This project provides some APIs to ease creating REST representations that follow the HATEOAS principle when working with Spring and especially Spring MVC. The core problem it tries to address is link creation and representation assembly.

© 2012-2019 The original authors.

Copies of this document may be made for your own use and for distribution to others, provided that you do not charge any fee for such copies and further provided that each copy contains this Copyright Notice, whether distributed in print or electronically.

1. Preface

1.1. Migrating to Spring HATEOAS 1.0

For 1.0 we took the chance to re-evaluate some of the design and package structure choices we had made for the 0.x branch. There had been an incredible amount of feedback on it and the major version bump seemed to be the most natural place to refactor those.

1.1.1. The changes

The biggest changes in package structure were driven by the introduction of a hypermedia type registration API to support additional media types in Spring HATEOAS. This lead to the clear separation of client and server APIs (packages named respectively) as well as media type implementations in the package mediatype.

The easiest way to get your code base upgraded to the new API is by using the migration script. Before we jump to that, here are the changes at a quick glance.

Representation models

The ResourceSupport/Resource/Resources/PagedResources group of classes never really felt appropriately named. After all, these types do not actually manifest resources but rather representation models that can be enriched with hypermedia information and affordances. Here’s how new names map to the old ones:

  • ResourceSupport is now RepresentationModel

  • Resource is now EntityModel

  • Resources is now CollectionModel

  • PagedResources is now PagedModel

Consequently, ResourceAssembler has been renamed to RepresentationModelAssembler and its methods toResource(…) and toResources(…) have been renamed to toModel(…) and toCollectionModel(…) respectively. Also the name changes have been reflected in the classes contained in TypeReferences.

  • RepresentationModel.getLinks() now exposes a Links instance (over a List<Link>) as that exposes additional API to concatenate and merge different Links instances using various strategies. Also it has been turned into a self-bound generic type to allow the methods that add links to the instance return the instance itself.

  • The LinkDiscoverer API has been moved to the client package.

  • The LinkBuilder and EntityLinks APIs have been moved to the server package.

  • ControllerLinkBuilder has been moved into server.mvc and deprecated to be replaced by WebMvcLinkBuilder.

  • RelProvider has been renamed to LinkRelationProvider and returns LinkRelation instances instead of Strings.

  • VndError has been moved to the mediatype.vnderror package.

1.1.2. The migration script

You can find a script to run from your application root that will update all import statements and static method references to Spring HATEOAS types that moved in our source code repository. Simply download that, run it from your project root. By default it will inspect all Java source files and replace the legacy Spring HATEOAS type references with the new ones.

Example 1. Sample application of the migration script
$ ./

Migrating Spring HATEOAS references to 1.0 for files : *.java

Adapting ./src/main/java/…


Note that the script will not necessarily be able to entirely fix all changes, but it should cover the most important refactorings.

Now verify the changes made to the files in your favorite Git client and commit as appropriate. In case you find method or type references unmigrated, please open a ticket in out issue tracker.

1.1.3. Migrating from 1.0 M3 to 1.0 RC1

  • Link.andAffordance(…) taking Affordance details have been moved to Affordances. To manually build up Affordance instances now use Affordances.of(link).afford(…). Also note the new AffordanceBuilder type exposed from Affordances for fluent usage. See Affordances for details.

  • AffordanceModelFactory.getAffordanceModel(…) now receives InputPayloadMetadata and PayloadMetadata instances instead of ResolvableTypes to allow non-type-based implementations. Custom media type implementations have to be adapted to that accordingly.

  • HAL Forms now does not render property attributes if their value adheres to what’s defined as default in the spec. I.e. if previously required was explicitly set to false, we now just omit the entry for required. We also now only force them to be non-required for templates that use PATCH as the HTTP method.

2. Fundamentals

This section covers the basics of Spring HATEOAS and its fundamental domain abstractions.

The fundamental idea of hypermedia is to enrich the representation of a resource with hypermedia elements. The simplest form of that are links. They indicate a client that it can navigate to a certain resource. The semantics of a related resource are defined in a so-called link relation. You might have seen this in the header of an HTML file already:

Example 2. A link in an HTML document
<link href="theme.css" rel="stylesheet" type="text/css" />

As you can see the link points to a resource theme.css and indicates that it is a style sheet. Links often carry additional information, like the media type that the resource pointed to will return. However, the fundamental building blocks of a link are its reference and relation.

