Ylläpidettävyyden metriikoiden toteutuksen arviointi staattisen analyysin työkaluissa
Riku Rauhala · Työväentutkimus Vuosikirja · 2026
Software maintenance work consists of tasks such as fixing bugs, patching vulnerabilities, and upgrading dependencies. Maintainability refers to how effortless it is to make changes. The level of maintainability of a code base can be quantified using metrics that measure different aspects of source code, such as size and complexity. Automated assessment of source code using the metrics is called static analysis. To enable software developers to assess the maintainability of their projects, various static analysis tools and services have been developed. It is not always clear which tools to select and which metrics best measure maintainability. This thesis presents an evaluation of four static analysis tools: SonarQube, Code Climate Quality, Qlty, and Codacy. For each of the tools, the implementation details of six core high-level metrics are evaluated: lines of code, cyclomatic complexity, cognitive complexity, duplication, code coverage, and code churn. The results of the evaluation show that SonarQube offers the most comprehensive set of metrics. Qlty implements the core metrics and offers a viable alternative to SonarQube.