When you run test using a mongodb database for testing, you will encountered multiple problems such as :

• Rapidity of execution: The time it takes to run the test
• Reliability and Avoiding Dependencies between tests: Setting up the database with a fixed set of data for each test case

Few weeks ago, I have discovered mongodb-memory-server which give the ability to run all the mongodb call in memory. With this package setup, we can start with a setup database using mongo-seeding and clear the db at each test in a fraction of second.

This project show how to setup the seeding with a mongodb running in memory. For also pushing it further, I find a way to run the server only once for all the tests.

I explain with all the details how I build the project and my way of working.

• Settings: How to setup mongodb
• Technologies
• ERD
• Documentation
• Organization
• Development
• Seeding
• Testing
• Admin
• Monitoring
• Security
• Running
• Deployment

For setting up the memory package for the test, you need to follow those step :

• Make sure when you run the server, the mongodb database runs depending of the environment. In test environment, the database should not be started since we will be using the memory database.
• Create a mongodb memory package such as inside the file: test/cases/libs/mongodb.js
• Then put inside the beforeEach and afterEach, the connection and deconnection to the memory database such as in the file: test/runner.js
• Now, if you run the test, all test should work independently.

The server is using the library Express but it can be switch easily in the server.js file for Fastify or Restify. The server is linked to the library Apollo-server for managing the data with GraphQL. The database is handle by MongoDB and can also be easily switch in database.js.

The models are found in the folder models. Their schema are described with mongoose and also typed for using GraphQL.

For the test, I am using Ava, mongo-seeding for setting up the database and mongodb-memory-server for running the database in memory.

Graphql-voyager has been added as a middleware so you can visualize the ERD on the following endpoint : /erd

http://localhost:5000/erd

The documentation of the server can be access at this endpoint. It gives information about the server, the tools and how to use the API. Open a browser and go to the following URL :

http://localhost:5000/documentation

The jsdoc can be generated locally with the following command :

npm run build:docs

The playground of the apollo server is activated, so the documentation and the json format can be access at the endpoint of the api. Open a browser and browse to the following URL :

http://localhost:5000/api/graphql

Since it's a graphql, anyone can generate the documentation by introspection. I suggest to use this tool 2fd/graphdoc with the following commands :

npm install -g @2fd/graphdoc
graphdoc -e http://localhost:5000/api/graphql -o ./doc/schema

A html doc will be then found inside the directory /doc/schema.