Spring HATEOAS lets you work with links through its immutable Link value type. Its constructor takes both a hypertext reference and a link relation, the latter being defaulted to the IANA link relation self. Read more on the latter in Link relations.

Example 3. Using links
Link link = Link.of("/something");

link = Link.of("/something", "my-rel");

Link exposes other attributes as defined in RFC-8288. You can set them by calling the corresponding wither method on a Link instance.

Find more information on how to create links pointing to Spring MVC and Spring WebFlux controllers in Building links in Spring MVC and Building links in Spring WebFlux.

2.2. URI templates

For a Spring HATEOAS Link, the hypertext reference can not only be a URI, but also a URI template according to RFC-6570. A URI template contains so-called template variables and allows expansion of these parameters. This allows clients to turn parameterized templates into URIs without having to know about the structure of the final URI, it only needs to know about the names of the variables.

Example 4. Using links with templated URIs
Link link = Link.of("/{segment}/something{?parameter}");
assertThat(link.isTemplated()).isTrue(); (1)
assertThat(link.getVariableNames()).contains("segment", "parameter"); (2)

Map<String, Object> values = new HashMap<>();
values.put("segment", "path");
values.put("parameter", 42);

assertThat(link.expand(values).getHref()) (3)
1 The Link instance indicates that is templated, i.e. it contains a URI template.
2 It exposes the parameters contained in the template.
3 It allows expansion of the parameters.

URI templates can be constructed manually and template variables added later on.

Example 5. Working with URI templates
UriTemplate template = UriTemplate.of("/{segment}/something")
  .with(new TemplateVariable("parameter", VariableType.REQUEST_PARAM);


To indicate the relationship of the target resource to the current one so-called link relations are used. Spring HATEOAS provides a LinkRelation type to easily create String-based instances of it.

The Internet Assigned Numbers Authority contains a set of predefined link relations. They can be referred to via IanaLinkRelations.

Example 6. Using IANA link relations
Link link = Link.of("/some-resource"), IanaLinkRelations.NEXT);


2.4. Representation models

To easily create hypermedia enriched representations, Spring HATEOAS provides a set of classes with RepresentationModel at their root. It’s basically a container for a collection of Links and has convenient methods to add those to the model. The models can later be rendered into various media type formats that will define how the hypermedia elements look in the representation. For more information on this, have a look at Media types.

Example 7. The RepresentationModel class hierarchy
class RepresentationModel
class EntityModel
class CollectionModel
class PagedModel

EntityModel -|> RepresentationModel
CollectionModel -|> RepresentationModel
PagedModel -|> CollectionModel

The default way to work with a RepresentationModel is to create a subclass of it to contain all the properties the representation is supposed to contain, create instances of that class, populate the properties and enrich it with links.

Example 8. A sample representation model type
class PersonModel extends RepresentationModel<PersonModel> {

  String firstname, lastname;

The generic self-typing is necessary to let RepresentationModel.add(…) return instances of itself. The model type can now be used like this:

Example 9. Using the person representation model
PersonModel model = new PersonModel();
model.firstname = "Dave";
model.lastname = "Matthews";

If you returned such an instance from a Spring MVC or WebFlux controller and the client sent an Accept header set to application/hal+json, the response would look as follows:

Example 10. The HAL representation generated for the person representation model
  "_links" : {
    "self" : {
      "href" : "https://myhost/people/42"
  "firstname" : "Dave",
  "lastname" : "Matthews"

2.4.1. Item resource representation model

For a resource that’s backed by a singular object or concept, a convenience EntityModel type exists. Instead of creating a custom model type for each concept, you can just reuse an already existing type and wrap instances of it into the EntityModel.

Example 11. Using EntityModel to wrap existing objects
Person person = new Person("Dave", "Matthews");
EntityModel<Person> model = EntityModel.of(person);

2.4.2. Collection resource representation model

For resources that are conceptually collections, a CollectionModel is available. Its elements can either be simple objects or RepresentationModel instances in turn.

