docusaurus-lunr-search
Offline Search for Docusaurus v3.9.2+
Screenshot
Prerequisites
- Docusaurus >v3.9.2
- Node.js >= 20.X
How to Use?
- Install this package
yarn add docusaurus-lunr-searchor
npm i docusaurus-lunr-search --saveIf npm install fails to install with error unable to resolve dependency tree, run npm i --legacy-peer-deps
2. Sometimes npm fails to install lunr package, in that case install lunr package manually
npm i lunr --save- Add the docusaurus-lunr-search plugin to your
docusaurus.config.js
module.exports = {
// ...
plugins: [require.resolve('docusaurus-lunr-search')],
}- Then build your Docusaurus project
yarn buildor
npm run build- Serve your application
yarn serveor
npm run serve Note: The Docusaurus search index is only generated from a full project build. The local development server does not provide search functionality.
Using an option (eg. languages) in the plugin
module.exports = {
// ...
plugins: [[ require.resolve('docusaurus-lunr-search'), {
languages: ['en', 'de'] // language codes
}]],
}Supports all the language listed here https://github.com/MihaiValentin/lunr-languages
Options available
| Option | Type | Default | Description |
|---|---|---|---|
languages |
Array |
['en'] |
Language codes to use for stemming, Supports all the language listed here https://github.com/MihaiValentin/lunr-languages |
indexBaseUrl |
Boolean |
false |
Base url will not indexed by default, if you want to index the base url set this option to true |
excludeRoutes |
Array |
[] |
Exclude certain routes from the search |
includeRoutes |
Array |
[] |
Include only specific routes for search |
stopWords |
Array |
[] |
Add stop words(words that are exclude from search result) to the search index |
excludeTags |
Array |
[] |
Exclude certain tags from the search |
highlightResult |
Boolean |
false |
Enable it to highlight the searched word in the result page. Used mark.js for highlighting. You can customize the highlight color using css mark { background-color: red !important; color: green !important } |
disableVersioning |
Boolean |
false |
Docs versions are displayed by default. If you want to hide it, set this plugin option to true |
assetUrl |
string |
\ |
Url from which the generated search doc files to be loaded, check issue #122 |
maxHits |
string |
5 |
Maximum number of hits shown |
fields |
object |
{} |
Lunr field definitions, allows "boosting" priority for different sources of keywords (e.g. title, content, keywords) |
maxThreads |
number |
os.cpus().length |
Maximum number of worker threads for indexing. Useful in containerized environments where os.cpus() returns the host's CPU count instead of the container's limit |
Options to configure Lunr fields
The fields config property is passed into Lunr directly as field attributes, and can be used to configure the relative priority of different field types (e.g. title, content, keywords).
docusaurus-lunr-search sets the default value for fields to:
{
title: { boost: 200 },
content: { boost: 2 },
keywords: { boost: 100 }
}Indexing non-direct children headings of .markdown
By default, this library will only search for headings that are
direct children of the .markdown element.
If you would like to render content inside the .markdown element on
a swizzled DocItem component, and want this library to index the
headings inside those custom elements even if they are not direct
children of the .markdown element, then add the attribute
data-search-children to a parent element of the headings you want to
index.
The data-search-children attribute will cause this library to look
for all headings inside that element, including both direct and
indirect children (E.g. 'grandchildren' nodes).
Check this issue #115 for more details.
Credits
Thanks to algolia/docsearch.js, I modified it to create this search component.
And thanks cmfcmf, I used the code from his library docusaurus-search-local for multi-language support.
Changelog
Checkout the releases page for changelog.