Responding to an idempotent HTTP POST request - http

Part of our RESTful API will allow users to register an item with a serial number. As the serial number is not globally unique it cannot be used as the identifier of the resource, so we'll use a POST to the parent resource which will generate an identifier, e.g.
POST /my/items
<item serial-number="ABCDEF" />
In the case where the item is not already registered, the HTTP semantics are well defined. We return a Location header, and the registered item as the entity body, e.g.
HTTP 201 Created
Location: /my/items/1234
<item id="1234" serial-number="ABCDEF" />
However, in the case where the item is already registered, the API should be idempotent and return the previously registered item without creating a new one. My best guess is that it should then return a 200 OK status code, and use the Content-Location header to indicate where the item actually came from, e.g.
HTTP 200 OK
Content-Location: /my/items/1234
<item id="1234" serial-number="ABCDEF" />
Would this seem reasonable? I'm not entirely clear whether the Location or Content-Location is more suitable in the second case.

I had similar requirement recently. For idempotent operations PUT is the best way. You're right there is a mismatch between the external-id and the internal one. I solved it by creating a dedicated resource for the external id target:
PUT /api-user/{username}/items/{serialNumber}
Internally I resolve it as a CREATE in case I have no item for 'username' and 'ABCDEF' serial number or UPDATE in case I do.
In case it was a CREATE I return 201 for an UPDATE 200. Further more the returned payload contains both homegrown id and the external serial number like you suggested in your payload.

Here is an interesting discussion on usages of the two headers. It claims Content-Location is not defined for PUT or POST so Location is possibly the better option in your case. It is certainly not clear cut which is better though.
Overall I think your approach makes sense though.

Related

Should I use PATCH or PUT in my REST API?

