This JavaScript library adds a fast and fully-featured auto-completion menu to your search box displaying results "as you type". It can easily be combined with Algolia's realtime search engine. The library is available as a jQuery plugin, an Angular.js directive or a standalone library.
- Displays suggestions to end-users as they type
- Shows top suggestion as a hint (i.e. background text)
- Supports custom templates to allow for UI flexibility
- Works well with RTL languages and input method editors
- Triggers custom events
The autocomplete.js
library must be included after jQuery, Zepto or Angular.js (with jQuery).
We recommend including it from a CDN:
<script src="https://cdn.jsdelivr.net/autocomplete.js/0/autocomplete.min.js"></script>
<!-- OR -->
<script src="https://cdn.jsdelivr.net/autocomplete.js/0/autocomplete.jquery.min.js"></script>
<!-- OR -->
<script src="https://cdn.jsdelivr.net/autocomplete.js/0/autocomplete.angular.min.js"></script>
<script src="https://cdnjs.cloudflare.com/ajax/libs/autocomplete.js/<VERSION>/autocomplete.min.js"></script>
<!-- OR -->
<script src="https://cdnjs.cloudflare.com/ajax/libs/autocomplete.js/<VERSION>/autocomplete.jquery.min.js"></script>
<!-- OR -->
<script src="https://cdnjs.cloudflare.com/ajax/libs/autocomplete.js/<VERSION>/autocomplete.angular.min.js"></script>
bower install algolia-autocomplete.js -S
Or you can fetch the source:
You can find the built version in dist/.
You can require it and use Browserify:
var autocomplete = require('autocomplete.js');
To turn any HTML <input />
into a simple and fast as-you-type auto-completion menu following one of the 2 next sections:
- Include
autocomplete.min.js
- Initialize the auto-completion menu calling the
autocomplete
function
<input type="text" id="search-input" />
<!-- [ ... ] -->
<script src="https://cdn.jsdelivr.net/algoliasearch/3/algoliasearch.min.js"></script>
<script src="https://cdn.jsdelivr.net/autocomplete.js/0/autocomplete.min.js"></script>
<script>
var client = algoliasearch('YourApplicationID', 'YourSearchOnlyAPIKey')
var index = client.initIndex('YourIndex');
autocomplete('#search-input', { hint: false }, [
{
source: autocomplete.sources.hits(index, { hitsPerPage: 5 }),
displayKey: 'my_attribute',
templates: {
suggestion: function(suggestion) {
return suggestion._highlightResult.my_attribute.value;
}
}
}
]).on('autocomplete:selected', function(event, suggestion, dataset) {
console.log(suggestion, dataset);
});
</script>
- Include
autocomplete.jquery.min.js
after includingjQuery
- Initialize the auto-completion menu calling the
autocomplete
jQuery plugin
<input type="text" id="search-input" />
<!-- [ ... ] -->
<script src="https://cdn.jsdelivr.net/algoliasearch/3/algoliasearch.min.js"></script>
<script src="https://cdn.jsdelivr.net/autocomplete.js/0/autocomplete.jquery.min.js"></script>
<script>
var client = algoliasearch('YourApplicationID', 'YourSearchOnlyAPIKey')
var index = client.initIndex('YourIndex');
$('#search-input').autocomplete({ hint: false }, [
{
source: $.fn.autocomplete.sources.hits(index, { hitsPerPage: 5 }),
displayKey: 'my_attribute',
templates: {
suggestion: function(suggestion) {
return suggestion._highlightResult.my_attribute.value;
}
}
}
]).on('autocomplete:selected', function(event, suggestion, dataset) {
console.log(suggestion, dataset);
});
</script>
- Include
autocomplete.angular.min.js
after includingjQuery
&Angular.js
- Inject the
algolia.autocomplete
module - Add the
autocomplete
,aa-datasets
and the optionalaa-options
attribute to your search bar
<div ng-controller="yourController">
<input type="text" id="search-input" autocomplete aa-datasets="getDatasets()" />
</div>
<!-- [ include jQuery + Angular.js ] -->
<script src="https://cdn.jsdelivr.net/algoliasearch/3/algoliasearch.angular.min.js"></script>
<script src="https://cdn.jsdelivr.net/autocomplete.js/0/autocomplete.angular.min.js"></script>
<script>
angular.module('myApp', ['algoliasearch', 'algolia.autocomplete'])
.controller('yourController', ['$scope', 'algolia', function($scope, algolia) {
var client = algolia.Client('YourApplicationID', 'YourSearchOnlyAPIKey');
var index = client.initIndex('YourIndex');
$scope.getDatasets = function() {
return {
source: algolia.sources.hits(index, { hitsPerPage: 5 }),
displayKey: 'my_attribute',
templates: {
suggestion: function(suggestion) {
return suggestion._highlightResult.my_attribute.value;
}
}
};
};
$scope.$on('autocomplete:selected', function(event, suggestion, dataset) {
console.log(suggestion, dataset);
});
}]);
</script>
Note: You need to rely on jQuery
, the lite version embedded in Angular.js won't work.
