A Survey on Program Analysis and Transformation

Program transformation is a process in which an input program is transformed into another program that achieves a specific goal. Such transformation is done by applying a sequence of transformation rules on the input program to generate another program as the output. Such transformations can be done manually with human intervention (software developer) or automatically by a transformation program applying a transformation algorithm. Automating this process has been of an interest to a myriad of researchers in the past years. Several researches were done to automatically find good transformation sequences to achieve different transformation goals including program optimization and test case generation. The most popular techniques used are search-based meta-heuristic algorithms including genetic algorithm. In this paper, we will survey previous works that used genetic algorithm to achieve optimization goals under the umbrella of program transformation.