I want to design my rest endpoint with the appropriate method for the following scenario.
There is a group. Each group has a status. The group can be activated or inactivated by the admin.
Should I design my end point as
PUT /groups/api/v1/groups/{group id}/status/activate
OR
PATCH /groups/api/v1/groups/{group id}
with request body like
{action:activate|deactivate}
The PATCH method is the correct choice here as you're updating an existing resource - the group ID. PUT should only be used if you're replacing a resource in its entirety.
Further information on partial resource modification is available in RFC 5789. Specifically, the PUT method is described as follows:
Several applications extending the Hypertext Transfer Protocol
(HTTP) require a feature to do partial resource modification. The
existing HTTP PUT method only allows a complete replacement of a
document. This proposal adds a new HTTP method, PATCH, to modify an
existing HTTP resource.
The R in REST stands for resource
(Which isn't true, because it stands for Representational, but it's a good trick to remember the importance of Resources in REST).
About PUT /groups/api/v1/groups/{group id}/status/activate: you are not updating an "activate". An "activate" is not a thing, it's a verb. Verbs are never good resources. A rule of thumb: if the action, a verb, is in the URL, it probably is not RESTful.
What are you doing instead? Either you are "adding", "removing" or "updating" an activation on a Group, or if you prefer: manipulating a "status"-resource on a Group. Personally, I'd use "activations" because they are less ambiguous than the concept "status": creating a status is ambiguous, creating an activation is not.
POST /groups/{group id}/activation Creates (or requests the creation of) an activation.
PATCH /groups/{group id}/activation Updates some details of an existing activation. Since a group has only one activation, we know what activation-resource we are referring to.
PUT /groups/{group id}/activation Inserts-or-replaces the old activation. Since a group has only one activation, we know what activation-resource we are referring to.
DELETE /groups/{group id}/activation Will cancel, or remove the activation.
This pattern is useful when the "activation" of a Group has side-effects, such as payments being made, mails being sent and so on. Only POST and PATCH may have such side-effects. When e.g. a deletion of an activation needs to, say, notify users over mail, DELETE is not the right choice; in that case you probably want to create a deactivation resource: POST /groups/{group_id}/deactivation.
It is a good idea to follow these guidelines, because this standard contract makes it very clear for your clients, and all the proxies and layers between the client and you, know when it is safe to retry, and when not. Let's say the client is somewhere with flaky wifi, and its user clicks on "deactivate", which triggers a DELETE: If that fails, the client can simply retry, until it gets a 404, 200 or anything else it can handle. But if it triggers a POST to deactivation it knows not to retry: the POST implies this.
Any client now has a contract, which, when followed, will protect against sending out 42 emails "your group has been deactivated", simply because its HTTP-library kept retrying the call to the backend.
Updating a single attribute: use PATCH
PATCH /groups/{group id}
In case you wish to update an attribute. E.g. the "status" could be an attribute on Groups that can be set. An attribute such as "status" is often a good candidate to limit to a whitelist of values. Examples use some undefined JSON-scheme:
PATCH /groups/{group id} { "attributes": { "status": "active" } }
response: 200 OK
PATCH /groups/{group id} { "attributes": { "status": "deleted" } }
response: 406 Not Acceptable
Replacing the resource, without side-effects use PUT.
PUT /groups/{group id}
In case you wish to replace an entire Group. This does not necessarily mean that the server actually creates a new group and throws the old one out, e.g. the ids might remain the same. But for the clients, this is what PUT can mean: the client should assume he gets an entirely new item, based on the server's response.
The client should, in case of a PUT request, always send the entire resource, having all the data that is needed to create a new item: usually the same data as a POST-create would require.
PUT /groups/{group id} { "attributes": { "status": "active" } }
response: 406 Not Acceptable
PUT /groups/{group id} { "attributes": { "name": .... etc. "status": "active" } }
response: 201 Created or 200 OK, depending on whether we made a new one.
A very important requirement is that PUT is idempotent: if you require side-effects when updating a Group (or changing an activation), you should use PATCH. So, when the update results in e.g. sending out a mail, don't use PUT.
I would recommend using PATCH, because your resource 'group' has many properties but in this case, you are updating only the activation field(partial modification)
according to the RFC5789 (https://www.rfc-editor.org/rfc/rfc5789)
The existing HTTP PUT method only allows a complete replacement of
a document. This proposal adds a new HTTP method, PATCH, to modify
an existing HTTP resource.
Also, in more details,
The difference between the PUT and PATCH requests is reflected in the
way the server processes the enclosed entity to modify the resource
identified by the Request-URI. In a PUT request, the enclosed entity
is considered to be a modified version of the resource stored on the
origin server, and the client is requesting that the stored version
be replaced. With PATCH, however, the enclosed entity contains a set
of instructions describing how a resource currently residing on the
origin server should be modified to produce a new version. The PATCH
method affects the resource identified by the Request-URI, and it
also MAY have side effects on other resources; i.e., new resources
may be created, or existing ones modified, by the application of a
PATCH.
PATCH is neither safe nor idempotent as defined by [RFC2616],
Section
9.1.
Clients need to choose when to use PATCH rather than PUT. For
example, if the patch document size is larger than the size of the
new resource data that would be used in a PUT, then it might make
sense to use PUT instead of PATCH. A comparison to POST is even more
difficult, because POST is used in widely varying ways and can
encompass PUT and PATCH-like operations if the server chooses. If
the operation does not modify the resource identified by the Request-
URI in a predictable way, POST should be considered instead of PATCH
or PUT.
The response code for PATCH is
The 204 response code is used because the response does not carry a
message body (which a response with the 200 code would have). Note
that other success codes could be used as well.
also refer thttp://restcookbook.com/HTTP%20Methods/patch/
Caveat: An API implementing PATCH must patch atomically. It MUST not
be possible that resources are half-patched when requested by a GET.
Since you want to design an API using the REST architectural style you need to think about your use cases to decide which concepts are important enough to expose as resources. Should you decide to expose the status of a group as a sub-resource you could give it the following URI and implement support for both GET and PUT methods:
/groups/api/groups/{group id}/status
The downside of this approach over PATCH for modification is that you will not be able to make changes to more than one property of a group atomically and transactionally. If transactional changes are important then use PATCH.
If you do decide to expose the status as a sub-resource of a group it should be a link in the representation of the group. For example if the agent gets group 123 and accepts XML the response body could contain:
<group id="123">
<status>Active</status>
<link rel="/linkrels/groups/status" uri="/groups/api/groups/123/status"/>
...
</group>
A hyperlink is needed to fulfill the hypermedia as the engine of application state condition of the REST architectural style.
One possible option to implement such behavior is
PUT /groups/api/v1/groups/{group id}/status
{
"Status":"Activated"
}
And obviously, if someone need to deactivate it, PUT will have Deactivated status in JSON.
In case of necessity of mass activation/deactivation, PATCH can step into the game (not for exact group, but for groups resource:
PATCH /groups/api/v1/groups
{
{ “op”: “replace”, “path”: “/group1/status”, “value”: “Activated” },
{ “op”: “replace”, “path”: “/group7/status”, “value”: “Activated” },
{ “op”: “replace”, “path”: “/group9/status”, “value”: “Deactivated” }
}
In general this is idea as #Andrew Dobrowolski suggesting, but with slight changes in exact realization.
I would generally prefer something a bit simpler, like activate/deactivate sub-resource (linked by a Link header with rel=service).
POST /groups/api/v1/groups/{group id}/activate
or
POST /groups/api/v1/groups/{group id}/deactivate
For the consumer, this interface is dead-simple, and it follows REST principles without bogging you down in conceptualizing "activations" as individual resources.

