Institutional Repository of Vilnius University
Programinio kodo aprėpties ataskaitų interpretavimo modelio kūrimas ir taikymo tyrimas /
Abstract
dc:descriptionTools such as JaCoCo produce code coverage reports with several coverage metrics but the raw data has four limitations: there is no overall quality indicator, package-level aggregation ignores code complexity, uncovered code is presented without prioritization and reports cannot be compared with each other. This work proposes an interpretation model that derives additional metrics from JaCoCo XML data, computes a composite quality score, aggregates class-level metrics to the package level using cyclomatic complexity weights, classifies risky code locations into four categories and compares two reports. The model is implemented as a web application, evaluated against the ISO/IEC 25010 standard and empirically applied to three open-source projects. The study showed that no single JaCoCo metric can serve as an overall quality indicator - in Jsoup, 71% of tested classes had a spread of more than 30% between their highest and lowest coverage metric. Complexity weighting at the package level reveals risks that a plain average hides: package rankings shifted by up to 13 positions. Risky code location detection reached 88% overall accuracy.
Degree
thesis:*- Grantor dc:publisher
- Institutional Repository of Vilnius University
- Year dc:date
- 2026
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Sabestinas, Tadas,
Rights
dc:rights- Statement dc:rights
-
- info:eu-repo/semantics/openAccess
- Language dc:language
- lit
Identifiers
dc:identifier.*- Repository record dc:identifier
- https://repository.vu.lt/VU:ELABAETD308338692&prefLang=en_US