Benchmark: js search benchmark

HTML Preparation code:

<script src="https://unpkg.com/lunr/lunr.js"></script>
<script type="text/javascript" src="https://rawgit.com/bvaughn/js-search/1.2.0/dist/js-search.js"></script>
<script type="text/javascript" src="https://cdn.jsdelivr.net/npm/minisearch@2.2.2/dist/umd/index.min.js"></script>
<script type="text/javascript" src="https://cdn.jsdelivr.net/npm/fuse.js/dist/fuse.js"></script>
<script>

        Benchmark.prototype.setup = function() {

          
  var books = [];
  var xmlhttp = new XMLHttpRequest();
  xmlhttp.onreadystatechange = function() {
    if (xmlhttp.readyState == 4 && xmlhttp.status == 200) {
      var json = JSON.parse(xmlhttp.responseText);
  
      books = json.books;
    }
  }
  xmlhttp.open('GET', 'http://bvaughn.github.io/js-search/books.json', true);
  xmlhttp.send();

        
  };


      
</script>

Tests:

Lunr
var idx = lunr(function () { this.ref('isbn'); this.field('title'); this.field('author'); books.forEach(function (doc) { this.add(doc) }, this); })
js search
var search = new JsSearch.Search('isbn'); search.searchIndex = new JsSearch.TfIdfSearchIndex('isbn'); search.addIndex('title'); search.addIndex('author'); search.addDocuments(books);
minisearch
var miniSearch = new MiniSearch({ fields: ['isbn', 'title', 'author'], // fields to index for full-text search }); miniSearch.addAll(books)
fuse.js
const options = { includeScore: true, keys: ['isbn', 'title', 'author'] } const fuse = new Fuse(books, options)

Rendered benchmark preparation results:

Suite status: <idle, ready to run>

Previous results

Test case name	Result
Lunr
js search
minisearch
fuse.js

Fastest: N/A

Slowest: N/A

Latest run results:

Run details: (Test run date: 7 days ago)

User agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/136.0.0.0 Safari/537.36

Browser/OS: Chrome 136 on Mac OS X 10.15.7

View result in a separate tab

Test name	Executions per second
Lunr	3739879.5 Ops/sec
js search	30802430.0 Ops/sec
minisearch	1817995.6 Ops/sec
fuse.js	730920.1 Ops/sec

Autogenerated LLM Summary (model llama3.2:3b, generated 7 months ago):

LLMs can make mistakes. Check important info.

Let's dive into the world of JavaScript microbenchmarks!

The provided JSON represents a benchmarking test, where users can create and run JavaScript microbenchmarks to compare different search libraries and their performance.

**Benchmark Definition**

The benchmark definition is a script that sets up the environment for the test. In this case, it includes four different search libraries: Lunr, js-search, minisearch, and fuse.js.

Each library has its own unique approach to indexing and searching data. Here's a brief overview of each:

1. **Lunr**: Lunr is a popular search library that uses an inverted index to store the data. It allows you to specify fields to index for full-text search. In this benchmark, Lunr is used with the ISBN field.
2. **js-search**: js-search is another search library that provides a simple and intuitive API for indexing and searching data. It uses a combination of TF-IDF and term frequency-inverse document frequency (TF-IDF) algorithms to calculate scores.
3. **minisearch**: minisearch is a lightweight search library that allows you to index and search data using a minimal number of lines of code. It provides an efficient way to store and retrieve data, making it suitable for smaller datasets.
4. **fuse.js**: fuse.js is a small and fast JavaScript search library that uses a simple and efficient algorithm to match keywords in the indexed data.

**Options Comparison**

Each library has its own set of options that can be used to customize the indexing and searching process. Here are some key options for each library:

1. **Lunr**: `includeScore` (boolean) - includes scores in search results, `keys` (array) - specifies fields to index for full-text search.
2. **js-search**: `includeScore` (boolean), `keys` (array) - similar to Lunr options, but with additional options like `threshold` and `stemmer`.
3. **minisearch**: `fields` (array) - specifies fields to index for full-text search, `indexSize` (number) - limits the size of the index.
4. **fuse.js**: `keys` (array), `limit` (number) - specifies fields to index for full-text search and sets a limit on the number of results.

**Pros and Cons**

Here's a brief summary of the pros and cons of each library:

1. **Lunr**:
	* Pros: Fast, efficient indexing and searching, supports multiple languages.
	* Cons: Can be complex to set up, limited support for non-standard fields.
2. **js-search**:
	* Pros: Simple and intuitive API, supports multiple search algorithms.
	* Cons: Can be slower than other libraries, requires more memory for indexing.
3. **minisearch**:
	* Pros: Lightweight, easy to use, fast searching.
	* Cons: Limited support for complex queries, may not work well with very large datasets.
4. **fuse.js**:
	* Pros: Fast and efficient, simple API, small footprint.
	* Cons: Limited support for advanced search features, may not work well with non-standard fields.

**Library Considerations**

Some libraries require additional setup or configuration beyond the provided benchmark definition:

1. **Lunr**: Requires a separate configuration file to define field mappings and indexing settings.
2. **js-search**: Requires a separate configuration file to define search algorithms and options.
3. **minisearch**: Requires a separate configuration file to limit the size of the index.
4. **fuse.js**: Does not require any additional setup or configuration.

**Other Alternatives**

If you're looking for alternative search libraries, here are some popular ones:

1. **Sphinx Search**: A powerful and feature-rich search library with a Python API.
2. **Elasticsearch**: A popular open-source search engine that provides robust querying capabilities.
3. **Solr**: A popular search library that provides efficient indexing and searching capabilities.

In conclusion, the choice of search library depends on your specific requirements, such as performance, ease of use, and feature set. This benchmark provides a good starting point for comparing different libraries and their options to find the best fit for your project.

Related benchmarks:

js search benchmark (version: 1)

Comparing performance of: Lunr vs js search vs minisearch vs fuse.js

Created: one year ago by: Registered User

Jump to the latest result

Lunr

js search

minisearch

fuse.js

Suite status: <idle, ready to run>

Experimental features:

Fastest: N/A

Slowest: N/A

Autogenerated LLM Summary (model llama3.2:3b, generated 7 months ago):