Simple NodeJS REST API for movies, built using express.js framework, which has two endpoints:
-
Create a new movie
POST /movies- Input:
- JSON Object
- properties: title: type: string description: type: string genre: type: string director: type: string year: type: string
- JSON Object
- Output:
- 200 OK or error code
- Input:
-
Getting all movies
GET /movies:- Input:
- filter object (optional)
- page number
- limit per page (N)
- Output:
- list of movies (no more than N)
- Input:
This app allows filter by multiple attributes, an example of it is
GET /movies?filter[year]=2002&filter[genre]=Action
In order to deploy and run this API whitout using docker, MongoDB and Redis should be installed in
the target host. Then just npm install and npm start. There are 2 environment variables required:
MONGO_HOST and REDIS_HOST, they can be stored in a .env file.
This projects uses Docker Compose for defining and running
multiple docker containers. There are 3 containers used: A container for MongoDB, another for Redis
and finally one for the app. For building and deploy the project just docker-compose build and
then MONGO_DB_NAME=movies-api REDIS_DB_INDEX=0 docker-compose up
This app uses a Redis instance in order of creating a cache for recent queries. This instance has a limit of available memory (that can be changed through env variables) and deletes less recently used keys when that limit is reached. This cache is emptied after a new registry is created, choosing consistency over fast results.
Tests are done using Mocha. A mix of unit tests and E2E tests, using
scenarios when creating, querying and making sure that the queries are correct. Test can be executed
just by running the command npm run test or
MONGO_DB_NAME=movies-api-test REDIS_DB_INDEX=1 docker-compose run web yarn test
Even though this is a small app with only 2 endpoints, it's always good to have enpoints well
documented. In order to do that this project contains an endpoint with the help of Swagger that
expose the usage of the API in /docs