Peter's Take-Home Test For Onica

This repository contains Peter's submission for Onica's take-home test. The deliverables as well as requirements. assumptions, tradeoffs, .. etc. are documented here as well

Deliverables

• Amazon DynamoDB Table
• AWS Lambda
• API Gateway

Chosen Technologies

• For the Infrastructure as Code technology, I chose serverless. Although this was my first time using it, I decided to learn something new while working on this take-home. Inside serverless.yml I also used raw CloudFormation when I needed to.
• Here are the serverless plugins I used
  • serverless-python-requirements: Used to automatically install the requirements inside requirements.txt to have our semi-reproducible build
  • serverless-pseudo-parameters: Used to be able to reference AWS Pseudo Parameters inside the serverless.yml
• For the application code, I chose Python and was careful to build a modular micro-service following the necessary patterns to build a loosely-coupled code-base. The architecture in this project follows the Layered Architecture pattern demonstrated in Eric Evan's Domain Driven Design (The blue book).

Installing Requirements

1. Install Python, at least version 3.7. Detailed instructions on this are found here.
2. Install virtualenv by typing python -m pip install virtualenv.
3. Create a virtualenv by typing python -m virtualenv -p python venv/ and activate by typing source venv/bin/activate.
4. Install Python requirements by typing python -m pip install -r requirements.txt
5. Install serverless. Detailed instructions on this are found here.
6. Install Node and npm from here.
7. Install serverless plugins by typing npm install.
8. Authenticate serverless with your AWS account. Many methods for accomplishing this can be found here.

Things I Would Have Changed In A Real Project

• Write tests, tests, and more tests. For the most part, it makes sense here to set using the doctest module. It's easy, simple, and gets the job done.
• Define more custom exceptions to handle different business logic cases, move things like UserId and Name to another file containing value objects because I just placed them the entities.py file for now while they are in fact, value objects.
• Add a CI/CD pipeline to automate the process of testing the code, integrating with dependencies, and faster deployment.
• I would have added pagination using limit-offset style.
• I would have added and configured a linter, style guide checker, and a static type checker and made them part of the CI/CD pipeline.
• I would have added a swagger file and based the API configuration off it; This would be useful because it would mean that API docs would never go out of date as the API gateway would be based off of them.