How redhatqe/tfacon works
Project type and purpose
• This is a command-line integration tool (tfacon) written in Go (with a Python pip package wrapper) whose job is to connect Test Management Platforms (currently Report Portal) to an AI/ML “Test Failure Analysis Classifier” (TFA) and an optional Recommendation Engine (RE).
• It is an open-source connector/CLI tool that fetches test items, invokes external ML services to predict defect types, and updates the test platform.
File structure
• Top-level directories:
• cmd/ → CLI commands (root, list, run, validate, init, revert)
• core/ → Orchestration of core logic (fetch data, drive predictions, assemble results)
• connectors/→ Abstraction and implementation of connectors (Report Portal “RPCon”, placeholders for Polarion, Jira)
• common/ → Shared types, helpers, config parsing and validation
• pip_package/ → Python packaging for pip install
• Configuration files:
• tfacon.yml (user parameters for launches, tokens, URLs)
• tfacon.cfg (advanced tool behavior: concurrency, retry, attributes)
• Build & CI: Makefile, Go modules, GitHub Workflows, linting, publishing to PyPI.
• Tests and examples live under test_data, examples, and *_test.go files.
README for additional insights
• Describes CLI commands and flags, sample outputs, and a UML graph of the internal package/module dependency.
• External systems: Report Portal API, TFA classifier API endpoint, optional Recommendation Engine endpoint.