v0.1.0

Changelog

[0.1.0]

Added

• Initial release of ibarrow
• High-performance ODBC to Arrow data conversion
• Two-level API (high-level wrappers + low-level functions)
• Zero-copy Arrow C Data Interface support
• Native type preservation (INT/DECIMAL/FLOAT → Arrow natives)
• Streaming pipelining for constant memory usage
• Support for Polars and Pandas DataFrames
• Comprehensive error handling with specific exception types
• Batch size configuration for performance tuning
• Connection timeout and query timeout support
• Transaction isolation level configuration
• Text and binary field size limits
• Read-only connection support
• Initial release
• Core ODBC to Arrow conversion functionality
• Python bindings with PyO3
• Rust backend for maximum performance

Performance

• Zero-copy data transfer via Arrow C Data Interface
• Streaming processing for large datasets (constant memory usage)
• Native type preservation avoiding string conversions
• Batch processing to avoid row-by-row fetching
• Optimized compilation with LTO and target-cpu=native

Security

• Safe PyCapsule management with Arc shared ownership
• No dangling pointer risks
• Automatic memory cleanup

Full Changelog: https://github.com/thomazyujibaba/ibarrow/commits/v0.1.0