Example 12. Using CollectionModel to wrap a collection of existing objects
Collection<Person> people = Collections.singleton(new Person("Dave", "Matthews"));
CollectionModel<Person> model = CollectionModel.of(people);

3. Server-side support

Now we have the domain vocabulary in place, but the main challenge remains: how to create the actual URIs to be wrapped into Link instances in a less fragile way. Right now, we would have to duplicate URI strings all over the place. Doing so is brittle and unmaintainable.

Assume you have your Spring MVC controllers implemented as follows:

class PersonController {

  HttpEntity<PersonModel> showAll() { … }

  @GetMapping(value = "/{person}", method = RequestMethod.GET)
  HttpEntity<PersonModel> show(@PathVariable Long person) { … }

We see two conventions here. The first is a collection resource that is exposed through @GetMapping annotation of the controller method, with individual elements of that collection exposed as direct sub resources. The collection resource might be exposed at a simple URI (as just shown) or more complex ones (such as /people/{id}/addresses). Suppose you would like to link to the collection resource of all people. Following the approach from above would cause two problems:

  • To create an absolute URI, you would need to look up the protocol, hostname, port, servlet base, and other values. This is cumbersome and requires ugly manual string concatenation code.

  • You probably do not want to concatenate the /people on top of your base URI, because you would then have to maintain the information in multiple places. If you change the mapping, you then have to change all the clients pointing to it.

Spring HATEOAS now provides a WebMvcLinkBuilder that lets you create links by pointing to controller classes. The following example shows how to do so:

import static org.sfw.hateoas.server.mvc.WebMvcLinkBuilder.*;

Link link = linkTo(PersonController.class).withRel("people");


The WebMvcLinkBuilder uses Spring’s ServletUriComponentsBuilder under the hood to obtain the basic URI information from the current request. Assuming your application runs at localhost:8080/your-app, this is exactly the URI on top of which you are constructing additional parts. The builder now inspects the given controller class for its root mapping and thus ends up with localhost:8080/your-app/people. You can also build more nested links as well. The following example shows how to do so:

Person person = new Person(1L, "Dave", "Matthews");
//                 /person                 /     1
Link link = linkTo(PersonController.class).slash(person.getId()).withSelfRel();
assertThat(link.getRel(), is(IanaLinkRelation.SELF.value()));
assertThat(link.getHref(), endsWith("/people/1"));

The builder also allows creating URI instances to build up (for example, response header values):

HttpHeaders headers = new HttpHeaders();

return new ResponseEntity<PersonModel>(headers, HttpStatus.CREATED);

3.1.1. Building links that point to methods

You can even build links that point to methods or create dummy controller method invocations. The first approach is to hand a Method instance to the WebMvcLinkBuilder. The following example shows how to do so:

Method method = PersonController.class.getMethod("show", Long.class);
Link link = linkTo(method, 2L).withSelfRel();


This is still a bit dissatisfying, as we have to first get a Method instance, which throws an exception and is generally quite cumbersome. At least we do not repeat the mapping. An even better approach is to have a dummy method invocation of the target method on a controller proxy, which we can create by using the methodOn(…) helper. The following example shows how to do so:

Link link = linkTo(methodOn(PersonController.class).show(2L)).withSelfRel();


methodOn(…) creates a proxy of the controller class that records the method invocation and exposes it in a proxy created for the return type of the method. This allows the fluent expression of the method for which we want to obtain the mapping. However, there are a few constraints on the methods that can be obtained by using this technique:

  • The return type has to be capable of proxying, as we need to expose the method invocation on it.

  • The parameters handed into the methods are generally neglected (except the ones referred to through @PathVariable, because they make up the URI).


3.3. Affordances

The affordances of the environment are what it offers …​ what it provides or furnishes, either for good or ill. The verb 'to afford' is found in the dictionary, but the noun 'affordance' is not. I have made it up.

