Abstract:
Program slicing is helpful in finding out the syntactic sub classes of the original program. It is mainly based on program variables and preserves the execution of the program. This knowledge is important in finding out which test cases are important or required for rerunning in post maintenance scenario. Historically Structural Testing is based on path selection for CFG and Data Flow Graph (DFG) and in CFG it is difficult to find where fault exists. The main task is to find all the relevant statements and conditions that are affected by the overall structural variables based on Abstract Syntax Tree (AST) which is normally based on data flow graphs and control flow graphs. We have introduced an alternative to control flow graph and slicing has been done considering AST and generate syntactic slices. We seed faults in both CFG and AST and compare test cases based on slices (execution paths) extracted from AST and CFG to demonstrate the effectiveness of our approach. The results show significant improvement in fault detection using AST and we get syntactic slices of code instead of semantic slices.
