추출식 소프트웨어 제품라인공학을 위한 자산 컴포넌트 탐색 및 재공학 방법

Abstract

Software product line engineering (SPLE) is a reuse paradigm that helps organizations increase productivity and improve software quality by developing software systems from reusable assets. The extractive approach to SPLE is an emerging approach that capitalizes on legacy systems when developing reusable assets. With the extractive approach, organizations can reduce time and costs required to initiate a product line, compared to the proactive approach. For the success of the extractive approach, it is important to reduce asset development time by analyzing legacy systems and extracting reusable assets from them. However, as developers who have developed these systems often combined many functions in one component, it is difficult to understand and separate reusable assets. To address these challenges, this thesis suggests an approach that contains processes and tactics for analyzing legacy systems and extracting reusable assets from it. This approach consists of two major processes: 1) To analize the legacy system structure, the first process includes steps that define “component type” which can be used to automatically identify legacy components of the defined component type. From the identified legacy components, the analyst can understand the role of each component in the legacy system. 2) To separate reusable assets from the legacy system, the second process includes steps that help analyst evaluate cohesiveness of candidate components based on the component type. To demonstrate the feasibility of the proposed approach, I defined one component type, a component identification method, and a seperation method, and developed a tool that supports the methods. Based on the result of the feasibility study, the legacy components had been engineered.