— James J. Gibson
The Ecological Approach to Visual Perception (page 126)

REST-based resources provide not just data but controls. The last ingredient to form a flexible service are detailed affordances on how to use the various controls. Because affordances are associated with links, Spring HATEOAS provides an API to attach as many related methods as needed to a link. Just as you can create links by pointing to Spring MVC controller methods (see Building links in Spring MVC for details) you …​

The following code shows how to take a self link and associate two more affordances:

Example 13. Connecting affordances to GET /employees/{id}
public EntityModel<Employee> findOne(@PathVariable Integer id) {

  Class<EmployeeController> controllerClass = EmployeeController.class;

  // Start the affordance with the "self" link, i.e. this method.
  Link findOneLink = linkTo(methodOn(controllerClass).findOne(id)).withSelfRel(); (1)

  // Return the affordance + a link back to the entire collection resource.
  return EntityModel.of(EMPLOYEES.get(id), //
      findOneLink //
          .andAffordance(afford(methodOn(controllerClass).updateEmployee(null, id))) (2)
          .andAffordance(afford(methodOn(controllerClass).partiallyUpdateEmployee(null, id)))); (3)
1 Create the self link.
2 Associate the updateEmployee method with the self link.
3 Associate the partiallyUpdateEmployee method with the self link.

Using .andAffordance(afford(…​)), you can use the controller’s methods to connect a PUT and a PATCH operation to a GET operation. Imagine that the related methods afforded above look like this:

Example 14. updateEmpoyee method that responds to PUT /employees/{id}
public ResponseEntity<?> updateEmployee( //
    @RequestBody EntityModel<Employee> employee, @PathVariable Integer id)
Example 15. partiallyUpdateEmployee method that responds to PATCH /employees/{id}
public ResponseEntity<?> partiallyUpdateEmployee( //
    @RequestBody EntityModel<Employee> employee, @PathVariable Integer id)

Pointing to those methods using the afford(…) methods will cause Spring HATEOAS to analyze the request body and response types and capture metadata to allow different media type implementations to use that information to translate that into descriptions of the input and outputs.

3.3.1. Building affordances manually

While the primary way to register affordances for a link, it might be necessary to build some of them manually. This can be achieved by using the Affordances API:

Example 16. Using the Affordances API to manually register affordances
var methodInvocation = methodOn(EmployeeController.class).all();

var link = Affordances.of(linkTo(methodInvocation).withSelfRel()) (1)

    .afford(HttpMethod.POST) (2)
    .withInputAndOutput(Employee.class) //
    .withName("createEmployee") //

    .andAfford(HttpMethod.GET) (3)
    .withOutput(Employee.class) //
        QueryParameter.optional("name"), //
        QueryParameter.optional("role")) //
    .withName("search") //

1 You start by creating an instance of Affordances from a Link instance creating the context for describing the affordances.
2 Each affordance starts with the HTTP method it’s supposed to support. We then register a type as payload description and name the affordance explicitly. The latter can be omitted and a default name will be derived from the HTTP method and input type name. This effectively creates the same affordance as the pointer to EmployeeController.newEmployee(…) created.
3 The next affordance is built to reflect what’s happening for the pointer to…). Here we define Employee to be the model for the response created and explicitly register QueryParameters.

Affordances are backed by media type specific affordance models that translate the general affordance metadata into specific representations. Please make sure to check the section on affordances in the Media types section to find more details about how to control the exposure of that metadata.

RFC-7239 forwarding headers are most commonly used when your application is behind a proxy, behind a load balancer, or in the cloud. The node that actually receives the web request is part of the infrastructure, and forwards the request to your application.

Your application may be running on localhost:8080, but to the outside world you’re expected to be at (and on the web’s standard port 80). By having the proxy include extra headers (which many already do), Spring HATEOAS can generate links properly as it uses Spring Framework functionality to obtain the base URI of the original request.

Anything that can change the root URI based on external inputs must be properly guarded. That’s why, by default, forwarded header handling is disabled. You MUST enable it to be operational. If you are deploying to the cloud or into a configuration where you control the proxies and load balancers, then you’ll certainly want to use this feature.

To enable forwarded header handling you need to register Spring’s ForwardedHeaderFilter for Spring MVC (details here) or ForwardedHeaderTransformer for Spring WebFlux (details here) in your application. In a Spring Boot application those components can be simply declared as Spring beans as described here.

Example 17. Registering a ForwardedHeaderFilter
ForwardedHeaderFilter forwardedHeaderFilter() {
    return new ForwardedHeaderFilter();

This will create a servlet filter that processes all the X-Forwarded-… headers. And it will register it properly with the servlet handlers.

