fix: show actual test file paths in failure log

[mashraf-222]

Problem Fixed

When tests fail to run, the error log shows misleading information:

Tests '[None, None, PosixPath(...)]' failed to run

This happens because the log prints original_file_path which is intentionally None for AI-generated tests, making it appear as if test paths weren't set correctly.

Root Cause

In parse_test_output.py:1222, the failure log message was:

f"Tests '{[test_file.original_file_path for test_file in test_files.test_files]}' failed to run"

For AI-generated tests:

• original_file_path = None (intentional - these are newly generated)
• instrumented_behavior_file_path = actual path being executed
• benchmarking_file_path = performance test path

The log should show the paths actually being used, not the intentionally-None original paths.

Solution Implemented

Changed the log to show instrumented_behavior_file_path or original_file_path (whichever is set):

test_paths_display = [
    str(test_file.instrumented_behavior_file_path or test_file.original_file_path)
    for test_file in test_files.test_files
]
logger.info(f"Tests {test_paths_display} failed to run, skipping")

Code Changes

File: codeflash/verification/parse_test_output.py

• Line 1220-1223: Updated failure log to show actual test paths
• Added comment explaining why we prefer instrumented_behavior_file_path

Testing

Verified the change improves log clarity when:

• AI-generated tests fail to compile/run
• Existing unit tests fail to execute
• Mixed test scenarios (generated + existing)

Impact

This is a logging-only change with no behavioral impact. It improves debuggability by:

• Showing actual file paths being executed
• Eliminating confusing None values in logs
• Making test execution failures easier to diagnose

🤖 Generated with Claude Code