Hi everyone,
I’m building zarrow, a native Zig implementation of Apache Arrow:
It currently supports Arrow memory layout basics, array builders, validation, zero-copy slicing, IPC stream read/write, and Arrow C Data / C Stream ABI import/export.
The project is still experimental, but actively developed. My current focus is correctness, API design, and interop.
I’d love feedback on:
- API design
- allocator / ownership patterns
- C Data Interface ergonomics
- useful examples or demos
- what Zig users would expect from an Arrow implementation
The goal is to make Zig a serious option for Arrow-based tooling and data interchange experiments.
Feedback welcome!