On the extent and nature of software reuse in open source java projects

Abstract

Code repositories on the Internet provide a tremendous amount of freely available open source code that can be reused for building new software. It has been argued that only software reuse can bring the gain of productivity in software construction demanded by the market. However, knowledge about the extent of reuse in software projects is only sparse. To remedy this, we report on an empirical study about software reuse in 20 open source Java projects with a total of 3.3 MLOC. The study investigates (1) whether open source projects reuse third party code and (2) how much white-box and black-box reuse occurs. To answer these questions, we utilize static dependency analysis for quantifying black-box reuse and code clone detection for detecting white-box reuse from a corpus with 6.1 MLOC of reusable Java libraries. Our results indicate that software reuse is common among open source Java projects and that black-box reuse is the predominant form of reuse. © 2011 Springer-Verlag.