Wolt Summer 2022 Software Engineering Internship (Backend)

Assignment

For the instructions refer to the official repository.

Overview

The application is implemented with a monolith-based architecture, since it is not complex enough for a microservice-based one, is completely Dockerized and PEP8 compliant.

In order to create OpenAPI documentation for the APIs, Swagger and Connexion framework were used. The application should be running on localhost:5000 (127.0.0.1:5000) and the Swagger interface, which allows to analyze and test the endpoints simply from the web-browser, can be accessed at localhost:5000/ui (127.0.0.1:5000/ui).

Some CI/CD tools (CircleCI, Codecov) were used in order to build and test the project, as can be seen from the above badges, and, from a security point of view, the application was tested with Bandit (static vulnerabilities' analysis).

Settings

Static data related to the application is stored in the configuation file settings.json.

{
    "delivery_fees": {
        "min_order_no_surcharge": 1000,
        "min_order_free_delivery": 10000,
        "distance_block_size": 500,
        "distance_fee": 100,
        "n_items_fee_free": 4,
        "item_fee": 50,
        "max_fee": 1500
    },
    "special_rush": {
        "weekday": 4,
        "start_time": 15,
        "end_time": 19,
        "multiplier": 1.1
    }
}

APIs

For what concern data validation, the marshmallow library was used in order to define schemas/models (models/delivery_order.py). I assumed that the input parameters that were integers could not be negative (trivial) and not equal to 0, and that the order datetime could not be in the past.

I would've added rate limiting and payload size limits in order to prevent possible DoS and brute force attacks. However, I thought this could be a bit overkill for this assignment and I decided to not implement them, since they were not requirements.

Instructions

You can decide to test/run the application with or without Docker, notice that if you're on Windows you MUST run the application with Docker (since time.tzset() is available only on Unix). Below you can find a short tutorial for both approaches.

Run the application (Docker version)

If you want to run the application with Docker, make sure you have Docker Desktop running (if you are on Windows) and follow these steps:

Go on the project's root
Build the project with docker-compose build
Run the application with docker-compose up

Run the application (Script version)

If you want to run the application without Docker, you have to follow these steps:

Go on the project's root
Create a virtual environment with virtualenv venv
Activate it with source venv/bin/activate or source venv/scripts/activate
Install all requirements needed with pip install -r requirements.txt
Run the script run.sh with bash run.sh

Testing the application

The application is already tested with CI/CD tools. However, if you would like to manually run the application's tests, follow the steps below.

If you have not yet created a virtual environment and installed the requirements:

Go on the project's root
Create a virtual environment with virtualenv venv
Activate it with source venv/bin/activate or source venv/scripts/activate
Install all requirements needed with pip install -r requirements.txt

Now simply execute the pytest command. The testing is set to fail if the coverage is not above 90%.

NennoMP/delivery-fee-calculator