Development of a System for Analysing Method Names in Java Source Code

Abstract

Naming code can seem like a simple task, however finding a good name can be rather challenging. Entity names should be consistent and brief yet comprehensive when representing the information each entity hold. What is considered a good name can be highly debatable, although it usually involves descriptive names that can contribute to readability and comprehensibility of source code. Bad code names can cause uncertainty, potential future bugs and be misleading. For this reason, the task of naming code is vital, hence there is a need of a system to improve and maintain it. To develop such a system, there are requirements required to be specified to define the expected implementation for certain entity names. These requirements are encoded into software in a domain-specific language, granting executable code to be generated from the expressed requirements. As a result, this name analysis tool provides programmers to perform code analysis on Java source code checking if the entities act in accordance with the requirements of their names. Additionally, the result shows insights of how contributions from linguistics can be valuable for software development and can be used to analyse software languages, such as entity names.