Can the Location header be used for multiple resource locations in a 201 Created response?

What is the correct verb and response to accept a batch PUT create and then return multiple locations? The Location header only appears to support one single Uri.
I assumed originally I could use PUT for a batch create and return an string array with a list of Uris, but in looking at the specification, that doesn't appear to be supported, but it isn't entirely clear either.
Are multiple Location headers permissible as an alternative?
Any advice?
No, you can have only one Location header field (and yes, it is clear from the spec).
That being said, PUT is for creating/updating a single resource. It seems that you're using it for something it is not designed for...
PUT can only create one resource, because according to the specification:
The PUT method requests that the state of the target resource be created or replaced with the state defined by the representation enclosed in the request message payload.
(https://www.rfc-editor.org/rfc/rfc7231#section-4.3.4)
However, it is possible for a request with a different verb (such as POST) to create multiple resources and return a 201 Created status. According to the specification:
The primary resource created by the request is identified by either a Location header field in the response or, if no Location field is received, by the effective request URI.
The 201 response payload typically describes and links to the resource(s) created.
(https://www.rfc-editor.org/rfc/rfc7231#section-6.3.2)
So it is OK to return the URIs of multiple created resources in a 201 response to, e.g., a POST request, but the Location header may only contain one URI.
RFC 5988 introduces Link HTTP Header. It can be used for this purpose.

Are PUT and POST requests required/expected to have a request body?

