markdimi / job-board

Take home project : job board

For this task the following technologies where used

• Python and Django for the backend.
• PostgreSQL for the required database.
• Bootstrap and JavaScript for the frontend.
• Docker & docker-compose for setting up the development environment (optional).
  • You can create the backend enviornment using the Dockerfile and the docker-compose.yml file.

Setting up the environment

For the backend you'll need the following:

• Python (3.5>) and the following packages:
  • Django, psycopg2-binary, djangorestframework, requests and django-cors-headers
• PostgreSQL with a database, user and its password, all set to 'postgres' (as it’s defined in the settings.py file in the DATABASES section).
  • Otherwise if your database setup is different you’ll need to change the settings accordingly.
• Then you can clone this project. The django project's name is jobs_coard and core is the name of the app.
• Now you can run the migrations for the Searches model to take place.
  • python manage.py makemigrations core
  • python manage.py migrate
• (Optional) You can also create a superuser to access Django's admin where you could observe the Searches model.
  • python manage.py createsuperuser
• Lastly you can run django server in port '9898'
  • python manage.py runserver 0.0.0.0:9898

For the frontend

• The frontend is an html file inside the Simple frontend folder.
• You will need an http server for serving the file in port '8000'
  • I used Python's Simple HTTP Server: python -m http.server

Execution

Now you can navigate to 0.0.0.0:8000 and you will see the form where you can choose a description and a location. With every search, the Django API is fetching results from Github's Jobs API and saves the client's data (description, location, ip and time) in the Searches model (and in a db table with the same name).

Final notes

Using Django for a project like this could be an overkill. However I used it because of my familiarity with it and its powerfull tools. Especially the Models API where you can create models that represent your data. Alternatively I could create a table inside the database and write all the logic in PHP or Node.

Also I could fetch Github's results directly from the frontend. Instead now, I get the form data in the backend and fetch the results from there.