v0.1.0 — Initial release

Initial release 🎉

Fast TOON (Token-Oriented Object Notation) encoder for Python, Rust, and CLI. Byte-identical to the TOON spec, 8× faster than toons, 2.7× faster than the official TypeScript SDK.

Performance (neptune 50-doc payload, 7480 bytes JSON → 4012 bytes TOON)

Encoder	Time	Relative
etoon (Python)	13.0 μs	baseline
@toon-format/toon (TS SDK, Node)	35.6 μs	2.73× slower
py-rtoon	85.9 μs	6.6× slower
toons	106.4 μs	8.2× slower

CLI via stdin pipe (Claude Code / Bash agents):

• etoon: 0.57 ms/call
• official toon CLI: 50.7 ms/call (89× slower due to stdin heuristics)

Architecture

Python dict → orjson.dumps → JSON bytes → sonic-rs (SIMD parse) → TOON emitter → String

Key optimizations:

• sonic-rs SIMD JSON parser (~7× faster than serde_json)
• orjson bridge — single PyO3 boundary crossing
• uniform-order table fast path — skips 300 key lookups on typical tabular payloads
• itoa specialized integer formatting
• 111/111 TOON spec fixtures passing (Apache-2.0 from toons)

Install

pip install etoon                          # Python (13μs in-process)
cargo add etoon --no-default-features      # Rust library
cargo install etoon --no-default-features  # Rust CLI from source

CLI binaries for Linux/macOS/Windows below ↓

Compatibility

Output is byte-identical to the toons Python package and the official toon-format/toon TypeScript SDK. Drop-in replacement for both on the Python side.

Limitations

• Integers > 2⁶³ coerced via f64 (arbitrary precision not supported)
• Custom indent hardcoded to 2 spaces (TOON spec default)
• Advanced spec features (delimiter/fold_keys) coming in v0.1.2

License: Apache-2.0 (test fixtures sourced from toons, also Apache-2.0)