Representing source code with Granular hierarchical structures

Quality factors namely testability, reliability, and maintainability are considered vulnerable to software complexity. Analyzing complexity of code is difficult though. Many techniques have been invented, including control flow graph (CFG) to aid program complexity analysis. However, the representation of code with dasiawebpsila structures exploited in CFG incurs some difficulty to human comprehension. Referring to Granular computing recently emerging from cognitive theories, this research thus proposes a novel approach to representing source code with ldquogranular hierarchical structuresrdquo. Instead of representing a program with dasiawebpsila, the method uses multiple dasiatreespsila to promisingly obtain more understanding during source code analysis. Preliminary experiments showed that representing source code with granular hierarchical structures gained more competent analysis of program complexity. The results were evaluated by the invented complexity measure called SCIM that satisfies more ldquobasic needs of good software measuresrdquo, compared to McCabe´s Cyclomatic complexity derived from control flow graph.