An Extensible Feature-Oriented Approach for Fine-Grained Code Quality Analysis
summary
Presents a concrete pipeline that maps static analysis issues to features using InsightMapper and feature-location output from Mobioos Forge, with SonarQube as the quality oracle and JaCoCo added for per-feature test coverage. Applies to ArgoUML (~400k LOC) and eShopOnContainers (~130k LOC); finds a small “result” feature with 3% of LOC causing 12% of issues, connective/login heavily cross-cutting, and basket consuming ~45–48% LOC but ~80% of estimated fix time. Outputs feature-oriented quality outcomes to enable feature-priority or criticality-based remediation.
AI-written summary · about
chapters (14)
- 0:00Introduction
- 1:29Code quality analysis limits
- 3:39Feature location for quality
- 4:42Insight Mapper approach
- 8:58ArgoUML validation
- 11:32ArgoUML results
- 12:16eShop validation
- 13:21eShop results
- 14:02Extensibility of Insight Mapper
- 15:12Live demo: Insight Mapper
- 18:47Adding projects to Insight Mapper
- 20:16Prioritizing fixes
- 22:03Future research
- 22:31Q&A
related talks
43mCode quality as an engine of your IT business
49mTreat Your Code as a Crime Scene
14mWhich Code Issues Actually Matter? | Prioritizing SonarQube Findings
42mFeature- and time-based software analysis - Balancing features, effort & quality
54mResolving technical debts in software architecture
49mMetrics for Architects
48mComprehensive Use of Analyzers to Improve Code Quality
33mCode Quality Metrics to Measure and Quantify Quality of Code