connects to my slack account using slack RTM library (yup, I have this live, working, just for the sake of it)
Every process was started by honcho (foreman port to python) to bypass the number of dynos restriction on heroku.

There is actually a blogpost explaining how I achieved that, so you might just check that out 😄.