Two Dimensional Visualization of Software Metrics

Static program analysis is a methodology which means the observation of a software without its execution. Static analysis techniques target testing, correctness checking, comprehending and other purposes [Crawford et al. 1985; Gyimothy et al. 2005]. CodeCompass [Ericsson 2017] is a static analysis framework which gives opportunity to perform various kind of analysis on the source code and present the results in different visualization methods. The architecture of this framework consists of two orthogonal layers. On a vertical layer it has a classical server-client architecture, since the statically collected information is presented by a web server towards a GUI or other querying client scripts. And on the horizontal layer the server and the client are implemented as independent plug-ins. Plug-ins provide the different functionalities of the framework. These can examine the code base from a wide range of aspects: some of them are language parsers which collect the named entities of a given language (variables, functions, classes, etc.) for further processing, others are inspecting the version control history of the project, yet others provide additional data from external sources, like third party databases. In this paper first in Section 2 we overview the main features of CodeCompass, an open source code comprehension framework. We discuss the most important software metrics we measure in Section 3.