Trading API

This project is a Ruby on Rails API for UN/LOCODE search.

The project download locodes zip file from UNECE, then unzip csv files and import all content to a PostgreSQL database.

The API endpoints could be checked with an Swagger documentation at the /api-docs endpoint.

Getting Started

There are two ways to run this project: with Docker or local.

Running Local

Ruby version: 2.6.3
Rails version: 6.0.2

If you run this project local, you need to change config/database.yml file, setting your own database configuration.

Example:

default: &default
  adapter: postgresql
  encoding: unicode
  host: localhost
  user: postgres
  password: your_db_pass
  port: 5432

After changes on database.yml, you may run the script command sh start.sh on the root project directory. This script file has all setup needed to install dependencies, create database ,run migrations, tests and also download and import locodes to database.

Running with docker

This project has a Dockerfile and docker-compose file to build all environment with two simple commands:

$ docker-compose build

$ docker-compose up

API documentation: http://localhost:3000/api-docs

Running with local kubernetes cluster (docker-desktop)

Is possible to deploy this project into a local kubernetes cluster, using the following command in the project root directory:

docker stack deploy --compose-file docker-compose.yml trading-api

After deployment, whe project will take some time to download whe zip file, unzip and then import all data from csv files.

If you want to follow the import proccess, is possible using the logs action on the menu below:

.

.

API Documentation

Is possible to test the API endpoints using the button "Try out" inside each endpoint on http://localhost:3000/api-docs.

Running the tests manually

If you would like to run tests manually, in a local installation, just run in your terminal (root project directory):

$ bundle exec rails rspec

A test coverage report will appear on the end.

Coverage report generated for RSpec to /app/coverage. 105 / 105 LOC (100.0%) covered.

Final considerations

It is not a good practice to leave the master.key file in the project directory, but I did because I leave this repository private.

The Postgres instance inside a docker container is not a good practice for production environments, just for testing.

There are a lot of possible improvements to this project, for example, using Redis for caching and writing more test units. But the main point was to show some skills that I have working in a backend project.