Web application showing auto generated tweets based on a Markov Chain model of a given Twitter user.
Prerequisites: you'll need node and npm installed
If you don't have yarn installed, install it:
$ npm i -g yarn
After cloning the repo, install the dependencies. Inside sanata folder, run:
$ yarn install
And then run the front-end and back-end with one command:
$ yarn start
By default this run the front-end at http://localhost:3000/ and the back-end at http://localhost:8080/. Both are watching for changes, so there is no need to re-start it after changes.
In order to use the twitter api you need an API key and API secret.
Once you have them, create a .env file inside the server folder, like this:
TWITTER_CONSUMER_KEY=yourkey
TWITTER_CONSUMER_SECRET=yoursecret
#MONGO_URL=mongodb://localhost:27017/sanata
If you don't specify a MONGO_URL in your .env file, Sanata will run with a mocked db. This means that every request will go to the Twitter API without using any cached information from the user.
So, for most cases you don't need a DB, but if you insist, you have two options: run mongodb with docker or use your own instance.
If you have docker installed, running this command should be enough:
yarn run start:mongo
Don't forget to add (or uncomment)
MONGO_URL=mongodb://localhost:27017/sanatain your .env file.
This project follows the "fork-and-pull" git workflow:
- Fork the repo on GitHub
- Clone the project to your own machine
- Commit changes to your own branch
- Push your work back up to your fork
- Submit a Pull request so that we can review your changes
Here is a good step-by-step guide.
- Decoupling front end and back end
- Automatic deployment of multiple Docker containers to Google Container Engine using Travis
- Triggering deploys with git tags
MIT © Hexacta