PATCH request method
The PATCH HTTP method applies partial modifications to a resource.
PATCH is somewhat analogous to the "update" concept found in CRUD (in general, HTTP is different than CRUD, and the two should not be confused).
In comparison with PUT, a PATCH serves as a set of instructions for modifying a resource, whereas PUT represents a complete replacement of the resource.
A PUT request is always idempotent (repeating the same request multiple times results in the resource remaining in the same state), whereas a PATCH request may not always be idempotent.
For instance, if a resource includes an auto-incrementing counter, a PUT request will overwrite the counter (since it replaces the entire resource), but a PATCH request may not.
Like POST, a PATCH request can potentially have side effects on other resources.
A server can advertise support for PATCH by adding it to the list in the Allow or Access-Control-Allow-Methods (for CORS) response headers.
Another implicit indication that PATCH is supported is the Accept-Patch header (usually after an OPTIONS request on a resource), which lists the media-types the server is able to understand in a PATCH request for a resource.
| Request has body | Yes |
|---|---|
| Successful response has body | May |
| Safe | No |
| Idempotent | No |
| Cacheable | Only if freshness information is included |
| Allowed in HTML forms | No |
Syntax
PATCH <request-target>["?"<query>] HTTP/1.1
<request-target>-
Identifies the target resource of the request when combined with the information provided in the
Hostheader. This is an absolute path (e.g.,/path/to/file.html) in requests to an origin server, and an absolute URL in requests to proxies (e.g.,http://www.example.com/path/to/file.html). <query>Optional-
An optional query component preceded by a question-mark
?. Often used to carry identifying information in the form ofkey=valuepairs.
Examples
>Successfully modifying a resource
Assume there is a resource on the server representing a user with a numeric ID of 123 in the following format:
{
"firstName": "Example",
"LastName": "User",
"userId": 123,
"signupDate": "2024-09-09T21:48:58Z",
"status": "active",
"registeredDevice": {
"id": 1,
"name": "personal",
"manufacturer": {
"name": "Hardware corp"
}
}
}
Instead of sending a JSON object to fully overwrite a resource, a PATCH modifies only specific parts of the resource.
This request updates the status field:
PATCH /users/123 HTTP/1.1
Host: example.com
Content-Type: application/json
Content-Length: 27
Authorization: Bearer ABC123
{
"status": "suspended"
}
The interpretation and authentication of the PATCH request depend on the implementation.
Success can be indicated by any of the successful response status codes.
In this example, a 204 No Content is used as there's no need to transmit a body with additional context about the operation.
An ETag is provided so the caller can perform a conditional request in future:
HTTP/1.1 204 No Content
Content-Location: /users/123
ETag: "e0023aa4f"
Specifications
| Specification |
|---|
| RFC 5789> |
Browser compatibility
The browser doesn't use the PATCH method for user-initiated actions, so "browser compatibility" doesn't apply.
Developers can set this request method using fetch().
See also
- HTTP request methods
- HTTP response status codes
- HTTP headers
204Allow,Access-Control-Allow-MethodsheadersAccept-Patch– specifies the patch document formats accepted by the server- JSON Patch Generator