#
Opendoor Engineering Problem is written in Python using Flask & Postgresql as a datastore. SQLAlchemy is used as an ORM and the whole application is deployed on Heroku.
Live version of this app can be found at - https://opendoor-listings-1000.herokuapp.com/listings
Try using a JSON Formatter browser extension to format the output in your browser.
##Setup Clone this project using -
git clone https://github.com/RahulShivkumar/opendoor-listings.git
Download the requirements using -
pip install -r requirements.txt
To run this application, you first need to setup the database and load it with data.
Set the location of your database in home.py(opendoor_listings/models/home.py). The current setup is for heroku in production, so you need to change this line to something like this
app.config['SQLALCHEMY_DATABASE_URI'] = postgresql://localhost/opendoor_listing
populate_script.py(opendoor_listings/models/populate_script.py) sets up the tables and loads with data from listings.csv
Run it using py populate_script.py within the models folder.
Once the data is setup, you can run py runserver.py that will start the server on local host port 5000 (http://127.0.0.1:5000/).
##Setup Using Heroku I used Heroku to deploy this application. Create a Heroku account here.
Once an account is created, you can download Heroku's Toolbelt here.
Login to your heroku account using -
heroku login
Create a new app called opendoor_listings from the first directory using -
heroku create opendoor-listings-1000
Once the app is created, go to your heroku dashboard (https://dashboard.heroku.com/) and select opendoor_listings. Create a new Postgresql add on for the application. This ensures that you have a datastore when you deploy the application to heroku (The address to which is stored as os.environ['DATABASE_URL']).
Just as in the previous setup, you first have to run populate_scripts.py to create the tables and load the data. The Procfile in the project tells Heroku to run populate_scripts.py after the app is pushed to it. You can push the app to heroku using
git push heroku
populate_scripts.py should have run and the data should be in the database. You can check if the script ran correctly by checking the heroku logs using heroku logs --tail. If the output doesn't show any exceptions, the tables have been created and the data is in there!
Now, change the Procfile using your favorite editor to -
web: python runserver.py
This tells Heroku to run runserver.py.
Commit this change and push it to heroku git push heroku.
Run the following command to allocate dynos.
heroku ps:scale web=1
Now run
heroku open
This will open the app in your browser. Add a /listings to the url to see all the house listings! Add the get parameters to see specific ones.
##Time Taken I timed myself while writing this application. It took me 3 hours and 15 minutes to build this app and deploy it. About an hour and 15 minutes went into deploying the app on Heroku since I had never used it before. I could have built it faster without proper error handling but I'm a pretty defensive programmer so I thought some good error handling and tests would be worth the time invested.
##Things Learnt Using Heroku - I really like the service and will definitely use it again for some future project.
GeoJson - Never heard of GeoJson before but it was cool plugging the results into geojson.io and seeing all the houses on a map!
##References Flask SQLAlchemy