school_schedule

School schedule task

Author

Maciej Zięba <[email protected]>

Assumptions made

When working on the task, I have made the following assumptions:

• Other API endpoints (for example, for management of the objects) were not mentioned in the task, so I did not create them.
• The task does not provide any requirements or restrictions for the objects so I didn't add any but I believe that real-life application could provide for example:
  • Constraint for unique class names.
  • Constraint for a class to have only a single subject on a given day for a given hour.
  • Constraint for a teacher to have only a single subject with a given name.
• Currently filtering by class_name uses exact match but depending on the requirements, it could be case insensitive.
• day_of_week is stored in the database as an integer with the range of 0 (Monday) to 6 (Sunday) for faster sorting and filtering but when returning the data the endpoint provides human readable names.

Possible improvements

The fact that school schedule usually does not change very often could allow for following performance improvements (if they are not against requirements):

1. Create a "search model" that would store all the required schedule information instead of gathering it "on the fly" (requiring joins, subqueries, etc.). Any changes to original models would update the "search models".

2. It should be possible to cache HTTP responses (for example using Varnish). This could drastically improve performance and the fact that a change would be reflected only after a few seconds, in most cases should not matter.

Also the Docker setup and configuration could be improved.

Docker installation

Build the application with docker-compose:

sudo docker-compose build

Run required database migrations:

sudo docker-compose run web python src/manage.py migrate

Running application

Start the application with docker-compose:

sudo docker-compose up

Running application in development mode

Start the application with docker-compose using docker-compose.dev.yml configuration file:

sudo docker-compose -f docker-compose.dev.yml up

Running tests

Run the tests with docker-compose:

sudo docker-compose run web pytest src/schedule/tests

Local installation

Create a python virtualenv and activate it:

python -m venv venv
source ./venv/bin/activate

Install application and its dependencies:

pip install -r requirements.txt

You can also install the application with various development tools:

pip install -r requirements-dev.txt

Run required database migrations:

python src/manage.py migrate

Configuration

The application can be configured through src/school_schedule/settings.py file.

Running application

To run the application, use the standard Django runserver command:

python src/manage.py runserver

or use the gunicorn WSGI server:

gunicorn --reload school_schedule.wsgi

Running tests

To run the tests, use the pytest command:

pytest src/schedule/tests

Usage

When using the default port 8000, the endpoint is available at http://localhost:8000/schedule/

Example calls:

http://localhost:8000/schedule/
http://localhost:8000/schedule/for_today=true
http://localhost:8000/schedule/class_name=5A
http://localhost:8000/schedule/for_today=true&class_name=5A