Pynguin 0.40.0

• Provide a (normalised) area under curve for timeline output variables.

  Variables for statistics that produce a timeline, e.g., coverage development over time imply in interesting property: while the final coverage of two distinct runs/configurations might be equal, one might have a faster coverage increase than the other, thus could be considered better. The area under curve, i.e., the integral over the development function allows to access this property easily.

• Add the ability to write the logging to a log file and not only STDOUT.

• Add an auto-deploy action to GitHub to push releases to PyPI.

• Fix a typo in the documentation (cf. GitHub issue #75).

• Switch to ruff-format.

Pynguin 0.39.0

• Fix a bug in the handling of infinite loops in global scope (thanks to @BergLucas in #65)

Pynguin 0.38.0

• Remove the dependency to our custom MutPy fork by integrating the relevant code into Pynguin directly (thanks to @BergLucas in #64)

Pynguin 0.37.0

• Add possibility to control number of mutations in MOSA.
• Compute the area under curve if Pynguin traces the overage over time; area under curve can be interpreted as the integral over the coverage function, which is computed via trapezoidal approximation.
• Fix the assertion generation for floats (thanks to @BergLucas in #70)
• Fix assertion and invalid function call (thanks to @BergLucas in #72)

Pynguin 0.36.0

• Remove unused code
• Fix ruff warnings
• Add sequence variable for type-evolution tracking
• Add CLI options to ignore methods and modules from analysis (see #62)

Pynguin 0.35.0

• Fix TypeError bug in instrumentation of bytecode (see GitHub PR #51)
• Add a dump method for type-information statistics
• Fix handling of aliased modules (see GitHub issue #57)
• Fix method-signature handling for C extensions (see GitHub issue #59)

Pynguin 0.34.0

• Activate a larger selection of checkers for ruff
• Implement various stopping conditions

Pynguin 0.33.0

• Fix an issue with type-information retrieval (see #35 )

Pynguin 0.32.0

• Breaking: Restructure the code
  Most of the code related to the generation algorithms has been moved. The code now resides in the ga package, where it fits much better. All accompanying modules have also been moved to different places. This might cause breakage if you have developed extensions for Pynguin that rely on the structure of the modules.
• Update the bytecode library to its latest GitHub version.
  Necessary to include a fix for a regression in that library, which was pointed out to us in GitHub issue #42 (hopefully fixed by this library upgrade).
• Adjust the header of generated test files; we now refer to Pynguin's website.
• Extend the usage of the ruff tool for code analysis.

Pynguin 0.31.0

• Update literature references in documentation
• Add issue templates for GitHub