GitHub - eranchetz/ChibiHashGo: Small, Fast 64 bit hash function - N-R-K/ChibiHash port for Go

ChibiHashGo

ChibiHashGo is a simple, fast 64-bit hash function implemented in Go. It is designed for performance and ease of use in applications where efficient hashing is required.

Features

• Minimal dependencies
• Optimized for performance
• Supports inputs of varying sizes (small keys to large datasets)
• Includes comprehensive benchmarks and GitHub Actions integration for CI

Installation

To install ChibiHashGo, use go get:

go get github.com/yourusername/ChibiHashGo

Usage

Import the package and use ChibiHash64 to compute hash values.

Example:

package main

import (
    "fmt"
    "github.com/eranchetz/ChibiHashGo"
)

func main() {
    key := []byte("example")
    seed := uint64(42)
    hash := chibihash.ChibiHash64(key, len(key), seed)
    fmt.Printf("Hash: %x\n", hash)
}

API Reference

func ChibiHash64(key []byte, length int, seed uint64) uint64

• key: The data to hash as a byte slice.
• length: The length of the data.
• seed: A 64-bit seed value for hashing.
• Returns: A 64-bit unsigned integer hash.

Benchmarks

The following benchmarks demonstrate the performance of ChibiHash64 for varying input sizes. These results were obtained using go test -bench=. -benchmem.

The function performs efficiently with no memory allocations (allocs/op = 0) across all test cases, ensuring minimal overhead.

CI/CD Integration

GitHub Actions is set up to automatically run tests and benchmarks for every push or pull request.

Workflow Configuration

The workflow file is located at .github/workflows/go.yml and includes:

• Installing dependencies (go mod tidy)
• Running tests (go test ./... -v)
• Running benchmarks (go test -bench=. -benchmem)

Testing

To test the module locally:

To run benchmarks:

go test -bench=. -benchmem

Contributing

Contributions are welcome! Fork the repository, make your changes, and submit a pull request. Ensure all tests pass and benchmarks are updated before submitting.

License

This project is licensed under the Unlicense.