CodelyTV Scala HTTP API

Project showing up how you could implement a HTTP API with Scala.

This is the first iteration of the project where you will find a very Object Oriented approach. We've followed the Ports & Adapters (or Hexagonal Architecture) Software Architecture using traits as the domain contracts/ports in order to be implemented by the infrastructure adapters.

Contents

• Endpoints
• Libraries and implementation examples
• Environment setup
  • Install the needed tools
  • Prepare the application environment
  • Run the tests and start the HTTP server
  • Pre-push Git hook
• Logs
• Deploy
• About
• License

Endpoints

One of the goals of this project is to serve as an example for the course on Scala HTTP API (Spanish) illustrating how to implement several concepts you would commonly find in any production application. In order to accomplish so, we have implemented the following 5 endpoints:

• GET /status: Application status health check.
• POST /videos: Create new video inserting it into the database and publishing the video_created domain event to the message queue.
• GET /videos: Obtain all the system videos.
• POST /users: Create new user inserting it into the database and publishing the user_registered domain event to the message queue.
• GET /users: Obtain all the system users.

Libraries and implementation examples

Feature	Library	Implementation examples
Build tool	SBT	Dependencies, configuration & build.sbt
Style formatting	ScalaFmt	Rules
HTTP server	Akka HTTP	Routes definition, server implementation, Video POST controller & its corresponding acceptance test
JSON marshalling	Spray JSON	User & User attributes marshallers
Database integration	Doobie	Video repository & its corresponding integration test
Domain events publishing	Akka RabbitMQ	Publisher implementation & its corresponding integration test
Infrastructure management	Docker	Docker Compose definition
Logging	ScalaLogging + Logback + Logstash encoder	Logback configuration, logger implementation & its corresponding integration test
Command line command	Scopt	Database tables creation script
Distribution/deploy	SBT Native packager	Build & deploy instructions
Continuous Integration	Travis CI	Travis definition
Acceptance tests	Akka HTTP TestKit	Previously specified acceptance tests
Integration tests	ScalaTest	Previously specified integration tests
Unit tests	ScalaTest + ScalaMock	Video creator use case test
Continuous Integration	Travis CI + SBT Coveralls	Travis definition

Environment setup

Install the needed tools

1. Clone this repository: git clone https://github.com/CodelyTV/scala-http-api.git scala-http-api
2. Download and install Docker compose. We'll need it in order to run all the project infrastructure.
3. Download and install SBT

Prepare the application environment

1. Copy the Docker environment variables config file and tune it with your desired values: cp docker/.env.dist docker/.env
2. Start Docker and bring up the project needed containers: cd docker/; docker-compose up -d; cd ..
3. Create the database tables in your Docker MySQL container: sbt createDbTables

Run the tests and start the HTTP server

1. Enter into the SBT console: sbt
2. Run the tests: t
3. Start the local server: run
4. Request for the server status: curl http://localhost:8080/status
5. Take a look at the courses related to this repository (Spanish) just in case you're interested into them!
   • Introducción a Scala
   • API HTTP con Scala y Akka

Pre-push Git hook

There's one Git hook included. It's inside the doc/hooks folder and it will run the prep SBT task before pushing to any remote.

This prep task is intended to run all the checks you consider before pushing. At this very moment, it try to compile and check the code style rules with ScalaFmt.

You can define what this task does modifying the prep task in the build.sbt file. We like the approach of just running 1 single SBT task as the hook instead of multiple tasks because it's more efficient (the hook doesn't have to run SBT multiple times), and also because this way we can control the pre push tasks with the SBT alias defined at the build.sbt without altering the hooks.

In order to install this hook, just cd doc/hooks and run ./install-hooks.sh.

Logs

We've added a logging mechanism thanks to logback and logstash-logback-encoder in order to:

• Output the log records through the standard output channel (usually, your terminal :P)
• Store the log records in JSON format in a log file available at var/log/app_log.json
• Compress the historical log files into var/log/app_log-%d{yyyy-MM-dd}.gz files
• Delete compressed historical logs older than 10 days

If you want more information on the logging policies and appenders, take a look at the logback.xml.

Deploy

We use SBT Native Packager in order to package the app in single Jar file that you can execute.

1. Create the universal package: sbt universal:packageBin.
2. Extract the generated zip: unzip target/universal/codelytv-scala-http-api-1.0.zip -d ~/var/www/ which will contain:
   • bin/: All the executable binaries of our main classes in Unix and Windows (bat) format
   • lib/: All the project dependencies jar files.
3. Run the main app binary:
   • Without specifying any parameters (OK for this example app): ~/var/www/codelytv-scala-http-api-1.0/bin/codelytv-scala-http-api
   • Specifying parameters for the JVM: ~/var/www/codelytv-scala-http-api-1.0/bin/codelytv-scala-http-api -Dconfig.resource=application/$CONFIG_PATH
   • Specifying application parameters: ~/var/www/codelytv-scala-http-api-1.0/bin/codelytv-scala-http-api -Dconfig.resource=application/$CONFIG_PATH -- -appParam

About

This hopefully helpful utility has been developed by CodelyTV and contributors.

We'll try to maintain this project as simple as possible, but Pull Requests are welcome!

License

The MIT License (MIT). Please see License for more information.