Understanding Semantic Impact of Source Code Changes: an Empirical Study

Since source code is the ultimate definition of the behavior of a software product, changes to source code become the critical factor in understanding behavioral changes and predicting faults. In studies on source code changes, text or syntactic approaches have been widely used. Textual analysis focuses on changed text fragments while syntactic analysis focuses on changed syntactic entities. Although both of them have demonstrated their advantages in experimental results, they have only focused on changed code. Because of semantic dependencies within programs, we believe that code impacted by changes is also helpful. Given a source code change, we identify its impact by program slicing along the variable def-use chains. To evaluate the effectiveness of change impacts in fault detection and prediction, we compare impacted code with changed code according to size and fault density. Our experiment on the change history of a successful industrial project shows that: for large changes, their impacts have relative small size and high fault density; while for small changes, change themselves have relative small size and high fault density. Our study suggests that: 1) change impacts are complementary to change themselves in detecting or predicting faults; 2) within the impact of a large change, a high degree of interference between impacts of different changed lines contributes to the high fault density; 3) high fault density in impacts aggravates the danger of large