@lexical/headless

Overview

The @lexical/headless package provides a headless version of the Lexical editor that can run without a DOM environment, perfect for server-side rendering, testing, or Node.js applications.

Installation

npm install @lexical/headless

For DOM operations, the headless package includes happy-dom as a dependency.

Core Function

createHeadlessEditor

Creates a Lexical editor that runs without requiring a browser DOM.

function createHeadlessEditor(
  editorConfig?: CreateEditorArgs
): LexicalEditor

editorConfig

CreateEditorArgs

Optional editor configuration (same as createEditor)

Returns: LexicalEditor instance with DOM methods disabled Example:

import { createHeadlessEditor } from '@lexical/headless';
import { $createParagraphNode, $createTextNode, $getRoot } from 'lexical';

const editor = createHeadlessEditor({
  namespace: 'ServerEditor',
  nodes: [/* your nodes */],
  onError: (error) => console.error(error)
});

editor.update(() => {
  const root = $getRoot();
  const paragraph = $createParagraphNode();
  paragraph.append($createTextNode('Hello from server!'));
  root.append(paragraph);
});

Disabled Methods

The following methods are disabled in headless mode and will throw an error if called:

registerDecoratorListener

() => never

Decorator listeners require DOM rendering

registerRootListener

() => never

Root element listeners require DOM

registerMutationListener

() => never

Mutation listeners require DOM observation

getRootElement

() => never

No root element in headless mode

setRootElement

() => never

Cannot set root element in headless mode

getElementByKey

() => never

No DOM elements in headless mode

focus

() => never

Cannot focus without DOM

blur

() => never

Cannot blur without DOM

Supported Operations

Headless editors support all non-DOM operations:

State manipulation: editor.update(), editor.read(), editor.getEditorState()
Node operations: Creating, modifying, and querying nodes
Selection: Creating and manipulating selections programmatically
Commands: Dispatching and handling commands
Serialization: toJSON(), fromJSON(), exportJSON(), importJSON()
Transforms: Registering node transforms
Listeners: Update listeners, command listeners
Collaboration: Yjs integration works in headless mode

Use Cases

Server-Side Rendering (SSR)

import { createHeadlessEditor } from '@lexical/headless';
import { $generateHtmlFromNodes } from '@lexical/html';
import { JSDOM } from 'jsdom';

// Set up JSDOM for HTML generation
const jsdom = new JSDOM('<!DOCTYPE html>');
global.document = jsdom.window.document;
global.window = jsdom.window as any;

function renderEditorContent(jsonState: string): string {
  const editor = createHeadlessEditor({
    nodes: [/* your nodes */]
  });
  
  // Load state from JSON
  const state = editor.parseEditorState(jsonState);
  editor.setEditorState(state);
  
  // Generate HTML
  return state.read(() => {
    return $generateHtmlFromNodes(editor);
  });
}

const html = renderEditorContent(savedJsonState);
console.log(html);

Content Processing

import { createHeadlessEditor } from '@lexical/headless';
import { $convertFromMarkdownString, $convertToMarkdownString } from '@lexical/markdown';

function convertMarkdownToJson(markdown: string): string {
  const editor = createHeadlessEditor({
    nodes: [/* markdown-compatible nodes */]
  });
  
  editor.update(() => {
    $convertFromMarkdownString(markdown);
  });
  
  return JSON.stringify(editor.getEditorState().toJSON());
}

const jsonState = convertMarkdownToJson('# Hello\n\nThis is **markdown**.');

Testing

import { describe, it, expect } from 'vitest';
import { createHeadlessEditor } from '@lexical/headless';
import { $getRoot, $createParagraphNode, $createTextNode } from 'lexical';

describe('Editor Tests', () => {
  it('should create and modify content', () => {
    const editor = createHeadlessEditor();
    
    editor.update(() => {
      const root = $getRoot();
      const paragraph = $createParagraphNode();
      const text = $createTextNode('Test content');
      paragraph.append(text);
      root.append(paragraph);
    });
    
    editor.read(() => {
      const root = $getRoot();
      expect(root.getTextContent()).toBe('Test content');
    });
  });
  
  it('should handle transforms', () => {
    const editor = createHeadlessEditor({
      nodes: [CustomNode]
    });
    
    const transformFn = vi.fn();
    editor.registerNodeTransform(CustomNode, transformFn);
    
    editor.update(() => {
      $getRoot().append(new CustomNode());
    });
    
    expect(transformFn).toHaveBeenCalled();
  });
});

