fast-markov

A high-performance Markov string chain for Node.js written in Rust.

Installation

npm install fast-markov
# or
yarn add fast-markov
# or
pnpm add fast-markov

Usage

Basic Example

const { MarkovChain } = require("fast-markov");

// Create a new chain
const chain = new MarkovChain();

// Add some training text
chain.addText("the quick brown fox jumps over the lazy dog");
chain.addText("the lazy cat sleeps all day long");

// Generate new text (10 words)
console.log(chain.generate(null, 10));

With Custom Start Word

const { MarkovChain } = require("fast-markov");

const chain = new MarkovChain();
chain.addText("the quick brown fox jumps over the lazy dog");

// Generate text starting with 'the'
console.log(chain.generate("the", 5));

Saving and Loading Chain State

const { MarkovChain } = require("fast-markov");

// Create and train chain
const chain = new MarkovChain();
chain.addText("some training text here");

// Export chain state
const exported = chain.exportCorpus();

// Create new chain and import state
const newChain = new MarkovChain();
newChain.importCorpus(exported);

With Custom Initial Capacity

const { MarkovChain } = require("fast-markov");

// Create chain with larger initial capacity for better performance
// when training with large datasets
const chain = new MarkovChain({ initialCapacity: 10000 });

API Reference

new MarkovChain(options?: MarkovChainOptions)

Creates a new Markov chain instance.

• options.initialCapacity?: number - Initial capacity for the internal hash map (default: 1000)

addText(text: string): void

Adds training text to the chain.

generate(startWord: string | null, length: number): string

Generates new text based on the trained patterns.

• startWord - Optional word to start generation with
• length - Number of words to generate
• Returns generated text string

getWordCount(): number

Returns the total number of words in the chain.

exportCorpus(): string

Exports the chain state as a JSON string.

importCorpus(json: string): void

Imports a previously exported chain state.

Supported Platforms

Pre-built binaries are available for the following platforms:

• Windows (x64, arm64)
• macOS (x64, arm64)
• Linux (x64, arm64)

License

MIT

Contributing

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