Turbolinks 5 is a ground-up rewrite with a new flow, new events, but the same core idea. It's available at turbolinks/turbolinks. This repository remains available for existing applications built on what we now call Turbolinks Classic.
Turbolinks makes following links in your web application faster. Instead of letting the browser recompile the JavaScript and CSS between each page change, it keeps the current page instance alive and replaces only the body (or parts of) and the title in the head. Think CGI vs persistent process.
This is similar to pjax, but instead of worrying about what element on the page to replace and tailoring the server-side response to fit, we replace the entire body by default, and let you specify which elements to replace on an opt-in basis. This means that you get the bulk of the speed benefits from pjax (no recompiling of the JavaScript or CSS) without having to tailor the server-side response. It just works.
Do note that this of course means that you'll have a long-running, persistent session with maintained state. That's what's making it so fast. But it also means that you may have to pay additional care not to leak memory or otherwise bloat that long-running state. That should rarely be a problem unless you're doing something really funky, but you do have to be aware of it. Your memory leaking sins will not be swept away automatically by the cleansing page change any more.
Turbolinks is designed to be as light-weight as possible (so you won't think twice about using it even for mobile stuff). It does not require jQuery or any other library to work. But it works great with the jQuery framework, or whatever else you have.
With Turbolinks pages will change without a full reload, so you can't rely on DOMContentLoaded
or jQuery.ready()
to trigger your code. Instead Turbolinks fires events on document
to provide hooks into the lifecycle of the page.
Event | Argument originalEvent.data |
Notes |
---|---|---|
page:before-change |
{url} |
The page is about to change. Cancellable with event.preventDefault() . Does not fire on history back/forward. |
page:fetch |
{url} |
A new page is about to be fetched from the server. |
page:receive |
{url} |
A page has been fetched from the server, but not yet parsed. |
page:before-unload |
[affectedNodes] |
Nodes are about to be changed. |
page:change |
[affectedNodes] |
Nodes have changed. Also fires on DOMContentLoaded . |
page:update |
Fired alongside both page:change and jQuery's ajaxSuccess (if available). |
|
page:load |
[newBody] |
A new body element has been loaded into the DOM. Does not fire on partial replacement or when a page is restored from cache, so as not to fire twice on the same body. |
page:partial-load |
[affectedNodes] |
New elements have been loaded into the DOM via partial replacement. |
page:restore |
A cached body element has been loaded into the DOM. | |
page:after-remove |
affectedNode |
An element has been removed from the DOM or body evicted from the cache and must be cleaned up. jQuery event listeners are cleaned up automatically. |
Example: load a fresh version of a page from the server
page:before-change
link clicked orTurbolinks.visit()
called (cancellable)page:fetch
about to send XHRpage:receive
received response from serverpage:before-unload
([currentBody]
) page has been parsed and is about to be changedpage:change
([newBody]
) new body is in placepage:update
page:load
([newBody]
) page has been loaded (progress bar hidden, scroll position updated)page:after-remove
(oldBody
) an old body has been evicted from the cache
Example: partial replacement with Turbolinks.replace()
page:before-unload
([currentNodes...]
) nodes are about to be changedpage:after-remove
(currentNode
) a node has been removed from the DOM and must be cleaned up (fires once per node)page:change
([newNodes...]
) new nodes are in placepage:update
page:partial-load
([newNodes...]
)
Example lifecycle setup:
// using jQuery for simplicity
$(document).on('ready', function(event) {
// initialize persistent state
});
$(document).on('ready page:load', function(event) {
// apply non-idempotent transformations to the body
});
$(document).on('page:partial-load', function(event) {
// apply non-idempotent transformations to the nodes in event.originalEvent.data
});
$(document).on('page:change', function(event) {
// idempotent function
});
$(document).on('page:after-remove', function(event) {
// delete references to the nodes in event.originalEvent.data to prevent memory leaks
});
By default, Turbolinks keeps 10 pages in memory (the full body element is kept in memory, so as not to lose state). On popstate, it attempts to restore pages from the cache. When a page exists in the cache, the following events are triggered:
page:before-unload
([currentBody]
) page is about to be changedpage:change
([cachedBody]
) body from cached page is in placepage:restore
The number of pages Turbolinks caches can be configured to suit your application's needs:
Turbolinks.pagesCached(); // View the current cache size
Turbolinks.pagesCached(20); // Set the cache size
If you need to make dynamic HTML updates in the current page and want it to be cached properly you can call:
Turbolinks.cacheCurrentPage();
Note: performing a partial replacement with URL change will remove the current page from the cache. This is because the replaced nodes cannot be brought back. If the user clicks the back button following a visit with partial replacement, the previous page will be fetched from the server.
