Building microservices testing framework

RESTful services and microservice architecture in general are big trends in the industry right now. I’ve noticed it also from this blog’s traffic analytics, where topics related with REST testing get the highest interest. Therefore I’ve decided to create some kind of example framework for RESTful APIs functional testing. Project is based on REST-Assured as a services client, Spock as a testing framework, Groovy and Gradle. You can clone full repository from my github. Tests are run against Wiremock API described in this post. Please consider this project as a kind of bootstrap, since it’s an example, not full-blown test framework. Ok, so as Linus said – talk is cheap, show me the code! (PS: and yes, I know that guy on the picture isn’t Linus Torvalds 😉 )

linus

Dependencies

Usually, first thing for me after importing new project is opening build.gradle (or pom.xml, in case of maven). In our example, the most important dependencies are REST-Assured and Spock. We’ve also Jackson, for serializing objects, Logback for our logs and snakeyaml for reading .yaml files, where we’d store properties. Let’s see the whole file:

Besides dependencies, interesting part here is testLogging.showStandardStreams = true property. This flag gives us ability to print log entries in stdout, which I find helpful in test debugging.

To have a greater understanding, let’s have a look at the project tree also. I’ve divided files into context packages, so we have: testflow package, with our test classes, next there is service, with api’s clients, dto and factory, with our entities and their builders, asserts with custom assertions and utils with yaml reader. There is also resources folder, where we store logback configuration and test.yaml with our properties and constants:
project

Test class

The core of our framework are testing classes in testflow package. Since it’s an example project, we have only one test (in TestCaseExampleSpec class):

Our test case is super simple – we build request with new user, post it to webservice, and then validate response. Since we’re using Spock, test steps are in BDD convention. You could noticed that test class extends another – BaseSpecification.java. For those who are not familiar with Spock, it’s for overriding default configuration. Every test class in Spock have to extend Specification class, and if we want to add some custom properties or setUp methods, We have to extend Specification and then extend out test classes:


Here, we’re setting up logging rule with test names, configuring default timeout, and setting up REST-Assured to use some less restricted https validation.

Service layer

Service classes are the engine of our framework. We’re storing here clients for our REST APIs. With use of REST-Assured, we’ve created following client:

I’m not going to dig down into details of framework. If you want to learn some more, you can view my past articles: REST-Assured: framework overview and REST-Assured: going deeper. In simple words, we get host of webservice under the key which we’ve declared in test.yaml file and post User entity to /addNewUser method. Then, we validate if response code was 200 and return response as a Map.

Entity and Factory

Entity, yes – no magic here. We have two properties:

Next thing is a UserFactory, where we build our entity. For the sake of test readability, I’ve created custom method buildNewUser() without any arguments, and then I invoke factory method and pass arguments to it. User login and www values are taken from test.yaml file, where we store them:

Assertions

If you’re familiar with REST-Assured, you can ask why we’ve created separate package with assertions, instead of doing them directly in service client methods. We could do that, but I prefer to be independent from specific frameworks as much as I can. Another point is possibility that we’d want to have some more complex assertions in the future. Therefore we’ve created UserValidation class, where we validate responses returned from our services:

Since it’s only example, we’re checking presence of login property only and it’s value. Although, validation classes would be ideal place for some serious validation and assertions.


Continue reading

If you want to continue reading and expand your knowledge in area of  REST and microservices, I recommend you these books:

  • Building Microservices – one of the most important books for me, everything you want to know about microservices is here
  • RESTful Web APIs – another great book about REST architecture. Lots of practical knowledge about designing and consuming RESTful APIs

Summary

My idea behind this post was to provide you some ideas for creating your own REST testing frameworks. We’ve omitted some parts of the project, like logback configuration or yaml reader – that’s because I found them less crucial. I encourage you to clone whole project from my github and to play a little with the code.

As a fan of continuous improvement, I’m sure that lots of you would have some thoughts and improvements for the code and idea of the project, so I hope that you’d share them with me in the comments!

  • Pingback: REST-Assured: going deeper | TEST DETECTIVE()

  • Pingback: REST-Assured framework overview | TEST DETECTIVE()

  • Abdul Qayum Mohammed

    Thanks for sharing an excellent article. Keep writing.

    Cheers,
    Abdul Qayum Mohammed

  • John Mike

    Hi,

    I’m a newbie in this field. I just have experience in C# and a little bit in Java. I downloaded your project source code already. However, i don’t how to open it. Could you please tell me what the editor used to develop this project is and if you can, please tell me the way to set up environment for it.

    • Łukasz Rosłonek

      Hi @disqus_GJbypW7edU:disqus !
      First of all, you need to have Java jdk installed end $JAVA_HOME system variable configured.
      As for the IDE (your editor), I recommend Intellij Idea, you can download community version for free. Having Intellij Idea installed, click on ‘import project’ and point directory of the source code you’ve downloaded. Then you chose ‘Import project from existing sources’ and go with ‘Gradle’. The last thing is to click ‘Finish’. Intellij will build the project and resolve dependencies for you – when it’s done, you’re ready to go!

      • Akshay Maldhure

        Once the project is built, how to run the tests?

  • Akshay Maldhure

    Hello,

    This is a great article, this is exactly what I’ve been looking for. Thanks for sharing it!
    However, I would like to develop a test framework for RESTful microservices in Java and REST Assured. And I do not understand what the code in each of your classes is doing. It would be great if you could elaborate more on that. Thank you!

  • Pingback: RESTful API Testing – VikramVI Knowledge Sharing()