For a Spring WebFlux application, the reactive counterpart is ForwardedHeaderTransformer:

Example 18. Registering a ForwardedHeaderTransformer
ForwardedHeaderTransformer forwardedHeaderTransformer() {
    return new ForwardedHeaderTransformer();

This will create a function that transforms reactive web requests, processing X-Forwarded-… headers. And it will register it properly with WebFlux.

With configuration as shown above in place, a request passing X-Forwarded-… headers will see those reflected in the links generated:

Example 19. A request using X-Forwarded-… headers
curl -v localhost:8080/employees \
    -H 'X-Forwarded-Proto: https' \
    -H 'X-Forwarded-Host:' \
    -H 'X-Forwarded-Port: 9001'
Example 20. The corresponding response with the links generated to consider those headers
  "_embedded": {
    "employees": [
        "id": 1,
        "name": "Bilbo Baggins",
        "role": "burglar",
        "_links": {
          "self": {
            "href": ""
          "employees": {
            "href": ""
  "_links": {
    "self": {
      "href": ""
    "root": {
      "href": ""
EntityLinks and its various implementations are NOT currently provided out-of-the-box for Spring WebFlux applications. The contract defined in the EntityLinks SPI was originally aimed at Spring Web MVC and doesn’t consider Reactor types. Developing a comparable contract that supports reactive programming is still in progress.

So far, we have created links by pointing to the web framework implementations (that is, the Spring MVC controllers) and inspected the mapping. In many cases, these classes essentially read and write representations backed by a model class.

The EntityLinks interface now exposes an API to look up a Link or LinkBuilder based on the model types. The methods essentially return links that point either to the collection resource (such as /people) or to an item resource (such as /people/1). The following example shows how to use EntityLinks:

EntityLinks links = …;
LinkBuilder builder = links.linkFor(Customer.class);
Link link = links.linkToItemResource(Customer.class, 1L);

EntityLinks is available via dependency injection by activating @EnableHypermediaSupport in your Spring MVC configuration. This will cause a variety of default implementations of EntityLinks being registered. The most fundamental one is ControllerEntityLinks that inspects SpringMVC controller classes. If you want to register your own implementation of EntityLinks, check out this section.

3.5.1. EntityLinks based on Spring MVC controllers

Activating entity links functionality causes all the Spring MVC controllers available in the current ApplicationContext to be inspected for the @ExposesResourceFor(…) annotation. The annotation exposes which model type the controller manages. Beyond that, we assume that you adhere to the following URI mapping setup and conventions:

  • A type level @ExposesResourceFor(…) declaring which entity type the controller exposes collection and item resources for.

  • A class level base mapping that represents the collection resource.

  • An additional method level mapping that extends the mapping to append an identifier as additional path segment.

The following example shows an implementation of an EntityLinks-capable controller:

@ExposesResourceFor(Order.class) (1)
@RequestMapping("/orders") (2)
class OrderController {

  @GetMapping (3)
  ResponseEntity orders(…) { … }

  @GetMapping("{id}") (4)
  ResponseEntity order(@PathVariable("id") … ) { … }
1 The controller indicates it’s exposing collection and item resources for the entity Order.
2 Its collection resource is exposed under /orders
3 That collection resource can handle GET requests. Add more methods for other HTTP methods at your convenience.
4 An additional controller method to handle a subordinate resource taking a path variable to expose an item resource, i.e. a single Order.

With this in place, when you enable EntityLinks @EnableHypermediaSupport in your Spring MVC configuration, you can create links to the controller as follows:

class PaymentController {

  private final EntityLinks entityLinks;

  PaymentController(EntityLinks entityLinks) { (1)
    this.entityLinks = entityLinks;

  ResponseEntity payment(@PathVariable Long orderId) {

    Link link = entityLinks.linkToItemResource(Order.class, orderId); (2)
1 Inject EntityLinks made available by @EnableHypermediaSupport in your configuration.
2 Use the APIs to build links by using the entity types instead of controller classes.

As you can see, you can refer to resources managing Order instances without referring to OrderController explicitly.

3.5.2. EntityLinks API in detail

Fundamentally, EntityLinks allows to build LinkBuilders and Link instances to collection and item resources of an entity type. Methods starting with linkFor… will produce LinkBuilder instances for you to extend and augment with additional path segments, parameters, etc. Methods starting with linkTo produce fully prepared Link instances.

