fix(forge test): ensure --rerun only runs failing tests from correct contracts

[dghelm]

Motivation

Fixes issue #10693 where forge test --rerun incorrectly runs passing tests that have the same name as failing tests in different contracts.

Solution

Implemented qualified test matching with contract-aware filtering while maintaining full backward compatibility:

Key Changes

1. Extended TestFilter trait with matches_qualified_test(contract_name, test_name) method for context-aware filtering
2. Enhanced FilterArgs system with qualified_failures field to store specific contract/test combinations
3. Smart failure persistence that intelligently chooses between qualified format (ContractName_testName) and legacy format (testName1|testName2) based on whether multiple contracts are involved
4. Updated test runner logic in is_matching_test_in_context to use qualified matching when appropriate

Architecture

• Intelligent format detection: Uses qualified format only when test names are ambiguous across multiple contracts
• Graceful fallback: Maintains legacy behavior for single-contract scenarios
• Conservative approach: Minimizes format changes to preserve existing workflows

Files Modified

• crates/common/src/traits.rs - Extended TestFilter trait
• crates/forge/src/cmd/test/filter.rs - Added qualified matching logic and fixed test name format consistency
• crates/forge/src/cmd/test/mod.rs - Enhanced failure persistence and loading with intelligent format selection
• crates/forge/src/multi_runner.rs - Updated test matching logic to use qualified context
• crates/forge/tests/cli/test_cmd.rs - Added comprehensive regression test

Backward Compatibility

All existing functionality is preserved:

• Single-contract scenarios continue using legacy format
• Existing test failure files work unchanged
• All CLI arguments and behavior remain the same

Side Effects and Considerations

Potential Side Effects

1. Test failure file format changes: Files may now contain qualified names (ContractTest_testName) in multi-contract scenarios
2. Filter matching behavior: Slightly different matching logic when qualified failures are present
3. Performance impact: Minimal additional processing to detect ambiguous test names

Mitigation

• Backward compatibility preserved: Legacy format still used for single-contract scenarios
• Graceful fallback: System falls back to legacy behavior when qualified parsing fails
• Conservative approach: Only uses qualified format when absolutely necessary

PR Checklist

• [x] Added Tests
  • Added should_replay_failures_only_correct_contract regression test
  • Verified existing should_replay_failures_only test continues to pass
• [x] Added Documentation
  • Comprehensive inline code comments explaining the qualified matching logic
  • Updated function documentation for new matches_qualified_test method
• [x] Breaking changes

This contribution was made to explore the viability of bug-fixing work via Claude Code assistance, as prompted by this Tweet. If it's a mess, please don't spend too much time helping me get it over the line.