Gain more confidence in your code with mutation testing

The core component of our work is code, and we all want to write the best one. But

how do we ensure that it is up to standard? We test it. The test results will tell us if

we did our job correctly. But how do we measure whether our code is good enough?

A common test metric like code coverage can be misleading, hitting 100% code

coverage does not necessarily mean effective tests. And if we find a metric, where

do we draw the line?

There is a methodology of testing which more objectively assesses if your test suite

is reliable enough or even how many unwanted changes to production code could go

undetected.

This methodology mutates your source code and runs the unit tests against it. If your

tests still pass, it means that your tests are not highly effective, or they leave some

potential bugs behind. If your tests fail, then you have written effective tests.

The presentation will have these key points:

• Applying this methodology will extend your insight into the quality of your tests

and your code.

• You will have an objective metric in test strength and gain confidence in your

code.

• We will show how we do it, highlight the approach, lessons learned and show you

interesting features.

After this talk, we will understand how much testing is important for our daily work

and will put effort into improving how we do it and therefore improve the production

code quality.

Also, it will help us to change mentality to test first and develop afterwards, making

mutation testing just a confirming tool of the excellent job that we are already doing.