Continuous Delivery with Docker and Jenkins

Code coverage

Think about the following scenario: you have a well-configured Continuous Integration process; however, nobody in your project writes unit tests. It passes all the builds, but it doesn't mean that the code is working as expected. What to do then? How to ensure that the code is tested?

The solution is to add the code coverage tool that runs all tests and verifies which parts of the code have been executed. Then, it creates a report showing not-tested sections. Moreover, we can make the build fail when there is too much untested code.

There are a lot of tools available to perform the test coverage analysis; for Java, the most popular are JaCoCo, Clover, and Cobertura.

Let's use JaCoCo and show how the coverage check works in practice. ...