@flicknote/defuddle

de·​fud·dle /diˈfʌdl/ transitive verb to remove unnecessary elements from a web page, and make it easily readable.

A fork of kepano/defuddle optimized for linkedom and Cloudflare Workers.

Changes from upstream

• Always outputs Markdown (optimized for LLM consumption)
• Removed browser-specific APIs (getComputedStyle, styleSheets, getBoundingClientRect)
• Removed Node.js/JSDOM bundle - use linkedom instead
• Replaced webpack with esbuild for faster builds
• ESM-only output for modern runtimes

See LINKEDOM_COMPAT.md for details on removed features and their impact.

Installation

npm install @flicknote/defuddle linkedom turndown

Peer Dependencies

| Package | Required | Purpose | |---------|----------|---------| | linkedom | Yes | DOM implementation for Defuddle.parse() | | turndown | Yes | HTML to Markdown conversion |

Usage

Cloudflare Workers / Edge Runtimes (Recommended)

Use the static Defuddle.parse() method for the simplest API:

import Defuddle from '@flicknote/defuddle';

export default {
  async fetch(request: Request): Promise<Response> {
    const url = new URL(request.url).searchParams.get('url');
    if (!url) {
      return new Response('Missing url parameter', { status: 400 });
    }

    const response = await fetch(url);
    const html = await response.text();

    // Simple! No need to manually parse HTML
    const result = Defuddle.parse(html, { removeImages: true, url });

    return Response.json(result);
  }
};

Or use the standalone parse function:

import { parse } from '@flicknote/defuddle';

const result = parse(html, { removeImages: true, url });

With Custom Document (JSDOM, Browser, etc.)

If you have your own Document instance:

import Defuddle from '@flicknote/defuddle';

const result = new Defuddle(document, { url }).parse();

Browser

import Defuddle from '@flicknote/defuddle';

// Browser has native document
const result = new Defuddle(document).parse();
console.log(result.content);
console.log(result.title);

Response

| Property | Type | Description | |----------|------|-------------| | content | string | Extracted content as Markdown | | title | string | Title of the article | | author | string | Author of the article | | description | string | Description or summary | | domain | string | Domain name | | favicon | string | URL of the favicon | | image | string | URL of the main image | | published | string | Publication date | | site | string | Name of the website | | wordCount | number | Word count | | parseTime | number | Parse time in milliseconds | | schemaOrgData | object | Schema.org data | | metaTags | object | Meta tags |

Options

| Option | Type | Default | Description | |--------|------|---------|-------------| | url | string | | URL of the page being parsed | | debug | boolean | false | Enable debug logging | | removeImages | boolean | false | Remove all images | | removeExactSelectors | boolean | true | Remove elements matching exact selectors | | removePartialSelectors | boolean | true | Remove elements matching partial selectors |

Bundles

Two bundles are available:

• Core (@flicknote/defuddle): Main bundle, no math library dependencies (~85KB)
• Full (@flicknote/defuddle/full): Includes math libraries for MathML/LaTeX conversion (~461KB)

License

MIT