agirton/react-select-fast-filter-options

react-select filterOptions function optimized to quickly filter large options lists

react-select-fast-filter-options

Fast filterOptions function for react-select; optimized to quickly filter huge options lists.

Getting started

Installation

The easiest way to install is using NPM:

npm install react-select-fast-filter-options --save

ES6, CommonJS, and UMD builds are available with each distribution. Use unpkg to access the UMD build:

<script src="https://unpkg.com/react-select-fast-filter-options/dist/umd/react-select-fast-filter-options.js"></script>

Here's how to fast filter with react-select or react-virtualized-select:

// Import the Select component from either react-select or react-virtualized-select
import Select from 'react-virtualized-select' // or from 'react-select'

// Import the fast-filter library
import createFilterOptions from 'react-select-fast-filter-options'

// Create a search index optimized to quickly filter options.
// The search index will need to be recreated if your options array changes.
// This index is powered by js-search: https://github.com/bvaughn/js-search
const filterOptions = createFilterOptions({ options })

// Render your Select, complete with the fast-filter index
function render ({ options }) {
  return (
    <Select
      filterOptions={filterOptions}
      options={options}
      {...otherSelectProps}
    />
  )
}

Configuration Options

By default, createFilterOptions returns a filter function configured to match all substrings, in a case-insensitive way, and return results in their original order. However it supports all of the underlying js-search configuration options.

The following table shows all supported parameters and their default values:

Property	Type	Default	Description
indexes	Array<String>		Optional array of attributes to build search index from; defaults to the labelKey attribute.
indexStrategy	IIndexStrategy	AllSubstringsIndexStrategy	See js-search docs
labelKey	string	"label"	Option key containing the display text
options	array	[]	Array of options objects
sanitizer	ISanitizer	LowerCaseSanitizer	See js-search docs
searchIndex	ISearchIndex	UnorderedSearchIndex	See js-search docs
tokenizer	ITokenizer	SimpleTokenizer	See js-search docs
valueKey	string	"value"	Option key containing the value

Advanced Configuration

The default filter configuration mimics react-search behavior. But you can also customize search. For example:

import {
  CaseSensitiveSanitizer,
  ExactWordIndexStrategy,
  Search,
  SimpleTokenizer,
  StemmingTokenizer,
  TfIdfSearchIndex
} from 'js-search'
import { stemmer } from 'porter-stemmer'
import createFilterOptions from 'react-select-fast-filter-options'

// Default index strategy is built for all substrings.
// In other word "c", "ca", "cat", "a", "at", and "t" all match "cat".
// Override to only allow exact-word matches like so:
const indexStrategy = new ExactWordIndexStrategy()

// Default sanitizer is case-insensitive
// Searches for "foo" will match "Foo".
// Override to be case-sensitive like so:
const sanitizer = new CaseSensitiveSanitizer()

// By default, search results are returned in the order they wre indexed.
// This means that your filtered options will match their unfiltered order.
// More advanced results orderings are possbile.
// For example TF-IDF ranking is an option.
// Learn more at https://github.com/bvaughn/js-search#tf-idf-ranking
const searchIndex = new TfIdfSearchIndex()

// Default tokenizer just splits search text on spaces.
// In other words "the boy" becomes 2 search tokens, "the" and "boy".
// The StemmingTokenizer can be used for fuzzier matching.
// For example, "searching" will match  "search", "searching", and "searched".
// Learn more at https://github.com/bvaughn/js-search#stemming
const tokenizer = new StemmingTokenizer(stemmer, new SimpleTokenizer())

const filterOptions = createFilterOptions({
  indexStrategy,
  options,
  sanitizer,
  searchIndex,
  tokenizer
})

In addition to the stemming tokenizer, other tokenizers are available as well, including StopWordsTokenizer which removes common words like "a", "and", and "the". For more information on available configuration options, see js-search documentation.