How does it work?

The mutation testing analysis of MuTalk works with two main elements: the source code to mutate and the tests that go with it.
With those, the analysis then goes through 4 phases:

• an initial test run
• a coverage analysis
• the generation of mutants
• the evaluation of mutants, also called the results generation

The initial test run

During this phase, the analysis will run every tests it was given. If a test fails, the analysis stops and inform the user that they should fix the tests. To avoid this, check the red test filter.

The coverage analysis

This phase is a regular coverage analysis used for test selection and mutant selection strategies.

The mutant generation

During this phase, the analysis creates the mutants of the source code using its mutant operators.

The results generation

In this phase, the analysis takes the mutants and apply them one by one. It installs a mutant and runs each selected tests with the mutant, this is the mutant evaluation. The evaluation stops at the first failing test or when every test has been run. Then the mutant is uninstalled, and it moves on to the next mutant.