Below is a faux mustache template describing the DOM structure of an autocomplete
dropdown menu. Keep in mind that header
, footer
, suggestion
, and empty
come from the provided templates detailed here.
<span class="aa-dropdown-menu{{#datasets}} aa-{{'with' or 'without'}}-{{name}}{{/datasets}}">
{{#datasets}}
<div class="aa-dataset-{{name}}">
{{{header}}}
<span class="aa-suggestions">
{{#suggestions}}
<div class="aa-suggestion">{{{suggestion}}}</div>
{{/suggestions}}
{{^suggestions}}
{{{empty}}}
{{/suggestions}}
</span>
{{{footer}}}
</div>
{{/datasets}}
{{empty}}
</span>
When an end-user mouses or keys over a .aa-suggestion
, the class aa-cursor
will be added to it. You can use this class as a hook for styling the "under cursor" state of suggestions.
Add the following CSS rules to add a default style:
.algolia-autocomplete {
width: 100%;
}
.algolia-autocomplete .aa-input, .algolia-autocomplete .aa-hint {
width: 100%;
}
.algolia-autocomplete .aa-hint {
color: #999;
}
.algolia-autocomplete .aa-dropdown-menu {
width: 100%;
background-color: #fff;
border: 1px solid #999;
border-top: none;
}
.algolia-autocomplete .aa-dropdown-menu .aa-suggestion {
cursor: pointer;
padding: 5px 4px;
}
.algolia-autocomplete .aa-dropdown-menu .aa-suggestion.aa-cursor {
background-color: #B2D7FF;
}
.algolia-autocomplete .aa-dropdown-menu .aa-suggestion em {
font-weight: bold;
font-style: normal;
}
Here is what the basic example looks like:
When initializing an autocomplete, there are a number of options you can configure.
-
autoselect
– Iftrue
, the first rendered suggestion in the dropdown will automatically have thecursor
class, and pressing<ENTER>
will select it. -
autoselectOnBlur
– Iftrue
, when the input is blurred, the first rendered suggestion in the dropdown will automatically have thecursor
class, and pressing<ENTER>
will select it. This option should be used on mobile, see #113 -
hint
– Iffalse
, the autocomplete will not show a hint. Defaults totrue
. -
debug
– Iftrue
, the autocomplete will not close onblur
. Defaults tofalse
. -
openOnFocus
– Iftrue
, the dropdown menu will open when the input is focused. Defaults tofalse
. -
dropdownMenuContainer
– If set with a DOM selector, it overrides the container of the dropdown menu. -
templates
– An optional hash overriding the default templates.dropdownMenu
– the dropdown menu template. The template should include all dataset placeholders.header
– the header to prepend to the dropdown menufooter
– the footer to append to the dropdown menuempty
– the template to display when none of the datasets are returning results. The templating function is called with a context containing the underlyingquery
.
-
cssClasses
– An optional hash overriding the default css classes.root
– the root classes. Defaults toalgolia-autocomplete
.prefix
– the CSS class prefix of all nested elements. Defaults toaa
.dropdownMenu
– the dropdown menu CSS class. Defaults todropdown-menu
.input
– the input CSS class. Defaults toinput
.hint
– the hint CSS class. Defaults tohint
.suggestions
– the suggestions list CSS class. Defaults tosuggestions
.suggestion
– the suggestion wrapper CSS class. Defaults tosuggestion
.cursor
– the cursor CSS class. Defaults tocursor
.dataset
– the dataset CSS class. Defaults todataset
.empty
– the empty CSS class. Defaults toempty
.