Transition Cache makes loading cached pages instantaneous. Once a user has visited a page, returning later to the page results in an instant load.
For example, if Page A is already cached by Turbolinks and you are on Page B, clicking a link to Page A will immediately display the cached copy of Page A. Turbolinks will then fetch Page A from the server and replace the cached page once the new copy is returned.
To enable Transition Cache, include the following in your javascript:
Turbolinks.enableTransitionCache();
The one drawback is that dramatic differences in appearance between a cached copy and new copy may lead to a jarring affect for the end-user. This will be especially true for pages that have many moving parts (expandable sections, sortable tables, infinite scrolling, etc.).
If you find that a page is causing problems, you can have Turbolinks skip displaying the cached copy by adding data-no-transition-cache
to any DOM element on the offending page.
Because Turbolinks skips the traditional full page reload, browsers won't display their native progress bar when changing pages. To fill this void, Turbolinks offers a JavaScript-and-CSS-based progress bar to display page loading progress (as of v3.0, the progress bar is turned on by default).
To disable (or re-enable) the progress bar, include one of the following in your JavaScript:
Turbolinks.ProgressBar.disable();
Turbolinks.ProgressBar.enable();
The progress bar is implemented on the <html>
element's pseudo :before
element and can be customized by including CSS with higher specificity than the included styles. For example:
html.turbolinks-progress-bar::before {
background-color: red !important;
height: 5px !important;
}
Control the progress bar manually using these methods:
Turbolinks.ProgressBar.start();
Turbolinks.ProgressBar.advanceTo(value); // where value is between 0-100
Turbolinks.ProgressBar.done();
Turbolinks will be enabled only if the server has rendered a GET
request.
Why not all request types? Some browsers track the request method of each page load, but triggering pushState
methods doesn't change this value. This could lead to the situation where pressing the browser's reload button on a page that was fetched with Turbolinks would attempt a POST
(or something other than GET
) because the last full page load used that method.
By default, all internal HTML links will be funneled through Turbolinks, but you can opt out by marking links or their parent container with data-no-turbolink
. For example, if you mark a div with data-no-turbolink
, then all links inside of that div will be treated as regular links. If you mark the body, every link on that entire page will be treated as regular links.
<a href="/">Home (via Turbolinks)</a>
<div id="some-div" data-no-turbolink>
<a href="/">Home (without Turbolinks)</a>
</div>
Note that internal links to files containing a file extension other than .html will automatically be opted out of Turbolinks. To whitelist additional file extensions to be processed by Turbolinks, use Turbolinks.allowLinkExtensions()
.
Turbolinks.allowLinkExtensions(); // => ['html']
Turbolinks.allowLinkExtensions('md'); // => ['html', 'md']
Turbolinks.allowLinkExtensions('coffee', 'scss'); // => ['html', 'md', 'coffee', 'scss']
Also, Turbolinks is installed as the last click handler for links. So if you install another handler that calls event.preventDefault()
, Turbolinks will not run. This ensures that you can safely use Turbolinks with things like data-method
, data-remote
, or data-confirm
from Rails.
Note: in v3.0, the default behavior of redirect_to
is to redirect via Turbolinks on XHR + non-GET requests. You can opt-out of this behavior by passing turbolinks: false
to redirect_to
.
By default, Turbolinks includes itself in ActionController::Base
. To opt out of the Turbolinks features in certain controllers (redirect_to
behavior, request_method
cookie, X-XHR-Referer
referrer check, etc.), set config.turbolinks.auto_include
to false
in application.rb
and include Turbolinks::Controller
in the controllers where you use Turbolinks.
If you have a lot of existing JavaScript that binds elements on jQuery.ready()
, you can pull the jquery.turbolinks library into your project that will trigger ready()
when Turbolinks triggers the page:load
event. It may restore functionality of some libraries.
Add the gem to your project, then add the following line to your JavaScript manifest file, after jquery.js
but before turbolinks.js
:
//= require jquery.turbolinks
Additional details and configuration options can be found in the jquery.turbolinks README.
You can track certain assets, like application.js
and application.css
, that you want to ensure are always of the latest version inside a Turbolinks session. This is done by marking those asset links with data-turbolinks-track
, like so:
<link href="/assets/application-9bd64a86adb3cd9ab3b16e9dca67a33a.css" rel="stylesheet"
type="text/css" data-turbolinks-track>
If those assets change URLs (embed an md5 stamp to ensure this), the page will do a full reload instead of going through Turbolinks. This ensures that all Turbolinks sessions will always be running off your latest JavaScript and CSS.
When this happens, you'll technically be requesting the same page twice. Once through Turbolinks to detect that the assets changed, and then again when we do a full redirect to that page.
Turbolinks will evaluate any script tags in pages it visits, if those tags do not have a type or if the type is text/javascript
. All other script tags will be ignored.