While for collection resources providing an entity type is sufficient, links to item resources will need an identifier provided. This usually looks like this:

Example 21. Obtaining a link to an item resource
entityLinks.linkToItemResource(order, order.getId());

If you find yourself repeating those method calls the identifier extraction step can be pulled out into a reusable Function to be reused throughout different invocations:

Function<Order, Object> idExtractor = Order::getId; (1)

entityLinks.linkToItemResource(order, idExtractor); (2)
1 The identifier extraction is externalized so that it can be held in a field or constant.
2 The link lookup using the extractor.

As controller implementations are often grouped around entity types, you’ll very often find yourself using the same extractor function (see EntityLinks API in detail for details) all over the controller class. We can centralize the identifier extraction logic even more by obtaining a TypedEntityLinks instance providing the extractor once, so that the actual lookups don’t have to deal with the extraction anymore at all.

Example 22. Using TypedEntityLinks
class OrderController {

  private final TypedEntityLinks<Order> links;

  OrderController(EntityLinks entityLinks) { (1)
    this.links = entityLinks.forType(Order::getId); (2)

  ResponseEntity<Order> someMethod(…) {

    Order order = … // lookup order

    Link link = links.linkToItemResource(order); (3)
1 Inject an EntityLinks instance.
2 Indicate you’re going to look up Order instances with a certain identifier extractor function.
3 Look up item resource links based on a sole Order instance.

3.5.3. EntityLinks as SPI

The EntityLinks instance created by @EnableHypermediaSupport is of type DelegatingEntityLinks which will in turn pick up all other EntityLinks implementations available as beans in the ApplicationContext. It’s registered as primary bean so that it’s always the sole injection candidate when you inject EntityLinks in general. ControllerEntityLinks is the default implementation that will be included in the setup, but users are free to implement and register their own implementations. Making those available to the EntityLinks instance available for injection is a matter of registering your implementation as Spring bean.

Example 23. Declaring a custom EntityLinks implementation
class CustomEntityLinksConfiguration {

  MyEntityLinks myEntityLinks(…) {
    return new MyEntityLinks(…);

An example for the extensibility of this mechanism is Spring Data REST’s RepositoryEntityLinks, which uses the repository mapping information to create links pointing to resources backed by Spring Data repositories. At the same time, it even exposes additional lookup methods for other types of resources. If you want to make use of these, simply inject RepositoryEntityLinks explicitly.

3.6. Representation model assembler

As the mapping from an entity to a representation model must be used in multiple places, it makes sense to create a dedicated class responsible for doing so. The conversion contains very custom steps but also a few boilerplate steps:

  1. Instantiation of the model class

  2. Adding a link with a rel of self pointing to the resource that gets rendered.

Spring HATEOAS now provides a RepresentationModelAssemblerSupport base class that helps reduce the amount of code you need to write. The following example shows how to use it:

class PersonModelAssembler extends RepresentationModelAssemblerSupport<Person, PersonModel> {

  public PersonModelAssembler() {
    super(PersonController.class, PersonModel.class);

  public PersonModel toModel(Person person) {

    PersonModel resource = createResource(person);
    // … do further mapping
    return resource;
createResource(…​) is code you write to instantiate a PersonModel object given a Person object. It should only focus on setting attributes, not populating Links.

Setting the class up as we did in the preceding example gives you the following benefits:

  • There are a handful of createModelWithId(…) methods that let you create an instance of the resource and have a Link with a rel of self added to it. The href of that link is determined by the configured controller’s request mapping plus the ID of the entity (for example, /people/1).

  • The resource type gets instantiated by reflection and expects a no-arg constructor. If you want to use a dedicated constructor or avoid the reflection performance overhead, you can override instantiateModel(…).

You can then use the assembler to either assemble a RepresentationModel or a CollectionModel. The following example creates a CollectionModel of PersonModel instances:

Person person = new Person(…);
Iterable<Person> people = Collections.singletonList(person);

PersonModelAssembler assembler = new PersonModelAssembler();
PersonModel model = assembler.toModel(person);
CollectionModel<PersonModel> model = assembler.toCollectionModel(people);

3.7. Representation Model Processors

Sometimes you need to tweak and adjust hypermedia representations after they have been assembled.

