ctcuff / Tweet-Tracker

A website built using React to track the occurrences of a given keyword on Twitter in real time

Tweet-Tracker

Like Tweety for Android, this is a website to track the occurrences of a single or multiple keywords. How does it work? When a user logs into Twitter, a Python server takes the auth tokens and uses those to connect to a status stream using tweepy. When a tweet is received, that tweet is sent to the front-end via sockets.

How do I build this?

Note that before you get started, you'll need to have Python 3.x and npm installed.

0. Head over to Twitter's developer site to apply for a developer account and create a new app.
1. Go to Google's Firebase site to create a new project.
2. In the Firebase console, navigate to the Authentication tab and, under the Sign-in-tab, enable Twitter Sign-in. Enter the API keys you (hopefully) got from step 0 and be sure to copy the callback URL as well.
3. Navigate back to Twitter's developer console and paste the callback URL from step 2 into the app's callback URL parameter.

There are 2 files are missing from this project. config.py and config.tsx. Create a config.py that looks like this:

# Place this file in the server directory
#
# The consumer and consumer secret keys come from the Twitter app you created
CONSUMER_KEY = ''
CONSUMER_SECRET = ''
# This is any random string. I recommend using something from the uuid library
SERVER_KEY = ''

In config.tsx:

// Place this file in the src directory of this project
//
// In the Firebase console, navigate to Project Settings > General. Scroll down and
// click "Add app" to add a web app. Copy the generated config, it should look
// like this.
const firebaseConfig = {
  apiKey: '',
  authDomain: '',
  databaseURL: '',
  projectId: '',
  storageBucket: '',
  messagingSenderId: '',
  appId: ''
};

export default firebaseConfig;

How do I run this?

You don't need a PhD! Once you've set up everything correctly (Firebase, Twitter auth, configs, ect...), run npm install then npm run build-development. Once that finishes, give server/app.py a run and head to http://localhost:5000 on your machine.