Software systems deal with many concerns that are defined by the user or by the system requirements. These concerns are often implemented in multiple places throughout the source code, despite any decomposition technique that may be used. There are approaches and tools that try to separate concerns present in the system, usually using methods as the smallest unit of concern granularity. In this work we present an overview of such existing tools. Furthermore, we have conducted a case study to learn what level of granularity is considered useful for identifying concerns in the source code. The case study involved programmers identifying concerns within familiar codebases that covered multiple programming languages. The study shows the statement-level granularity is being generally used, while finer-grained concern resolution may not be of a great value.
展开▼
