White-Box Testing

In contrary to blackbox tests, whitebox testing refers to testing activity which is based on knowledge about the tested system. An example for whitebox tests are tests which are based on coverage criteria, e.g. to achieve that every statement was executed at least once or for model-based software, that every block was executed once during testing.

Key Benefits

  • Record and evaluate internal signals
  • Override internal signals
  • Inject faults internally
  • Coverage analysis throughout the model
White-Box Model

Besides the support for measurements of coverage metrices during test executions, Arttest extends the possible tests with knowledge about system interna on model level. It enables the tester to select arbitrary signals from the model and use them in the test either to be evaluated, to be observed and used as trigger or condition for specified events or to be overridden by specified signal data. Thus, it is possible to test internal system behaviours for correct behaviour, e.g. after an error was fixed. Moreover, the mentioned features enable you to profit from many additional benefits, depending on your application, such as

  • Speeding up test executions, since models can be driven faster to initial states for further testing purpose
  • Closed loop tests, where the system under test is combined with a model of its environment
  • Fault injection tests on model level, which would otherwise only be possible in later development stages
  • Integration tests, where many sub-models are integrated and tested together
  • And many more…