GitHub Actions

Instantly attribute failed tests to changed lines

The LLM-Based module is designed to enhance the efficiency of code reviews by automatically analyzing GitHub Actions log data. This module fetches log data, extracts information about failures, and attributes these failures to specific symbols in the code. By providing precise context and actionable insights, it helps PR authors quickly resolve issues, facilitating smoother and faster merges.

How it Works

1. Data from GitHub Actions is fetched upon every commit and retrieves the latest logs for a given workflow run.

2. Once the log data is retrieved, the module processes the logs to extract information about test failures. This involves scanning the logs for common failure patterns and capturing relevant details.

3. After identifying the failures, the module attributes each failure to specific symbols in the codebase. This step involves:

   • Parsing the failure messages to extract file paths and line numbers.

   • Mapping these locations to functions, methods, or classes in the source code.

4. The module generates detailed insights based on the extracted failure information and the corresponding code symbols. These insights include:

   • A summary of each failure.

   • The exact location in the code where the failure occurred.