A perfect example is when you have a controller that deals with order fulfillment, but you need to add links related to making payments.

Imagine having your ordering system producing this type of hypermedia:

  "orderId" : "42",
  "state" : "AWAITING_PAYMENT",
  "_links" : {
    "self" : {
      "href" : "http://localhost/orders/999"

You wish to add a link so the client can make payment, but don’t want to mix details about your PaymentController into the OrderController.

Instead of polluting the details of your ordering system, you can write a RepresentationModelProcessor like this:

public class PaymentProcessor implements RepresentationModelProcessor<EntityModel<Order>> { (1)

  public EntityModel<Order> process(EntityModel<Order> model) {

    model.add( (2)
        Link.of("/payments/{orderId}").withRel(LinkRelation.of("payments")) //

    return model; (3)
1 This processor will only be applied to EntityModel<Order> objects.
2 Manipulate the existing EntityModel object by adding an unconditional link.
3 Return the EntityModel so it can be serialized into the requested media type.

Register the processor with your application:

public class PaymentProcessingApp {

  PaymentProcessor paymentProcessor() {
    return new PaymentProcessor();

Now when you issue a hypermedia respresentation of an Order, the client receives this:

  "orderId" : "42",
  "state" : "AWAITING_PAYMENT",
  "_links" : {
    "self" : {
      "href" : "http://localhost/orders/999"
    "payments" : { (1)
      "href" : "/payments/42" (2)
1 You see the LinkRelation.of("payments") plugged in as this link’s relation.
2 The URI was provided by the processor.

This example is quite simple, but you can easily:

  • Use WebMvcLinkBuilder or WebFluxLinkBuilder to construct a dynamic link to your PaymentController.

  • Inject any services needed to conditionally add other links (e.g. cancel, amend) that are driven by state.

  • Leverage cross cutting services like Spring Security to add, remove, or revise links based upon the current user’s context.

Also, in this example, the PaymentProcessor alters the provided EntityModel<Order>. You also have the power to replace it with another object. Just be advised the API requires the return type to equal the input type.

3.8. Using the LinkRelationProvider API

When building links, you usually need to determine the relation type to be used for the link. In most cases, the relation type is directly associated with a (domain) type. We encapsulate the detailed algorithm to look up the relation types behind a LinkRelationProvider API that lets you determine the relation types for single and collection resources. The algorithm for looking up the relation type follows:

  1. If the type is annotated with @Relation, we use the values configured in the annotation.

  2. If not, we default to the uncapitalized simple class name plus an appended List for the collection rel.

  3. If the EVO inflector JAR is in the classpath, we use the plural of the single resource rel provided by the pluralizing algorithm.

  4. @Controller classes annotated with @ExposesResourceFor (see Using the EntityLinks interface for details) transparently look up the relation types for the type configured in the annotation, so that you can use LinkRelationProvider.getItemResourceRelFor(MyController.class) and get the relation type of the domain type exposed.

A LinkRelationProvider is automatically exposed as a Spring bean when you use @EnableHypermediaSupport. You can plug in custom providers by implementing the interface and exposing them as Spring beans in turn.

4. Media types

4.1. HAL – Hypertext Application Language

JSON Hypertext Application Language or HAL is one of the simplest and most widely adopted hypermedia media types adopted when not discussing specific web stacks.

It was the first spec-based media type adopted by Spring HATEOAS.

4.1.1. Building HAL representation models

As of Spring HATEOAS 1.1, we ship a dedicated HalModelBuilder that allows to create RepresentationModel instances through a HAL-idiomatic API. These are its fundamental assumptions:

  1. A HAL representation can be backed by an arbitrary object (an entity) that builds up the domain fields contained in the representation.

  2. The representation can be enriched by a variety of embedded documents, which can be either arbitrary objects or HAL representations themselves (i.e. containing nested embeddeds and links).

  3. Certain HAL specific patterns (e.g. previews) can be directly used in the API so that the code setting up the representation reads like you’d describe a HAL representation following those idioms.

