v5.0 pre-release

DOM Events


Create a custom DOM event in the browser for the purpose of implementing custom functionality.
dispatch_event immediately dispatches a DOM CustomEvent in the browser. These events bubble up through the DOM and are cancelable.
If jQuery is in use (and available in the global window scope) dispatch_event will also trigger a jQuery event with the same name.
batch: String, # [null] - add the operation to a named batch
cancel: Boolean, # [false] - cancel the operation (for use on client)
delay: Integer, # [0] - wait for n milliseconds before running
detail: Object, # [null] - assigned to event.detail
name: String, # required - the name of the DOM event to dispatch (can be custom)
select_all: Boolean, # [false] - operate on list of elements returned from selector
selector: String, # [document] - string containing a CSS selector or XPath expression
xpath: Boolean # [false] - process the selector as an XPath expression
While the event itself is frequently the payload, you can build very powerful functionality by passing a Hash of metadata to the detail parameter.
The detail parameter will convert user_id to userId - snake_case to camelCase - for JS consumption on the client.
Developers frequently use dispatch_event to notify the client when long-running server processes are completed. You can see an example in Leveraging Stimulus.

Life-cycle Callback Events

  • cable-ready:before-dispatch-event
  • cable-ready:after-dispatch-event
Life-cycle events for dispatch_event are raised on document.



Add a meta tag to your document head. If a meta tag with the same name already exists, update the content to a new value.
batch: String, # [null] - add the operation to a named batch
cancel: Boolean, # [false] - cancel the operation (for use on client)
content: String, # required - the content attribute of the meta tag
delay: Integer, # [0] - wait for n milliseconds before running
name: String # required - the name attribute of the meta tag
meta tags are under-appreciated as a semantically-appropriate place to place data that can be consumed by multiple components on a page. It's even less common to see scenarios where a meta tag is updated after the page has been rendered. This is a shame, because there's absolutely no good reason not to do so.


Here is a Stimulus controller that watches a meta tag for content changes, which update the attached element. In this example, imagine a notification "pill" component that is not displayed if there is no text value.
import { Controller } from 'stimulus'
export default class extends Controller {
static values = { tag: String }
initialize () {
this.metaTag = document.head.querySelector(
this.element.textContent = this.metaTag.content = new MutationObserver(this.meta)
connect () {, {
attributeFilter: ['content']
disconnect () {
meta = mutation => {
this.element.textContent = mutation[0].target.content || ''
<meta name="pill-messages" content="4" />
<span class="badge text-danger" data-controller="pill" data-pill-tag-value="messages"></span>
This technique allows the developer to update data that concerns the page without a potentially brittle expectation that the server knows the id of every component that needs to be updated.

Life-cycle Callback Events

  • cable-ready:before-set-meta
  • cable-ready:after-set-meta
Life-cycle events for set_meta are raised on document.