Code coverage model

This library Provides a Java API to parse and collect code coverage results. It is used by my Jenkins' coverage plug-in to visualize the coverage of individual builds.

Additionally, this library is used by my additional Quality Monitor GitHub Action, that monitors the quality of projects based on a configurable set of metrics and gives feedback on pull requests (or single commits) in GitHub. There are also two additional actions available to autograde student software projects based on these metrics: GitHub Autograding action and GitLab Autograding action.

This library consists basically of two separate parts. The first part is the model to manage several metrics in a hierarchical software project. Metrics can be attached to modules, packages, files, classes, or methods. The values can be aggregated or merged so that it is possible to zoom into every detail level. For each node of the tree, the value of a metric can be computed using one of the following aggregations:

• OVERALL (default): the value of the whole subtree, e.g., the line coverage of all lines, the sum of the cyclomatic complexity of all methods, or the mean value of percentages like the class cohesion.
• MAXIMUM and MINIMUM: the largest or smallest value of the elements that provide the metric, e.g., the method with the highest cyclomatic complexity or the file with the worst line coverage. These elements are the methods for method metrics, the files for coverage metrics, and the classes for all other metrics. If a parser reports a metric on a coarser level only (e.g., the complexity of files), then these values are used.

The aggregated values of two trees can be compared as well, e.g., to compute the change of the maximum cyclomatic complexity with respect to a reference build.

Supported metrics are code coverage (line, branch, instruction), mutation coverage (mutation killed rate, test strength), tests (number of tests), and general software metrics (lines of code, non-commenting source statements, cyclomatic complexity, cognitive complexity, NPath-complexity, and class cohesion).

The second part consists of parsers for several report formats:

• Code Coverage
  • Cobertura code coverage results
  • Open Clover code coverage results
  • Go Coverage results
  • JaCoCo code coverage results
  • LCOV code coverage results
  • OpenCover code coverage results
  • VectorCAST code coverage results including MC/DC, Function, Function Call coverages
  • TRACE32 code coverage results

• Mutation Coverage
  • PIT mutation coverage results
  • Stryker mutation coverage results

• Test Results
  • JUnit test results
  • NUnit test results
  • XUnit test results

• Software Metrics
  • PMD software metrics via a patched version of PMD

All source code is licensed under the MIT license. Contributions to this library are welcome!