I'm writting a RESTful api, and at I'm thinking about the process of a user creating a key. I have the following possibilities:
GET request to /new/<keyname> - although it's very easy I think I won't use this, because I heard GET is for retrieving and/or listing information;
POST request to /<keyname> - This seemed to me easy and simple enough, but does not pass any data in the request body. Can I do it this way ? Is this weird ?
POST request to /keys passing in the request body "keyname=SomeKey" - Is this the correct way ?
I looked at this API from joyent and in all their PUT and POST requests they pass some data in the request body. Is this expected ? Is it really wrong not to require a request body in a PUT and POST request ?
I asked this question on the Http-WG. This was the most precise answer I got http://lists.w3.org/Archives/Public/ietf-http-wg/2010JulSep/0276.html
In summary, POST does not require a body. I would expect the same justification can be applied to PUT.
RFC2616 is the base RFC for HTTP 1.1
In the most general form, an HTTP message is this (note the optional body):
generic-message = start-line
*(message-header CRLF)
CRLF
[ message-body ]
start-line = Request-Line | Status-Line
Reading further gives this:
9.5 POST
The POST method is used to request that the origin server accept the
entity enclosed in the request as a new subordinate of the resource
identified by the Request-URI in the Request-Line. ...
and
9.6 PUT
The PUT method requests that the enclosed entity be stored under the
supplied Request-URI. ...
The fundamental difference between the POST and PUT requests is
reflected in the different meaning of the Request-URI. The URI in a
POST request identifies the resource that will handle the enclosed
entity. That resource might be a data-accepting process, a gateway to
some other protocol, or a separate entity that accepts annotations.
In contrast, the URI in a PUT request identifies the entity enclosed
with the request -- the user agent knows what URI is intended and the
server MUST NOT attempt to apply the request to some other resource.
Both POST and PUT include the phrase entity enclosed in the request.
Based on my reading, I believe that a body is desired (a non-normative description, I know) for both POST and PUT.
In the context of REST, POST is create and PUT is update. I can imagine creating an empty object (perhaps a placeholder for future information), but I don't imagine much use of an empty update.
It is not required. You can send a POST/PUT request without a body and instead use query string parameters. But be careful if your parameters contain characters that are not HTTP valid you will have to encode them.
For example if you need to POST 'hello world' to and end point you would have to make it look like this: http://api.com?param=hello%20world
Probably the best way is your third option: POST to /keys with keyname=SomeKey.
Here's why: You may wish to add another function to your API, for example create_new_user. It would then be difficult to tell the difference between a user trying to POST a key called create_new_user and a user trying to use the create_new_user function.
You are correct in saying that you should not be using GET to do this operation as the GET operation "SHOULD NOT have the significance of taking an action
other than retrieval." (RFC 2616).
To answer your question in one line. Yes it is expected to have Body/Content in body, but it is not required(Mandatory).
According to okHttp3 (an HTTP library for android): the following methods need a body: POST, PUT, PATCH, PROPPATCH (WebDAV) and REPORT (source). It even crashes if you try to do a request with the given methods without a body.

How should I implement a COUNT verb in my RESTful web service?

I've written a RESTful web service that supports the standard CRUD operations, and that can return a set of objects matching certain criteria (a SEARCH verb), but I'd like to add a higher-order COUNT verb, so clients can count the resources matching search criteria without having to fetch all of them.
A few options that occur to me:
Ignoring the HTTP specification and returning the object count in the response body of a HEAD request.
Duplicating the SEARCH verb's logic, but making a HEAD request instead of a GET request. The server then would encode the object count in a response header.
Defining a new HTTP method, COUNT, that returns the object count in the response body.
I'd prefer the API of the first approach, but I have to strike that option because it's non-compliant. The second approach seems most semantically correct, but the API isn't very convenient: clients will have to deal with response headers, when most of the time they want to be able to do something easy like response.count. So I'm leaning toward the third approach, but I'm concerned about the potential problems involved with defining a new HTTP method.
What would you do?
The main purpose of rest is to define a set of resources that you interact with using well defined verbs. You must thus avoid to define your own verbs. The number of resources should be considered as a different resource, with its own uri that you can simply GET.
For example:
GET resources?crit1=val1&crit2=val2
returns the list of resources and
GET resources/count?crit1=val1&crit2=val2
Another option is to use the conneg: e.g. Accept: text/uri-list returns the resources list and Accept: text/plain returns only the count
You can use HEAD without breaking the HTTP specification and you can indicate the count by using an HTTP Range header in the response:
HEAD /resource/?search=lorem
Response from the service, assuming that you return the first 20 results by default:
...
Content-Range: resources 0-20/12345
...
This way you transfer the amount of resources to the client within the header of the response message without the need to return a message body.
Update:
The solution suggested Yannick Loiseau will work fine. Just wanted to provide one other alternative approach which can be used to achieve what you need without the need to define a new resource of verb.
You can use GET and add the count into the body of the message. Then, if you API allows clients to request a range of results, you can use that in order to limit the size of message body to a minimum (since you only want the count). One way to do that would be to request an empty range (from 0 to 0), for example:
GET /resource/?search=lorem&range=0,0
The service could then respond as follows, indicating that there are 1234 matching resources in the result set:
<?xml version="1.0" encoding="UTF-8" ?>
<resources range="0-0/1234" />
Ignoring the HTTP specification and returning the object count in the response body of a HEAD request.
IMHO, this is a very bad idea. It may not work simply because you might have intermediaries that don't ignore the HTTP spec.
Defining a new HTTP method, COUNT, that returns the object count in the response body.
There is no problem with this approach. HTTP is extendable and you can define your own verbs. Some firewalls prohibit this, but they are usually also prohibit POST and DELETE and X-HTTP-Method-Override header is widely supported.
Another option, to add a query param to your url, something like: ?countOnly=true

