Relating Source-model Data with Version and Change Data
Florian Havlicek · 2008
As time passes by, software systems undergo a lot of changes and each work over costs money. Adaptations often result in the majority of funds being invested into these makeovers. Hence keeping the costs for changes as low as possible should be a priority of all companies. To perform this task the structure of software systems has to be analysed. Not only should the current state of the source code, but also the evolution of it be investigated. The data stored in the version control system is enough to do those analyses. Based on the source model data of one release of a software project, the new addresses of the entities in another release should be calculated. Both versions have to be stored in the local file system. Each of the pairs of equivalent source code files is compared with the diff-command. To analyse the differences made to the compared release, six cases of the relative positioning of the source code changes to the entities are defined. During the computation process statistical values for each entity are logged. This allows us to get evolution data on the method level which can be used for deeper analysis of the software project.