Batch Content Conversion

import { createHeadlessEditor } from '@lexical/headless';
import { $generateHtmlFromNodes } from '@lexical/html';

interface Document {
  id: string;
  lexicalState: string;
}

async function batchConvertToHtml(documents: Document[]): Promise<Map<string, string>> {
  const editor = createHeadlessEditor({
    nodes: [/* your nodes */]
  });
  
  const results = new Map<string, string>();
  
  for (const doc of documents) {
    const state = editor.parseEditorState(doc.lexicalState);
    editor.setEditorState(state);
    
    const html = state.read(() => $generateHtmlFromNodes(editor));
    results.set(doc.id, html);
  }
  
  return results;
}

Content Validation

import { createHeadlessEditor } from '@lexical/headless';
import { $getRoot } from 'lexical';

function validateEditorContent(jsonState: string): {
  isValid: boolean;
  errors: string[];
} {
  const errors: string[] = [];
  
  const editor = createHeadlessEditor({
    onError: (error) => {
      errors.push(error.message);
    },
    nodes: [/* your nodes */]
  });
  
  try {
    const state = editor.parseEditorState(jsonState);
    editor.setEditorState(state);
    
    state.read(() => {
      const root = $getRoot();
      
      // Custom validation logic
      if (root.getChildrenSize() === 0) {
        errors.push('Content is empty');
      }
      
      // Validate specific nodes
      const children = root.getAllTextNodes();
      children.forEach(node => {
        const text = node.getTextContent();
        if (text.includes('forbidden-word')) {
          errors.push('Content contains forbidden words');
        }
      });
    });
  } catch (error) {
    errors.push(`Parse error: ${error.message}`);
  }
  
  return {
    isValid: errors.length === 0,
    errors
  };
}

With Markdown

import { createHeadlessEditor } from '@lexical/headless';
import { 
  $convertFromMarkdownString, 
  $convertToMarkdownString, 
  TRANSFORMERS 
} from '@lexical/markdown';
import { HeadingNode, QuoteNode } from '@lexical/rich-text';
import { CodeNode } from '@lexical/code';
import { LinkNode } from '@lexical/link';
import { ListNode, ListItemNode } from '@lexical/list';

function processMarkdown(input: string): string {
  const editor = createHeadlessEditor({
    nodes: [
      HeadingNode,
      QuoteNode,
      CodeNode,
      LinkNode,
      ListNode,
      ListItemNode
    ]
  });
  
  // Convert markdown to Lexical
  editor.update(() => {
    $convertFromMarkdownString(input, TRANSFORMERS);
  });
  
  // Process/modify content
  editor.update(() => {
    const root = $getRoot();
    // ... make modifications ...
  });
  
  // Convert back to markdown
  return editor.getEditorState().read(() => {
    return $convertToMarkdownString(TRANSFORMERS);
  });
}

const processed = processMarkdown('# Title\n\nContent');

With Yjs Collaboration

import { createHeadlessEditor } from '@lexical/headless';
import { createBinding } from '@lexical/yjs';
import * as Y from 'yjs';

const doc = new Y.Doc();
const docMap = new Map([['main', doc]]);

const editor = createHeadlessEditor({
  nodes: [/* your nodes */]
});

// Collaboration works in headless mode!
const binding = createBinding(
  editor,
  provider,
  'main',
  doc,
  docMap
);

Important Notes

Headless editors cannot:

Render to the DOM
Handle user input events
Use decorators that require rendering
Access or manipulate DOM elements
Use plugins that depend on DOM APIs

For HTML generation in headless mode, use @lexical/html with a DOM implementation like jsdom or happy-dom.

Performance Benefits

Headless editors are ideal for:

Server-side processing: No browser overhead
Batch operations: Process multiple documents efficiently
Testing: Fast unit tests without DOM rendering
Build-time generation: Pre-render content during build
API endpoints: Convert/validate content in backend services

Comparison with Regular Editor

Feature	Regular Editor	Headless Editor
DOM Rendering	Yes	No
User Input	Yes	No
State Management	Yes	Yes
Commands	Yes	Yes
Transforms	Yes	Yes
Serialization	Yes	Yes
Collaboration	Yes	Yes
SSR	No	Yes
Testing Speed	Slower	Faster
Memory Usage	Higher	Lower

Core

React

Packages

@lexical/headless

Overview

Installation

Core Function

createHeadlessEditor

Disabled Methods

Supported Operations

Use Cases

Server-Side Rendering (SSR)

Content Processing

Testing

Batch Content Conversion

Content Validation

With Markdown

With Yjs Collaboration

Important Notes

Performance Benefits

Comparison with Regular Editor

Build docs developers (and LLMs) love

Core

React

Packages

​Overview

​Installation

​Core Function

​createHeadlessEditor

​Disabled Methods

​Supported Operations

​Use Cases

​Server-Side Rendering (SSR)

​Content Processing

​Testing

​Batch Content Conversion

​Content Validation

​With Markdown

​With Yjs Collaboration

​Important Notes

​Performance Benefits

​Comparison with Regular Editor

Build docs developers (and LLMs) love

Overview

Installation

Core Function

createHeadlessEditor

Disabled Methods

Supported Operations

Use Cases

Server-Side Rendering (SSR)

Content Processing

Testing

Batch Content Conversion

Content Validation

With Markdown

With Yjs Collaboration

Important Notes

Performance Benefits

Comparison with Regular Editor