@chakra-docs/search-pagefind

Pagefind document record helpers for Chakra Docs search manifests.

Created by Ryan Hefner and Commune Software.

This package maps generated DocsSearchRecord data (from a Chakra Docs manifest's search array) into Pagefind-friendly document records with url, title, content, and meta fields.

Install

npm install @chakra-docs/search-pagefind pagefind

Optional peer dependency: pagefind (>=1 <2) — only needed if you index the records with Pagefind's Node API.

Usage

import { createPagefindDocumentRecords } from '@chakra-docs/search-pagefind';
import * as pagefind from 'pagefind';
import { docsManifest } from './.chakra-docs/generated';

const records = createPagefindDocumentRecords(docsManifest.search);

const { index } = await pagefind.createIndex();

for (const record of records) {
  await index.addCustomRecord({
    url: record.url,
    content: record.content,
    language: 'en',
    meta: { title: record.title, ...record.meta },
  });
}

await index.writeFiles({ outputPath: 'public/pagefind' });

Page-level records map to the page route and heading-level records map to route#heading-id, so results can deep-link into sections. Canonically encoded route segments are preserved as-is in Pagefind URLs. Tags and aliases are appended to the indexed content so they participate in matching. Each record's meta carries the available document identity, display, source, tag, alias, and search-priority fields as strings (empty when absent).

API

• createPagefindDocumentRecords(records) — map DocsSearchRecord[] to PagefindDocumentRecord[] (url, title, content, meta).
• PagefindDocumentRecord — result type.

Help and contributing

See the project README, open an issue, or read the contribution guidelines. Report vulnerabilities privately through the security policy.

License

MIT