Recent Releases of pi-scanner
pi-scanner - v1.2.1 - Critical Bug Fixes
🐛 Bug Fixes
Critical Fixes
- CRITICAL: Phase 2 reports now correctly show LLM validation results instead of duplicating Phase 1 data
- Fixed SWIFT/BIC detection false positives on SQL keywords (DISTINCT, ORIGINAL, INPUTFORMAT, COALESCE)
- Fixed name detection false positives in SQL contexts (e.g., 'Everyday Banking', 'Staging Table', 'Contract Account')
Improvements
- Added comprehensive ISO 3166-1 country code validation for SWIFT codes (249 countries + XK for Kosovo)
- Improved LLM progress tracking to show "findings" being validated instead of "files"
- Enhanced SQL keyword filtering with extensive list of SQL-specific terms
📋 Technical Details
- Corrected SWIFT pattern from
[A-Z]{6}to proper[A-Z]{4}[A-Z]{2}format per ISO 9362 - CSV export now includes LLM validation fields:
LLMValidated,LLMRiskLevel,LLMExplanation - Added context-aware detection with
currentFiletracking for SQL filtering - Fixed test failures in banking and driver license tests
🚀 Getting Started
Download the appropriate binary for your platform below, or use Docker:
bash
docker pull ghcr.io/macattak/pi-scanner:1.2.1
Note: Docker images will be pushed separately after resolving authentication.
- Go
Published by MacAttak 12 months ago
pi-scanner - PI Scanner v1.1.0 - Enhanced Detection & Enterprise Validation
PI Scanner v1.1.0 Release Notes
Release Date: June 20, 2025
Version: v1.1.0
🎯 Major Improvements
✅ Enhanced Detection Engine
- BSB detection rate: 100.0% (3/3) - significantly exceeded 80% target
- Multi-language support enhanced: Java (100%), Scala (94.4%), Python (90.0%)
- Context-aware filtering: 100% accuracy in distinguishing test vs production code
- Real-world pattern recognition: Improved detection in complex code structures
🏢 Enterprise Features
- Comprehensive business validation metrics with accuracy, precision, recall, F1 scores
- Performance benchmarking: 1058.5 files/second processing speed
- Risk assessment and compliance reporting for enterprise deployment
- Quality assessment framework with empirical testing methodology
🛡️ Repository Modernization
- Removed all ML dependencies - simplified deployment (no more tokenizer libraries)
- Simplified Docker build - Alpine-based container for smaller footprint
- Updated documentation and developer onboarding guide
- Streamlined build system - cleaned Makefile and removed obsolete targets
- Enhanced security - proper .gitignore for security audit reports
🧪 Test Infrastructure
- Real-world test datasets - 6 comprehensive scenarios across different domains
- Multi-language test framework - automated testing across Java, Scala, Python
- Business validation test suite - enterprise-grade quality metrics
- Context-specific detection validation - test vs production code differentiation
📊 Performance Metrics
| Metric | Result | Target | Status | |--------|--------|--------|---------| | BSB Detection Rate | 100.0% | 80% | ✅ Exceeded | | Java Test Pass Rate | 100.0% | 70% | ✅ Exceeded | | Scala Test Pass Rate | 94.4% | 70% | ✅ Exceeded | | Python Test Pass Rate | 90.0% | 70% | ✅ Exceeded | | Context Filtering | 100.0% | 70% | ✅ Exceeded | | Processing Speed | 1058.5 files/sec | 10 files/sec | ✅ Exceeded |
🚀 What's New
Detection Improvements
- Fixed BSB detection in Python test cases (literal newline issue)
- Enhanced pattern matching for complex code structures
- Improved confidence scoring for different PI types
- Better handling of formatted PI data (dashes, spaces)
Enterprise Validation
- Added comprehensive business validation framework
- Performance benchmarking and reporting
- Risk assessment with compliance recommendations
- Quality metrics aligned with enterprise standards
Developer Experience
- Modernized developer guide with Make-based workflow
- Simplified build process (no ML dependencies)
- Enhanced documentation structure
- Better onboarding for new developers
🐳 Docker Usage
```bash
Pull the latest image
docker pull ghcr.io/macattak/pi-scanner:v1.1.0
Run scan
docker run --rm -e GITHUBTOKEN=$GITHUBTOKEN \ -v $(pwd)/output:/home/scanner/output \ ghcr.io/macattak/pi-scanner:v1.1.0 \ scan --repo github/docs ```
📦 Installation
Binary Downloads
Download the appropriate binary for your platform:
- macOS (ARM64):
pi-scanner-darwin-arm64 - macOS (Intel):
pi-scanner-darwin-amd64 - Linux (AMD64):
pi-scanner-linux-amd64 - Linux (ARM64):
pi-scanner-linux-arm64 - Windows (AMD64):
pi-scanner-windows-amd64.exe
From Source
bash
git clone https://github.com/MacAttak/pi-scanner.git
cd pi-scanner
make build
🔒 Security
All binaries are signed and checksums are provided in checksums.txt. Verify downloads:
```bash
Verify checksum (example for macOS ARM64)
shasum -a 256 pi-scanner-darwin-arm64
Should match: ac1c2bc0e7c663a19a5bc0ed1e0d957f93093d01d7a9448dcbaf29d2ac802e21
```
🛠️ Breaking Changes
- Removed ML dependencies: No longer requires tokenizer libraries or ONNX runtime
- Simplified deployment: Docker image significantly smaller
- Updated configuration: Some configuration options simplified
- Repository structure: Cleaned up obsolete ML-related files
🐛 Bug Fixes
- Fixed BSB detection in multi-language test framework
- Resolved context filtering edge cases
- Improved pattern matching stability
- Fixed Docker build optimization
📚 Documentation
🤝 Contributing
We welcome contributions! See CONTRIBUTING.md for guidelines.
📞 Support
- Issues: GitHub Issues
- Documentation: docs/
- Security: Report security issues via GitHub Security tab
🙏 Acknowledgments
Special thanks to all contributors and the open-source community for making this release possible.
Full Changelog: v1.0.0...v1.1.0
- Go
Published by MacAttak about 1 year ago