Abstract
Many statechart-based testing strategies result in specifying a set of paths to be executed
through a (flattened) statechart. These techniques can usually be easily automated so that
the tester does not have to go through the tedious procedure of deriving paths manually to
comply with a coverage criterion. The next step is then to take each test path individually
and derive test data, i.e., fully specified test cases. This requires that we determine the
system state required for each event/transition that is part of the path to be tested and the
input parameter values for all events and actions associated with the transitions. We
propose here a methodology towards the automation of this procedure, which is based on
a careful normalization and analysis of operation contracts and transition guards written
with the Object Constraint Language (OCL). It is illustrated by two case studies that
exemplify the steps and provide a first validation. Though many of the steps are
automated, some inputs are still required from the modeller and tester and further work is
necessary to push the automation even further and solve a number of remaining issues.
This report is a first attempt at clarifying the issues with the objective to support as much
of the UML statechart notation as possible and to identify the analysis steps required in
such an endeavour.
Users
Please
log in to take part in the discussion (add own reviews or comments).