Search Based Slicing for Program Dependency Structures of Interest
Tao Jiang · 2008
Mark Weiser introduced program slicing: creating a subprogram defined by a set of variables and a specific statement. A large number of applications have since been developed based on this technique, from debugging and testing through program maintenance and integration to program metrics and refactoring. Recent develop-ments in the efficiency of slicing tools (i.e. Codesurfer) allow for slices to be computed for each program statement in a reasonable time. However, the technique presents a new challenge: How are we to identify useful subsets of slices to represent dependency structures of particular interest? This the-sis explores dependency structures in programs using a Search-Based Slicing (SBS) framework as a way of responding to this challenge. The framework combines general Search-Based Software Engineering (SBSE) theory and the program slicing technique to make it possible to explore structures of interest in the large number of potential subsets of slices. The goal of the SBS framework put forward in this thesis is as follows: to identify