Kubernetes & Infra-as-Code Tooling
Parse multi-document YAML manifests concurrently across Rayon thread pools with zero heap allocations on hot scanning loops.
Seamless 1-line migration from archived serde_yaml to 100% safe Rust, 100% test-suite compliance, zero unsafe code, and native MCP AI support.
// Cargo.toml: noyalib = { version = "0.0.17", features = ["simd", "rayon"] }
use noyalib::{parallel, Value, Result};
fn main() -> Result<()> {
let yaml_data = r#"
name: noyalib Engine
version: 0.0.17
features: [SIMD, Rayon, SafeRust, MCP, LSP]
performance: 520MB/s
"#;
// Parallel Rayon multi-document deserialization
let parsed: Value = parallel::from_str(yaml_data)?;
println!("Parsed in parallel: {}", parsed["name"]);
Ok(())
}Architected for infrastructure platforms, AI agent runtimes, and high-performance Rust microservices.
Parse multi-document YAML manifests concurrently across Rayon thread pools with zero heap allocations on hot scanning loops.
Expose native MCP tools (`noyalib-mcp`) allowing Claude 3.5 and GPT-4o to parse, validate, and edit structured YAML configs losslessly.
Compile to sub-50 KB gzipped WASM (`noyalib-wasm`) with zero native C/C++ dependencies for browser applications and Cloudflare Workers.
Power real-time syntax recovery, schema hover documentation, and diagnostics in VS Code with `noyalib-lsp` (`parse_lenient`).
noyalib combines SWAR vector scanners, zero-copy byte dispatch, and multi-thread Rayon document pre-scanning to outpace both maintained 2026 Rust crates and legacy parsers. All benchmarks are open source and reproducible.
Benchmarked using Criterion.rs v0.5 across 10,000 iterations (95% Confidence Interval Β±0.8%). Hardware: Apple M2 Max & AMD EPYC 9654. Dataset: Official 387-case YAML Test Suite corpus and 100 MB multi-document enterprise telemetry payload.
cargo bench --bench throughputSurgical SWAR vector scanners, zero-copy byte dispatch, and 100% safe Rust memory invariants.
Multi-byte SWAR bitwise scanning finds structural delimiters in a single CPU pass with zero unsafe transmutes on stable Rust. Scans up to 8 bytes per CPU instruction loop.
Enforces #![forbid(unsafe_code)] across the entire codebase. Eliminates undefined behavior, null pointer dereferences, and buffer overflow vulnerability vectors on untrusted inputs.
Native MCP AI server tool allowing Claude 3.5 Sonnet, GPT-4o, and AI agents to query, parse, and validate data streams.
Enforces cargo audit, cargo deny, and cargo vet on every push. Zero unmaintained or archived dependencies.
Pre-scans multi-document boundaries and deserializes documents concurrently across all CPU cores.
Fault-tolerant parse_lenient engine powers live diagnostics, schema hover docs, and completion in VS Code.
noyalib is designed as a synchronized suite of satellite crates serving every execution context.
use noyalib::{parallel, Value};
let docs = parallel::values("--- doc1: val --- doc2: val")?;
assert_eq!(docs.len(), 2);Convert YAML, benchmark WASM parsing throughput live, and explore JSON Schema 2020-12 real-time diagnostics running on the noyalib-wasm WebAssembly engine.
{
"name": "noyalib",
"version": "0.0.17",
"architecture": [
"SIMD",
"Rayon",
"SafeRust"
],
"performance": {
"speed": "520MB/s",
"safe": true
}
}Choose your preferred installation method below.
Add to Cargo project with SIMD and Rayon acceleration.
cargo add noyalib --features simd,rayon
Install command-line parser & validator binary.
cargo install noya-cli
Install NPM package for Web & Cloudflare Workers.
npm install noyalib-wasm
Run native MCP server for AI tools & LLMs.
cargo install noyalib-mcp
noyalib is an open-source, high-performance data & YAML engine written in 100% safe Rust featuring zero-copy SIMD parsing, Rayon parallel streaming, WebAssembly runtime, Model Context Protocol (MCP) AI server, and LSP support. Validated against the official YAML Test Suite at 100% strict compliance (387/387 cases passing).