Year: 2014
Type: dissertation
Abstract: Background: Modifiability and Maintainability are important software qualities for existing software systems.However, structural complexities such as dependency cycles, are known to inhibit these software qualities.These architectural hotspots are common in many software systems, and the larger they get, the higher the chance that such flaws exists.Several studies have also shown that these structural complexities are difficult to refactor and can be difficult to detect without using code analysis tools.Research Goals: If we could improve the overall architecture by refactoring a select few components, this could translate into a better, more maintainable software system.The main goal of this research is therefore to create a plugin that could help developers to reduce these structural complexitites and finding suitable candidates to refactor.The plugin will be a simulation tool which can simulate refactoring and visualize dependencies between components in a graph that can be manipulated.Approach: My main researh methods will be Desgin Science and I will use several software engineering frameworks to create the architecture of the plugin.I will perform tests to verify that the plugin meets the requirements that I have set, and I will use case studies to verify how useful the plugin is and how it can help developers to refactor source code.Expected Results: A plugin for Visual Studio, for analysing C# source code.The plugin should be able to simulate refactorings based on the analysis of the source code.The user should be able to simulate breaking cycles, moving edges and adding intermediate nodes.This tool should help developers to restructure and decouple the architecture at the class granularity level in order to make the system more maintainable.
Source: BIBSYS Brage (BIBSYS (Norway))
Author: Christian Thurmann-Nielsen
Language: English
Subfield: Information Systems
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W2502088138