If POST to an URL and how can I tell the client to look at a fragment of the response?

I would like to POST an entity as follows
POST /example.org/MyEntity/100
Based on the passed entity, the server would like to draw the users attention to a particular part of the response using a fragment identifier. e.g.
/example.org/MyEntity/100#InterestingPart
How do I return this new URL to the client. I am assuming I could do some form of redirect using a 3XX response code, but I actually do not want the client to do another request because the only difference between the two URLs is the fragment. At the moment it seems that a 307 return code would be the most appropriate because according to the spec you should not automatically redirect a POST.
Is there are better way?
Update: My client is not limited to the constraints of a web browser. I am just looking at this from the perspective of HTTP.
Update2: Based on my reading of RFC2616, I can see nothing stopping me from returning a 200 and a Location header that contains the fragment identifier. Anyone know of a reason why I cannot do that?
I think the only sensible solution is to have action URL have static fragment identifier, like <form method="post" action="/action#anchored"> and then put an anchor wherever you want user to look at while generating page.
But, to answer the Update2: no, there's no reason to avoid it.
My inclination is to return 201 - and have the location header point to the URI you want the client to GET.
I didn't look, but IIRC nothing dictates that the location header points to the resource created, so it should be spec legal.
You should normally redirect every POST to avoid problems with refreshing the page and the use of the back button. This is known as the PRG (POST Redirect Get) pattern:
http://blog.httpwatch.com/2007/10/03/60-of-web-users-can%E2%80%99t-be-wrong-%E2%80%93-don%E2%80%99t-break-the-back-button/
Although, this does incur the cost of another round trip to the server it makes your web application much more user friendly.
You could then add the fragment onto the redirected URL.
There's an example of PRG with a fragment on this page:
http://www.httpwatch.com/httpgallery/redirection/
POSTing to the URI:
http://example.org/MyEntity/100
implies to me that a MyEntity resource called "100" already exists. If that's the case, why not use PUT instead? Is this an update or a create operation?
An alternative might be:
POST http://example.org/MyEntities
Now your service has a choice to make from at least two possibilities:
Return 201 Created. Set the Location header to be the URI you want the client to use (e.g.: http://example.org/MyEntities/100#InterestingPart). Add the representation of the new resource to the body.
Return 204 No Content. Same as above, but no body. This option requires a subsequent GET to fetch the representation, which sounds like what you're trying to avoid.
Neither approach requires redirection and both can return as specific a URI as you desire.
I am curious though, why is the #InterestingPart significant? Why not just return the entire representation and its URI http://example.org/MyEntities/100 in the Location header - and let clients decide for themselves what's interesting or not? If the answers have something to do with only a small part of the resource being of interest (or being modified) during a request, how feasible would it be to break MyResource into a main resource and one or more subordinate resources? For example:
/MyResources/100/CoolThings
/MyResources/100/CoolThings/42
/MyResources/100/InterestingThings
/MyResources/100/InterestingThings/109

Resources