Pythia: a regression test selection tool based on textual differencing

Regression testing is a commonly used activity whose purpose is to determine whether the modifications made to a software system have introduced new faults. For many large, complex, software systems the retest all strategy is not practical: the resources required to reexecute and verify all available test cases (i.e., time and human effort) are prohibitive. Ad hoc methods are not desirable, as they can compromise the reliability of the regression test activity and consequently the reliability of the software system being tested. In this paper we present a new technique for selecting regression test cases based on the modifications that have been made on the program. The technique, which is based on the idea of directly comparing source files from the old and the new version of the program, has been implemented in a tool called Pythia. A novel characteristic of Pythia, which is capable of analyzing large software systems written in C, is that it has been implemented primarily through the integration of standard, well known, UNIX1 programs.