Using software metrics to select refactoring for long method bad smell

摘要

Refactoring is a technique for improving software structure without changing its behavior which can be used to remove bad smells and increases software maintainability. But only few approaches have been proposed to address the identification of appropriate refactorings. Specifically, our research proposes a method to select refactoring based on software metrics which are defined in terms of data flow and control flow graphs. The method consist of 4 steps: 1) calculate metrics, 2) find candidate refactoring by using refactoring filtering condition (RFC), 3) apply a suite of candidate refactorings and compute maintainability, and 4) identify the refactoring that gives the highest maintainability. We demonstrate out approach by giving an example of removing a long method bad smell in a customer class in a movie rental system. Our approach proves to be able to suggest an appropriate set of refactoring techniques such as extract method, replace temp with query, and decompose condition, to solve the long method bad smell.