-
keyboardShortcuts
- Array of shortcut that will focus the input. For example if you want to binds
and/
you can specify:keyboardShortcuts: ['s', '/']
<script type="text/template" id="my-custom-menu-template">
<div class="my-custom-menu">
<div class="row">
<div class="col-sm-6">
<div class="aa-dataset-d1"></div>
</div>
<div class="col-sm-6">
<div class="aa-dataset-d2"></div>
<div class="aa-dataset-d3"></div>
</div>
</div>
</div>
</script>
<script>
$('#search-input').autocomplete(
{
templates: {
dropdownMenu: '#my-custom-menu-template',
footer: '<div class="branding">Powered by <img src="https://www.algolia.com/assets/algolia128x40.png" /></div>'
}
},
[
{
source: $.fn.autocomplete.sources.hits(index1, { hitsPerPage: 5 }),
name: 'd1',
templates: {
header: '<h4>List 1</h4>',
suggestion: function(suggestion) {
// FIXME
}
}
},
{
source: $.fn.autocomplete.sources.hits(index2, { hitsPerPage: 2 }),
name: 'd2',
templates: {
header: '<h4>List 2</h4>',
suggestion: function(suggestion) {
// FIXME
}
}
},
{
source: $.fn.autocomplete.sources.hits(index3, { hitsPerPage: 2 }),
name: 'd3',
templates: {
header: '<h4>List 3</h4>',
suggestion: function(suggestion, answer) {
// FIXME
}
}
}
]
);
</script>
minLength
– The minimum character length needed before suggestions start getting rendered. Defaults to1
.
An autocomplete is composed of one or more datasets. When an end-user modifies the value of the underlying input, each dataset will attempt to render suggestions for the new value.
Datasets can be configured using the following options.
-
source
– The backing data source for suggestions. Expected to be a function with the signature(query, cb)
. It is expected that the function will compute the suggestion set (i.e. an array of JavaScript objects) forquery
and then invokecb
with said set.cb
can be invoked synchronously or asynchronously. -
name
– The name of the dataset. This will be appended tott-dataset-
to form the class name of the containing DOM element. Must only consist of underscores, dashes, letters (a-z
), and numbers. Defaults to a random number. -
displayKey
– For a given suggestion object, determines the string representation of it. This will be used when setting the value of the input control after a suggestion is selected. Can be either a key string or a function that transforms a suggestion object into a string. Defaults tovalue
. Example function usage:displayKey: function(suggestion) { return suggestion.nickname || suggestion.firstName }
-
templates
– A hash of templates to be used when rendering the dataset. Note a precompiled template is a function that takes a JavaScript object as its first argument and returns a HTML string.-
empty
– Rendered when0
suggestions are available for the given query. Can be either a HTML string or a precompiled template. The templating function is called with a context containingquery
,isEmpty
, and any optional arguments that may have been forwarded by the source:function emptyTemplate({ query, isEmpty }, [forwarded args])
. -
footer
– Rendered at the bottom of the dataset. Can be either a HTML string or a precompiled template. The templating function is called with a context containingquery
,isEmpty
, and any optional arguments that may have been forwarded by the source:function footerTemplate({ query, isEmpty }, [forwarded args])
. -
header
– Rendered at the top of the dataset. Can be either a HTML string or a precompiled template. The templating function is called with a context containingquery
,isEmpty
, and any optional arguments that may have been forwarded by the source:function headerTemplate({ query, isEmpty }, [forwarded args])
. -
suggestion
– Used to render a single suggestion. The templating function is called with thesuggestion
, and any optional arguments that may have been forwarded by the source:function suggestionTemplate(suggestion, [forwarded args])
. Defaults to the value ofdisplayKey
wrapped in ap
tag i.e.<p>{{value}}</p>
.
-
A few helpers are provided by default to ease the creation of Algolia-based sources.
To build a source based on Algolia's hits
array, just use:
{
source: autocomplete.sources.hits(indexObj, { hitsPerPage: 2 }),
templates: {
suggestion: function(suggestion, answer) {
// FIXME
}
}
}
To build an Amazon-like autocomplete menu, suggesting popular queries and for the most popular one displaying the associated categories, you can use the popularIn
source:
{
source: autocomplete.sources.popularIn(popularQueriesIndexObj, { hitsPerPage: 3 }, {
source: 'sourceAttribute', // attribute of the `popularQueries` index use to query the `index` index
index: productsIndexObj, // targeted index
facets: 'facetedCategoryAttribute', // facet used to enrich the most popular query
maxValuesPerFacet: 3 // maximum number of facets returned
}, {
includeAll: true, // should it include an extra "All department" suggestion
allTitle: 'All departments' // the included category label
}),
templates: {
suggestion: function(suggestion, answer) {
var value = suggestion.sourceAttribute;
if (suggestion.facet) {
// this is the first suggestion
// and it has been enriched with the matching facet
value += ' in ' + suggestion.facet.value + ' (' + suggestion.facet.count + ')';
}
return value;
}
}
}
The source
options can also take a function. It enables you to have more control of the results returned by Algolia search. The function function(query, callback)
takes 2 parameters
query: String
: the text typed in the autocompletecallback: Function
: the callback to call at the end of your processing with the array of suggestions
source: function(query, callback) {
var index = client.initIndex('myindex');
index.search(query, { hitsPerPage: 1, facetFilters: 'category:mycat' }).then(function(answer) {
callback(answer.hits);
}, function() {
callback([]);
});
}
Or by reusing an existing source:
var hitsSource = autocomplete.sources.hits(index, { hitsPerPage: 5 });
source: function(query, callback) {
hitsSource(query, function(suggestions) {
// FIXME: Do stuff with the array of returned suggestions
callback(suggestions);
});
}
The autocomplete component triggers the following custom events.