• @admin-bro/express: A middleware for connecting express with admin-bro. I use it for creating the route of the admin bro.
• @admin-bro/mongoose: A middleware for connecting mongoose with admin bro. I use it for connecting admin bro to the mongoose schema.
• admin-bro: AdminBro is an admin interface for managing/editing the information in the database. It removes the need of building an admin interface. I use it because I cannot rely on someone else for building an admin interface if I work with a frontend.
• apollo-server-express: Apollo serves as an abstraction layer that decouples services and apps so that each can be developed independently of the other, in any language and on any platform. I use it for managing a graphQl.
• basic-auth: Basic Auth is used for protecting url and ressource by a htaccess authentication. I use it for protecting the admin bro route.
• bcrypt: A library for hashing information. I use it for hashing the password.
• dotenv: Dotenv is a zero-dependency module that loads environment variables from a .env file into process.env. I use it for creating an environment for the prod, stage, dev and local development.
• email-validator: Checking an email is a simple task but it can be tricky to check all the possibilities given by a particular RFC. I rely on this small package for checking the validity of an email.
• express: xpress is a minimal and flexible Node.js web application framework that provides a robust set of features for web and mobile applications. I use it for managing the routes and the API.
• express-session: A middleware for managing the session through express. I use it for creating the account of the app.
• express-status-monitor: Simple, self-hosted module based on Socket.io and Chart.js to report realtime server metrics for Express-based node servers. I use this for checking the status of the server without connecting to aws.
• graphQl: GraphQL is an open-source data query and manipulation language for APIs, and a runtime for fulfilling queries with existing data. I use it because it give more flexibility for the frontend and it optimizes the query way better than REST.
• graphql-tools: This package provides a few useful ways to create a GraphQL schema. I use this for creating an introspection of the API and creating thing such as documentation and erd with the schema generated.
• graphql-voyager: With graphql-voyager you can visually explore your GraphQL API as an interactive graph. This is a great tool when designing or discussing your data model. I use it for letting people see the erd and the documentation behind the API without doing any effort.
• helmet: Helmet helps you secure your Express apps by setting various HTTP headers. I use it for hidding some header and protecting the app against well know weakness of Express.
• isomorphic-fetch: The Fetch API is currently not implemented consistently across browsers. This module will enable you to use fetch in your Node code in a cross-browser compliant fashion. I use it for testing the unit test, I call every endpoint with a fetch as any browser will do.
• jsonwebtoken: JSON Web Token (JWT) is a compact, URL-safe means of representing claims to be transferred between two parties. I use it for creating and managing the Bearer Token.
• mailgun-js: Node.js module for interacting with the Mailgun API. I use it for sending mail with Mailgun.
• marked: Low-level compiler for parsing markdown without caching or blocking for long periods of time. I use it for reading the content of the Readme and passing it to a webpage. That way the README is readable in a browser through an endpoint.
• module-alias: Create aliases of directories and register custom module paths in NodeJS. I use it for avoiding transversal path inside the call, it makes the development way more simpler and cleaner.
• mongo-uri-builder: A zero dependency Node.js module to easily create MongoDB connection strings using configuration objects. I use it for passing the uri to mongodb and making the connection with the express server.
• mongoose: Mongoose is a MongoDB object modeling tool designed to work in an asynchronous environment. I use it like an ORM and for designing the schema of the app easily.
• mustache: Mustache is a logic-less template syntax. It can be used for HTML, config files, source code - anything. It works by expanding tags in a template using values provided in a hash or object. I use it for parsing tags inside the email template and replacing those tags by a value from the app.
• node-cron: The node-cron module is tiny task scheduler in pure JavaScript for node.js based on GNU crontab. I use it for managing the calls that need to be execute following a schedule.
• react: React is a JavaScript library for creating user interfaces. I use it for creating the user interface for the ERD and documentation.
• react-dom: This package serves as the entry point to the DOM and server renderers for React. I use it for managing the interaction in the dom for the voyager.
• winston: A logger for just about everything. I use it for saving into a log file any interaction in the server.
• ava: AVA is a test runner for Node.js with a concise API, detailed error output, embrace of new language features and process isolation that lets you develop with confidence. I use it for managing effiently the test of the app.
• coveralls: Coveralls.io support for Node.js. I use it for linking the app to coveralls website and getting the percentage of coverage of the test.
• eslint: ESLint is a tool for identifying and reporting on patterns found in ECMAScript/JavaScript code. In many ways, it is similar to JSLint and JSHint with a few exceptions. I use it for fixing the style of the app.
• faker: Generate massive amounts of fake data in the browser and node.js I use it inside the test for generating random data, it make the test more powerful.
• heml: HEML is an open source markup language for building responsive email. I use it for building the email template and making responsive email.
• husky: Husky is a modern native Git hooks. I use it for restricting the commit to the interesting one and for forcing the test to be run before merging anything.
• jsdoc: An API documentation generator for JavaScript. I use it for generating the documentation using the comments in the code.
• mongo-seeding: The ultimate solution for populating your MongoDB database. I use it for populating the test database and also for populating the server at first installation.
• npx: Executes either from a local node_modules/.bin, or from a central cache, installing any packages needed in order for to run. I use it for running package from my local node module directory such as esLint.
• nyc: Istanbul's state of the art command line interface. I use it for creating the report for the coverall and making it available in the browser.

For having a database with some data from the start, you can seed it with a single command. It will fill up all the db with dummy data using mongo-seed.

npm run seed

For automatic tests, I use Ava. All the tests can be run with a single command. The command will also provide details on the coverage.

npm run test

For connecting to the API during the development, I use localtunnel for exposing the server through a domain format. The server works on the port 5000, so I use the following commands :

npm install -g localtunnel
lt --port 5000 --subdomain couple-api

The server have a version of AdminBro for managing some of the content. Open a browser and go to the following URL :

http://localhost:5000/admin

WARNING : At February 06 2020, the AdminBro in production create a Teaser error with infinite call. The trick is to modify the env production or to force the minify to false in the source code.

The monitoring is enable by default and can be viewed live on the endpoint /status :

http://localhost:5000/status

Helmet is automatically installed. It will setup the HTTP header of the app aappropriately and will hide some of them for the hacker. Helmet is not perfect but will help with cross-site scripting attack, clickjacking and some others.

For running the API, a single command is needed. You might want to use the SSH Tunneling.

npm run start

The continuous deployment is made with Travis with the informations inside the .travis.yml. The SSH key for accessing the server from Travis has been encoded and is passed to travis with the file deploy_key.enc. The env variable key and iv needed for decrypting the file has been added to the setting of travis.

MIT - Copyright © JUSTAL Kevin