This project adheres to the [Open Code of Conduct][code-of-conduct]. By participating, you are expected to uphold this code. [code-of-conduct]: http://todogroup.org/opencodeofconduct/#fetch/[email protected]
This module exposes the fetch
function, which is an easier way to make web requests and handle
responses than using an XMLHttpRequest. This module is written as closely as
possible to the standard Fetch specification at https://fetch.spec.whatwg.org.
You'll need a Promise polyfill for older browsers.
$ bower install es6-promise
For a node.js implementation, try node-fetch.
For use with webpack, refer to Using WebPack with shims and polyfills.
Differences with github/fetch
The major difference is the addition of the following methods and properties to the Request
object:
abort
: cancels a requestsend
: sends a requestthen
,catch
: both update the internal promise object, making requests look like promisesgetXHR
: returns the used XHR objectisAborted
: set totrue
when the request is abortedpromise
: promise of the requestxhr
: XMLHttpRequest object used for the request
As a result, this module is not a polyfill. Instead, it exports the following objects to enable their import in environments that already have an implementation of fetch
:
Body
Request
Response
Headers
fetch
The fetch
function supports any HTTP method. We'll focus on GET and POST
example requests.
fetch('/users.html')
.then(function(response) {
return response.text()
}).then(function(body) {
document.body.innerHTML = body
})
fetch('/users.json')
.then(function(response) {
return response.json()
}).then(function(json) {
console.log('parsed json', json)
}).catch(function(ex) {
console.log('parsing failed', ex)
})
fetch('/users.json').then(function(response) {
console.log(response.headers.get('Content-Type'))
console.log(response.headers.get('Date'))
console.log(response.status)
console.log(response.statusText)
})
var form = document.querySelector('form')
fetch('/users', {
method: 'post',
body: new FormData(form)
})
fetch('/users', {
method: 'post',
headers: {
'Accept': 'application/json',
'Content-Type': 'application/json'
},
body: JSON.stringify({
name: 'Hubot',
login: 'hubot',
})
})
var input = document.querySelector('input[type="file"]')
var data = new FormData()
data.append('file', input.files[0])
data.append('user', 'hubot')
fetch('/avatars', {
method: 'post',
body: data
})
The fetch
specification differs from jQuery.ajax()
in mainly two ways that
bear keeping in mind:
-
The Promise returned from
fetch()
won't reject on HTTP error status even if the response is a HTTP 404 or 500. Instead, it will resolve normally, and it will only reject on network failure, or if anything prevented the request from completing. -
By default,
fetch
won't send any cookies to the server, resulting in unauthenticated requests if the site relies on maintaining a user session.
To have fetch
Promise reject on HTTP error statuses, i.e. on any non-2xx
status, define a custom response handler:
function checkStatus(response) {
if (response.status >= 200 && response.status < 300) {
return response
} else {
var error = new Error(response.statusText)
error.response = response
throw error
}
}
function parseJSON(response) {
return response.json()
}
fetch('/users')
.then(checkStatus)
.then(parseJSON)
.then(function(data) {
console.log('request succeeded with JSON response', data)
}).catch(function(error) {
console.log('request failed', error)
})
To abort (cancel) a fetch
, simply call abort()
on the request object returned by fetch:
var request = fetch('/big-data')
.then(checkStatus)
.then(parseJSON)
.catch(function(error) {
if (request.isAborted) {
console.log('request was aborted')
}
})
// Later…
request.abort()
To automatically send cookies for the current domain, the credentials
option
must be provided:
fetch('/users', {
credentials: 'same-origin'
})
This option makes fetch
behave similar to XMLHttpRequest with regards to
cookies. Otherwise, cookies won't get sent, resulting in these requests not
preserving the authentication session.
Use the include
value to send cookies in a cross-origin resource sharing (CORS) request.
fetch('https://example.com:1234/users', {
credentials: 'include'
})
Like with XMLHttpRequest, the Set-Cookie
response header returned from the
server is a forbidden header name and therefore can't be programatically
read with response.headers.get()
. Instead, it's the browser's responsibility
to handle new cookies being set (if applicable to the current URL). Unless they
are HTTP-only, new cookies will be available through document.cookie
.
Due to limitations of XMLHttpRequest, the response.url
value might not be
reliable after HTTP redirects on older browsers.
The solution is to configure the server to set the response HTTP header
X-Request-URL
to the current URL after any redirect that might have happened.
It should be safe to set it unconditionally.
# Ruby on Rails controller example
response.headers['X-Request-URL'] = request.url
This server workaround is necessary if you need reliable response.url
in
Firefox < 32, Chrome < 37, Safari, or IE.
Latest ✔ | Latest ✔ | 10+ ✔ | Latest ✔ | 6.1+ ✔ |