A case study in branch testing automation

We present areaI world experience with some recent research results aimed at improving the branch testing processo The results considered consist into (i) a method for the automatic derivation of sets of paths that satisfy the branch testing criterion and into (ii) a bound on the number of test paths needed for achieving 100% branch coverage. The derivation of test path sets (point i) is based on a new method for the static analysis of the program flowgraph; the method has been implemented within a prototype rool, called BAT. The computation of the bound (point ii), called Pbranch, considers those control flow paths with a 10w number of predicates, that are more likely feasible. BAT and Pbranch, have been validated on object oriented C++ software, developed to control a new generation of telecommunications systems, within the test environment of Ericsson Telecomunicazioni. The case study regarded the basic test phase, that is the first step of the testing process and is appropriate for the application of the proposed theory, since it includes the branch testing of program units up to a prescribed coverage measure. While the experimentation is stili on-going, we describe some preliminary results.