An #openfooddata project started at Open Food Data Hackdays Zug 2018
Based on flask-vuejs-template. See the upstream project for updates and further details.
Before getting started, you should have the following installed and running:
- Yarn - instructions
- Vue Cli 3 - instructions
- Python 3
- Pipenv (optional)
- Heroku Cli (if deploying to Heroku)
-
Clone this repository:
$ git clone https://github.com/AndizzleB/Eatinder.git
-
Setup virtual environment, install dependencies, and activate it:
$ pipenv install --dev $ pipenv shell
-
Install JS dependencies
$ yarn install
Run Flask Api development server:
$ python run.py
From another tab in the same directory, start the webpack dev server:
$ yarn serve
The Vuejs application will be served from localhost:8080
and the Flask Api
and static files will be served from localhost:5000
.
The dual dev-server setup allows you to take advantage of webpack's development server with hot module replacement.
Proxy config in vue.config.js
is used to route the requests
back to Flask's Api on port 5000.
If you would rather run a single dev server, you can run Flask's
development server only on :5000
, but you have to build build the Vue app first
and the page will not reload on changes.
$ yarn build
$ python run.py
This template is configured to work with Heroku + Gunicorn and it's pre-configured to have Heroku build the application before releasing it.
Heroku's nodejs buidlpack will handle install for all the dependencies from the packages.json
file.
It will then trigger the postinstall
command which calls yarn build
.
This will create the bundled dist
folder which will be served by whitenoise.
The python buildpack will detect the Pipfile
and install all the python dependencies.
The Procfile
will run Django migrations and then launch Django's app using gunicorn, as recommended by heroku.