-
autocomplete:opened
– Triggered when the dropdown menu of the autocomplete is opened. -
autocomplete:shown
– Triggered when the dropdown menu of the autocomplete is shown (opened and non-empty). -
autocomplete:empty
– Triggered when all datasets are empty. -
autocomplete:closed
– Triggered when the dropdown menu of the autocomplete is closed. -
autocomplete:updated
– Triggered when a dataset is rendered. -
autocomplete:cursorchanged
– Triggered when the dropdown menu cursor is moved to a different suggestion. The event handler will be invoked with 3 arguments: the jQuery event object, the suggestion object, and the name of the dataset the suggestion belongs to. -
autocomplete:selected
– Triggered when a suggestion from the dropdown menu is selected. The event handler will be invoked with 3 arguments: the jQuery event object, the suggestion object, and the name of the dataset the suggestion belongs to. -
autocomplete:cursorremoved
– Triggered when the cursor leaves the selections or its current index is lower than 0 -
autocomplete:autocompleted
– Triggered when the query is autocompleted. Autocompleted means the query was changed to the hint. The event handler will be invoked with 3 arguments: the jQuery event object, the suggestion object, and the name of the dataset the suggestion belongs to.
All custom events are triggered on the element initialized as the autocomplete.
Turns any input[type="text"]
element into an auto-completion menu. options
is an
options hash that's used to configure the autocomplete to your liking. Refer to
Options for more info regarding the available configs. Subsequent
arguments (*datasets
), are individual option hashes for datasets. For more
details regarding datasets, refer to Datasets.
$('.search-input').autocomplete({
minLength: 3
},
{
name: 'my-dataset',
source: mySource
});
Removes the autocomplete functionality and reverts the input
element back to its
original state.
$('.search-input').autocomplete('destroy');
Opens the dropdown menu of the autocomplete. Note that being open does not mean that the menu is visible. The menu is only visible when it is open and has content.
$('.search-input').autocomplete('open');
Closes the dropdown menu of the autocomplete.
$('.search-input').autocomplete('close');
Returns the current value of the autocomplete. The value is the text the user has
entered into the input
element.
var myVal = $('.search-input').autocomplete('val');
Sets the value of the autocomplete. This should be used in place of jQuery#val
.
$('.search-input').autocomplete('val', myVal);
Returns a reference to the autocomplete plugin and reverts jQuery.fn.autocomplete
to its previous value. Can be used to avoid naming collisions.
var autocomplete = jQuery.fn.autocomplete.noConflict();
jQuery.fn._autocomplete = autocomplete;
The standalone version API is similiar to jQuery's:
var search = autocomplete('#search', { hint: false }, [{
source: autocomplete.sources.hits(index, { hitsPerPage: 5 }
}]);
search.autocomplete.open();
search.autocomplete.close();
search.autocomplete.getVal();
search.autocomplete.setVal('Hey Jude');
search.autocomplete.destroy();
You can also pass a custom Typeahead instance in Autocomplete.js constructor:
var search = autocomplete('#search', { hint: false}, [{ ... }], new Typeahead({ ... }));
To start developing, you can use the following commands:
$ npm install
$ npm run dev
$ open http://localhost:8888/test/playground.html
Linting is done with eslint and Algolia's configuration and can be run with:
$ npm run lint
Unit tests are written using Jasmine and ran with Karma. Integration tests are using Mocha and Saucelabs.
To run the unit tests suite run:
$ npm test
To run the integration tests suite run:
$ npm run build
$ npm run server
$ ngrok 8888
$ TEST_HOST=http://YOUR_NGROK_ID.ngrok.com SAUCE_ACCESS_KEY=YOUR_KEY SAUCE_USERNAME=YOUR_USERNAME./node_modules/mocha/bin/mocha --harmony -R spec ./test/integration/test.js
- Bump the version in
package.json
andbower.json
npm run build
to update thedist/
files- Update CHANGELOG.md
- Commit and push to GitHub
- Create & push a new tag named
X.Y.Z
- Run
npm publish
This library has originally been forked from Twitter's typeahead.js library.