Allyanonimiser

Added
- NSW legacy driver's license pattern support (licenses issued until 1990)
- Comprehensive pattern reference table in documentation
- Enhanced pattern management examples in README

Fixed
- Fixed package build with proper dependency specifications
- Added missing dependencies in pyproject.toml (pandas, tqdm)
- Updated documentation with clearer pattern examples

Added
- Comprehensive reporting system with rich visualization capabilities:
- Added `AnonymizationReport` class for tracking statistics and metrics
- Added `ReportingManager` class for handling report sessions
- Added integration with Jupyter notebooks for rich visualizations
- Added multiple export formats (HTML, JSON, CSV) for report sharing
- Added detailed entity type tracking and distribution analysis
- Added performance metrics tracking for anonymization operations
- Added document-level statistics and batch reporting features
- Added visual charts for entity distribution and operator usage
- Created example scripts demonstrating reporting capabilities

Enhanced
- Integrated reporting system with existing functionality:
- Added reporting to anonymize(), process(), and process_files() methods
- Added report display capabilities for Jupyter notebook environments
- Added methods to start, retrieve, and finalize reports
- Added automated reporting for batch operations
- Improved reporting of anonymization effectiveness with rate metrics
- Enhanced documentation with comprehensive reporting examples
- Added test suite for reporting functionality

Added
- Stream processing for very large files with Polars integration:
- Implemented `StreamProcessor` class for memory-efficient processing
- Added chunk-by-chunk processing with minimal memory footprint
- Created streaming file reader for multi-gigabyte CSV files
- Added streaming API for custom chunk processing
- Added comprehensive error handling for stream processing failures
- Implemented graceful fallbacks when Polars is not available
- Added adaptive chunk sizing based on data volume
- Added example script (`example_stream_processing.py`) demonstrating stream processing

Improved
- Enhanced memory efficiency when processing multi-gigabyte datasets
- Added streaming progress bars for better user feedback
- Updated documentation with comprehensive stream processing examples
- Added tests for stream processing functionality

Added
- Simplified API with unified interface methods for common operations:
- Added `manage_acronyms(action, data, ...)` to replace multiple acronym methods
- Added `manage_patterns(action, data, ...)` to replace multiple pattern methods
- Added unified `process_dataframe(operation, ...)` to consolidate DataFrame methods
- Added configuration objects for improved parameter organization:
- `AnalysisConfig` for grouping analysis parameters
- `AnonymizationConfig` for grouping anonymization parameters
- Added stream processing for very large files with Polars integration:
- Implemented `StreamProcessor` class for memory-efficient processing
- Added chunk-by-chunk processing with minimal memory footprint
- Created streaming file reader for multi-gigabyte CSV files
- Added streaming API for custom chunk processing
- Added comprehensive error handling for stream processing failures
- Implemented graceful fallbacks when Polars is not available
- Added adaptive chunk sizing based on data volume
- Added example script (`example_stream_processing.py`) demonstrating stream processing
- Created new example script (`example_simplified_api.py`) demonstrating the simplified API
- Maintained backward compatibility with deprecated method support
- Added comprehensive docstrings for all new methods

Improved
- Reduced total public API method count while preserving all functionality
- Consolidated parameter handling for better code organization
- Enhanced code maintainability and reduced duplication
- Simplified pattern and acronym management workflows
- Optimized large file processing with adaptive chunk sizing
- Enhanced memory efficiency when processing multi-gigabyte datasets
- Added streaming progress bars for better user feedback

Breaking Changes
- Removed legacy analyzer factory functions (`create_au_analyzer`, `create_insurance_analyzer`, `create_au_insurance_analyzer`)
- Added cleaner `create_analyzer()` function as the main entry point
- Increased major version number to indicate production readiness
- Removed backward compatibility with 0.x versions
- Streamlined and simplified internal implementation
- Improved documentation and examples with clearer API usage

Changed
- Updated minimum Python version requirement to 3.10+
- Added support for Python 3.11 and 3.12
- Removed support for Python 3.8 and 3.9 due to dependency requirements (particularly NumPy 2.0+)
- Updated GitHub Actions workflows to test on Python 3.10-3.12 only

Fixed
- Fixed circular import issues in the insurance module
- Enhanced batch processing capabilities for better performance
- Addressed compatibility issues with newer NumPy (2.0+) requirements
- Fixed package build and CI/CD processes
- Improved documentation for Python version requirements