As a rule of thumb when switching to Turbolinks, move all of your javascript tags inside the head
and then work backwards, only moving javascript code back to the body if absolutely necessary. If you have any script tags in the body you do not want to be re-evaluated then you can set the data-turbolinks-eval
attribute to false
:
<script type="text/javascript" data-turbolinks-eval=false>
console.log("I'm only run once on the initial page load");
</script>
Turbolinks will not re-evaluate script tags on back/forward navigation, unless their data-turbolinks-eval
attribute is set to always
:
<script type="text/javascript" data-turbolinks-eval=always>
console.log("I'm run on every page load, including history back/forward");
</script>
You can use Turbolinks.visit(path)
to go to a URL through Turbolinks.
You can also use redirect_to path, turbolinks: true
in Rails to perform a redirect via Turbolinks.
Turbolinks
Function | Arguments | Notes |
---|---|---|
visit() |
path , options |
Load a new page and change the URL. |
replace() |
stringOrDocument , options |
Replace the current page without changing the URL. |
Option | Type | Notes |
---|---|---|
change |
Array |
Replace only the nodes with the given ids. |
append |
Array |
Append the children of nodes with the given ids. |
prepend |
Array |
Prepend the children of nodes with the given ids. |
keep |
Array |
Replace the body but keep the nodes with the given ids. |
flush |
Boolean |
Replace the body, including data-turbolinks-permanent nodes. |
title |
Boolean or String |
If false , don't update the document title. If a string, set the value as title. |
scroll |
Boolean |
If false , don't scroll to top (or #target ) after the page is loaded. |
cacheRequest |
Boolean |
Enable/disable the request cache. |
showProgressBar |
Boolean |
Show/hide the progress bar during the request. |
Function | Arguments | Notes |
---|---|---|
pagesCached() |
None or Number |
Get or set the maximum number of pages that should be cached. |
cacheCurrentPage() |
||
enableTransitionCache() |
||
disableRequestCaching() |
||
allowLinkExtensions() |
String ... |
Whitelist additional file extensions to be processed by Turbolinks. |
Property | Notes |
---|---|
supported |
true if the browser fully supports Turbolinks. |
EVENTS |
Map of event names. |
Turbolinks.ProgressBar
Function | Arguments | Notes |
---|---|---|
enable() |
||
disable() |
||
start() |
||
advanceTo() |
Number |
Value must be between 0 and 100. |
done() |
Like pjax, this naturally only works with browsers capable of pushState
. But of course we fall back gracefully to full page reloads for browsers that do not support it.
Note: there is currenty no fallback for partial replacement on browsers that don't support pushState
.
Turbolinks is designed to work with any browser that fully supports pushState
and all the related APIs. This includes Safari 6.0+ (but not Safari 5.1.x!), IE10, and latest Chromes and Firefoxes.
Do note that existing JavaScript libraries may not all be compatible with Turbolinks out of the box due to the change in instantiation cycle. You might very well have to modify them to work with Turbolinks's new set of events. For help with this, check out the Turbolinks Compatibility project.
Turbolinks works with Rails 3.2 and newer.
- External scripts are not guaranteed to execute in DOM order (#513)
- Iframes in
data-turbolinks-permanent
nodes are reloaded on page load (#511) - Audio and video elements in
data-turbolinks-permanent
nodes are paused on page load (#508) - Partial replacement removes pages from the cache (#551)
- Add
gem 'turbolinks'
to your Gemfile. - Run
bundle install
. - Add
//= require turbolinks
to your Javascript manifest file (usually found atapp/assets/javascripts/application.js
). If your manifest requires both turbolinks and jQuery, make sure turbolinks is listed after jQuery. - Restart your server and you're now using turbolinks!
Ruby:
rake test:all
BUNDLE_GEMFILE=Gemfile.rails42 bundle
BUNDLE_GEMFILE=Gemfile.rails42 rake test
JavaScript:
bundle install
npm install
script/test # requires phantomjs >= 2.0
script/server # https://localhost:9292/javascript/index.html
These projects are not affiliated with or endorsed by the Rails Turbolinks team.
- Flask Turbolinks (Python Flask)
- Django Turbolinks (Python Django)
- ASP.NET MVC Turbolinks
- PHP Turbolinks Component (Symfony Component)
- PHP Turbolinks Package (Laravel Package)
- Grails Turbolinks (Grails Plugin)
Thanks to Chris Wanstrath for his original work on Pjax. Thanks to Sam Stephenson and Josh Peek for their additional work on Pjax and Stacker and their help with getting Turbolinks released. Thanks to David Estes and Nick Reed for handling the lion's share of post-release issues and feature requests. And thanks to everyone else who's fixed or reported an issue!