letsbench

🏁 LetsBench

A simple CLI tool to run head-to-head function benchmarking across NPM packages. Perfect for comparing the performance of similar functions from different libraries.

[!TIP] A helpful tool to evaluate performance changes between the main package and the updated pkg-pr-new version.

Quick Start

Run LetsBench directly with npx:

npx letsbench

Features

• 🏆 Head-to-head comparison of functions from two NPM packages
• ⚡ Performance metrics including execution time and memory usage
• 🔍 Automatic function discovery from package exports
• 📊 Detailed results with system information and winner declaration
• 🎯 Multiple runs support for more accurate averages
• 🎨 Colorful CLI interface with visual feedback
• ⚡ Direct CLI syntax for quick comparisons

Usage

Interactive Mode

Simply run the command and follow the interactive prompts:

npx letsbench

Direct CLI Mode

Compare functions directly from the command line using natural syntax:

# Basic comparison
npx letsbench lodash cloneDeep '{"a":{"b":{"c":1}}}' vs ramda clone

# Different arguments for each function
npx letsbench lodash cloneDeep '{"a":{"b":{"c":1}}}' vs ramda clone '{"a":{"b":{"c":1}}}'

# Multiple runs for better accuracy
npx letsbench --runs 20 lodash cloneDeep '{"a":{"b":{"c":1}}}' vs ramda clone

CLI Syntax:

npx letsbench [options] <package1> <function1> <args1> vs <package2> <function2> [args2]

• If args2 is omitted, args1 will be used for both functions
• Arguments are parsed the same way as in interactive mode

Options

• --runs, -r: Number of runs per function (1-100, default: 1)

Example Sessions

Interactive Mode

➜ npx letsbench

  _         _         ____                  _
 | |    ___| |_ ___  | __ )  ___ _ __   ___| |__
 | |   / _ \ __/ __| |  _ \ / _ \ '_ \ / __| '_ \
 | |__|  __/ |_\__ \ | |_) |  __/ | | | (__| | | |
 |_____\___|\__|___/ |____/ \___|_| |_|\___|_| |_|

A simple CLI to run head-to-head function benchmarking across NPM packages

✔ First NPM package: demo-package1
✔ Second NPM package: demo-package2
✔ demo-package1 loaded
✔ demo-package2 loaded
✔ Choose function from demo-package1: pascalCase
✔ Choose function from demo-package2: casePascal
✔ Arguments for demo-package1.pascalCase: hello world
✔ Arguments for demo-package2.casePascal: ["hello world", {"normalize": true}]
✔ Benchmarks completed

🏆 BENCHMARK RESULTS
==================================================

💻 System Info:
Platform: darwin arm64
CPU: Apple M1
Memory: 8GB
Node: v23.10.0
Runs: 1

📊 Results:

1. demo-package1.pascalCase
   ⏱️  Time: 0.1453ms
   🧠 Memory: +6344 bytes
   ✅ Result: "HelloWorld"

2. demo-package2.casePascal
   ⏱️  Time: 0.4080ms
   🧠 Memory: +12936 bytes
   ✅ Result: "HelloWorld"

🚀 Winner: demo-package1.pascalCase
   2.81x faster than demo-package2.casePascal

Direct CLI Mode

➜ npx letsbench lodash map "[1,2,3]" vs ramda map

🏆 BENCHMARK RESULTS
==================================================

💻 System Info:
Platform: darwin arm64
CPU: Apple M1
Memory: 8GB
Node: v23.10.0
Runs: 1

📊 Results:

1. lodash.map
   ⏱️  Time: 0.0234ms
   🧠 Memory: +2456 bytes
   ✅ Result: [1,2,3]

2. ramda.map
   ⏱️  Time: 0.0891ms
   🧠 Memory: +4123 bytes
   ✅ Result: [1,2,3]

🚀 Winner: lodash.map
   3.81x faster than ramda.map

Function Arguments

LetsBench supports flexible argument parsing in both interactive and CLI modes:

Argument Examples

Input	Parsed As	Description
hello world	["hello world"]	Single string (auto-parsed)
[]	[]	No arguments
["hello world"]	["hello world"]	Single string (explicit)
["hello", {"normalize": true}]	["hello", {"normalize": true}]	String with options object
[42, 100]	[42, 100]	Two numbers
[[1,2,3]]	[[1,2,3]]	Array as argument

Argument Parsing Rules

• Empty input: Returns empty array []
• Valid JSON: Parses as JSON (arrays remain arrays, objects become single arguments)
• Invalid JSON: Treats as single string argument

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

License

MIT