eleventy-plugin-search

eleventy-plugin-search

Early Development Notice: This plugin is under active development. The API may change before reaching v1.0.0. Please report issues and feedback on GitHub.

HTML-first Eleventy search plugin with Fuse.js and Cheerio. Processes final rendered HTML files after templates for accurate search indexing.

Features

• HTML-first processing: Works with final rendered HTML, not source files
• Accurate content extraction: Uses Cheerio for fast, reliable HTML parsing
• Automatic heading IDs: Generates URL-safe anchor IDs for headings without them
• Fuse.js integration: Creates search indexes optimized for client-side fuzzy search
• Configurable exclusions: Remove nav, header, footer, and other elements from indexing
• Debug support: Enable detailed logging with DEBUG=Eleventy:search

Installation

npm install eleventy-plugin-search

Usage

Add the plugin to your Eleventy configuration:

// eleventy.config.js
import searchPlugin from 'eleventy-plugin-search';

export default function (eleventyConfig) {
  eleventyConfig.addPlugin(searchPlugin, {
    // Options (all optional)
    indexPath: 'search-index.json',
    excludeSelectors: ['nav', 'header', 'footer']
  });
}

The plugin runs after Eleventy builds your site and creates a search-index.json file in your output directory.

Options

Option	Type	Default	Description
pattern	string	'**/*.html'	Glob pattern for HTML files to process
ignore	string | string[]	['**/search-index.json']	Files to exclude from indexing
indexPath	string	'search-index.json'	Output path for the search index
excludeSelectors	string[]	['nav', 'header', 'footer']	CSS selectors to remove before extraction
maxContentLength	number	10000	Maximum characters of content to index
fuseOptions	object	See below	Fuse.js configuration

Default Fuse.js Options

{
  keys: [
    { name: 'title', weight: 10 },
    { name: 'content', weight: 5 },
    { name: 'excerpt', weight: 3 }
  ],
  threshold: 0.3,
  includeScore: true,
  includeMatches: true,
  minMatchCharLength: 3
}

Search Index Structure

The generated search-index.json contains:

{
  "version": "1.0.0",
  "generator": "eleventy-plugin-search",
  "generated": "2025-01-22T...",
  "totalEntries": 10,
  "config": {
    "fuseOptions": { /* Fuse.js config for client reconstruction */ }
  },
  "stats": {
    "totalEntries": 10,
    "entriesByType": { "page": 10 },
    "averageContentLength": 2048,
    "totalContentLength": 20480
  },
  "entries": [
    {
      "id": "page:/about",
      "type": "page",
      "url": "/about",
      "title": "About Us",
      "content": "Full page text content...",
      "excerpt": "First 250 characters...",
      "headings": [
        { "level": "h2", "id": "our-mission", "title": "Our Mission" },
        { "level": "h2", "id": "our-team", "title": "Our Team" }
      ],
      "wordCount": 523
    }
  ]
}

Client-Side Search Example

import Fuse from 'fuse.js';

// Fetch the search index
const response = await fetch('/search-index.json');
const searchIndex = await response.json();

// Initialize Fuse.js with the index config
const fuse = new Fuse(searchIndex.entries, searchIndex.config.fuseOptions);

// Perform a search
const results = fuse.search('your search query');

// Results include score and matches
results.forEach((result) => {
  console.log(result.item.title, result.score);

  // Use headings for scroll-to functionality
  result.item.headings.forEach((heading) => {
    console.log(`  ${heading.level}: ${heading.title} (#${heading.id})`);
  });
});

URL Generation

The plugin generates clean URLs from file paths:

• index.html → /
• about/index.html → /about
• blog/first-post.html → /blog/first-post

Heading ID Generation

For headings without id attributes, the plugin generates URL-safe IDs:

• "Hello World" → hello-world
• "Chapter 1: Introduction" → chapter-1-introduction
• Duplicate headings get numeric suffixes: section, section-1, section-2

Debugging

Enable debug output to see what the plugin is doing:

DEBUG=Eleventy:search npx @11ty/eleventy

Code Quality

• ESLint: Enforces code quality rules including prefer-const, no-var, strict equality, and complexity limits
• Prettier: Ensures consistent code formatting (single quotes, no trailing commas, 100 char line width)

Run npm run prerelease before committing to ensure your code passes all checks.

Test Coverage

This plugin is tested using mocha with c8 for code coverage. Current coverage: 97%.

Requirements

• Node.js >= 20.0.0
• Eleventy >= 2.0.0

License

MIT

Development Transparency

Portions of this project were developed with the assistance of AI tools including Claude and Claude Code. These tools were used to:

• Generate or refactor code
• Assist with documentation
• Troubleshoot bugs and explore alternative approaches

All AI-assisted code has been reviewed and tested to ensure it meets project standards. See the included CLAUDE.md file for more details.