Search Engine Mechanics

Verdocs includes a high-performance, zero-config search engine that operates entirely on the client side. This ensures that search functionality is instantaneous and works even in offline or static hosting environments.

How it Works

When you run verdocs generate or while the preview server is running, Verdocs:

1. Scans Content: Iterates through all Markdown files within each version folder.
2. Extracts Metadata: Identifies page titles (from H1 tags or file names) and indexes all text content.
3. Generates JSON Index: A highly optimized .json search index is created for each version and stored in out/search-index/.
4. Client-Side Resolution: The browser downloads only the index for the version currently being viewed, making search results available without further network requests.

Usage

You can trigger the search interface in two ways:

1. Clicking: Click the "Search..." bar in the top navigation area.
2. Keyboard Shortcut: Press Cmd + K (macOS) or Ctrl + K (Linux/Windows).

Search Results

Results are ranked based on matches in titles and body content. Each result provides a snippet of the matching line with the query terms highlighted, allowing for quick context assessment.

Optimization Tips

To make your documentation more searchable:

• Use Clear H1 Headings: Verdocs uses the first H1 heading in each file as the primary title in search results.
• Structured Hierarchy: Clear heading structures (H2, H3, etc.) help index the document effectively.
• Descriptive Content: Since Verdocs indexes the full text, ensure your documentation contains relevant keywords for better search discovery.