BRomano / voxy

Voxy interview

Voxy

The application is Running on this IP check it if you want.

To run the test, just git pull && docker-compose up

Some decision on implementations

1. For package manager I choose to use pipenv, it is integrated on Docker build, and on a small CI process;
2. Also I choose to create two applications, one for frontend using vue.js and another for backend in Python/Flask;
3. For Backend I am using OpenAPI 3.0 (Swagger) to document every endpoint;
4. Algo for helping to run the project I am putting the project inside a docker-compose
   1. So I created 2 inner projects one for frontend with vue.js, tailwind and nginx for reverse proxy.
   2. For backend the container will be created based on voxy/Dockerfile, and run inside a gunicorn, it will run 4 instances, it is all configured on docker-compose yml;
5. To show a small knowledge with tests, I created some tests on tests folders using pytest
   1. And a github action CI to run these tests with linter and coverage.
6. To run the whole project just need to pull the repository and run docker-compose up, docker-compose need to be installed.
   1. Or you can access the application in this link

Applications

Front

1. Choosed to create two way to submit sentences, one by text input, and another to submit a text file (*.txt)
   1. If you just write some words on textarea and submit, it will count only words, not pontuations.
   2. If you choose to submit a file, frontend will load the first 200 characters and show on textarea, but will submit the whole file to backend;
   3. The frontend are using tailwind.css, typescript and vue.js

Backend

1. I choosed to create two endpoints
   1. One to count words on a sentence
   2. Another to count words on a text file
2. Both endpoints are documented by openApi
   1. And can be tested by API.
3. The Coverage can be seen on this link
4. For security and enviroments I have configured MAX_CONTENT_LENGTH
5. The application on backend was builded as a package so it can scale horizontally, acordly to framework's documentation
6. The application is using multiple routes, and is started as a module;
7. And for better management on multiple enviroments the application has a config.py so it can have multiple configurations for test, stagging, deployment;
   1. It can be seen when gunicorn is run, and on create_app() like "voxy:create_app('DevConfig')"