Here’s an example of the API used:

// An order
var order = new Order(…); (1)

// The customer who placed the order
var customer = customer.findById(order.getCustomerId());

var customerLink = Link.of("/orders/{id}/customer") (2)

var additional = …

var model = HalModelBuilder.halModel(order)
  .preview(new CustomerSummary(customer)) (3)
  .forLink(customerLink) (4)
  .embed(additional) (5)
  .link(Link.of(…, IanaLinkRelations.SELF));
1 We set up some domain type. In this case, an order that has a relationship to the customer that placed it.
2 We prepare a link pointing to a resource that will expose customer details
3 We start building a preview by providing the payload that’s supposed to be rendered inside the _embeddable clause.
4 We conclude that preview by providing the target link. It transparently gets added to the _links object and its link relation is used as the key for the object provided in the previous step.
5 Other objects can be added to show up under _embedded. The key under which they’re listed is derived from the objects relation settings. They’re customizable via @Relation or a dedicated LinkRelationProvider (see Using the LinkRelationProvider API for details).
  "_links" : {
    "self" : { "href" : "…" }, (1)
    "customer" : { "href" : "/orders/4711/customer" } (2)
  "_embedded" : {
    "customer" : { … }, (3)
    "additional" : { … } (4)
1 The self link as explicitly provided.
2 The customer link transparently added through ….preview(…).forLink(…).
3 The preview object provided.
4 Additional elements added via explicit ….embed(…).

In HAL _embedded is also used to represent top collections. They’re usually grouped under the link relation derived from the object’s type. I.e. a list of orders would look like this in HAL:

  "_embedded" : {
    "orders : [
      … (1)
1 Individual order documents go here.

Creating such a representation is as easy as this:

Collection<Order> orders = …;


That said, if the order is empty, there’s no way to derive the link relation to appear inside _embedded, so that the document will stay empty if the collection is empty.

If you prefer to explicitly communicate an empty collection, a type can be handed into the overload of the ….embed(…) method taking a Collection. If the collection handed into the method is empty, this will cause a field rendered with its link relation derived from the given type.

  .embed(Collections.emptyList(), Order.class);
  // or
  .embed(Collections.emptyList(), LinkRelation.of("orders"));

will create the following, more explicit representation.

  "_embedded" : {
    "orders" : []

4.1.2. Configuring link rendering

In HAL, the _links entry is a JSON object. The property names are link relations and each value is either a link object or an array of link objects.

For a given link relation that has two or more links, the spec is clear on representation:

Example 24. HAL document with two links associated with one relation
  "_links": {
    "item": [
      { "href": "https://myhost/cart/42" },
      { "href": "https://myhost/inventory/12" }
  "customer": "Dave Matthews"

But if there is only one link for a given relation, the spec is ambiguous. You could render that as either a single object or as a single-item array.

By default, Spring HATEOAS uses the most terse approach and renders a single-link relation like this:

Example 25. HAL document with single link rendered as an object
  "_links": {
    "item": { "href": "https://myhost/inventory/12" }
  "customer": "Dave Matthews"

Some users prefer to not switch between arrays and objects when consuming HAL. They would prefer this type of rendering:

Example 26. HAL with single link rendered as an array
  "_links": {
    "item": [{ "href": "https://myhost/inventory/12" }]
  "customer": "Dave Matthews"

If you wish to customize this policy, all you have to do is inject a HalConfiguration bean into your application configuration. There are multiple choices.

Example 27. Global HAL single-link rendering policy
public HalConfiguration globalPolicy() {
  return new HalConfiguration() //
      .withRenderSingleLinks(RenderSingleLinks.AS_ARRAY); (1)
1 Override Spring HATEOAS’s default by rendering ALL single-link relations as arrays.

If you prefer to only override some particular link relations, you can create a HalConfiguration bean like this:

Example 28. Link relation-based HAL single-link rendering policy
public HalConfiguration linkRelationBasedPolicy() {
  return new HalConfiguration() //
      .withRenderSingleLinksFor( //
          IanaLinkRelations.ITEM, RenderSingleLinks.AS_ARRAY) (1)
      .withRenderSingleLinksFor( //
          LinkRelation.of("prev"), RenderSingleLinks.AS_SINGLE); (2)
1 Always render item link relations as an array.
2 Render prev link relations as an object when there is only one link.

If neither of these match your needs, you can use an Ant-style path pattern:

Example 29. Pattern-based HAL single-link rendering policy
public HalConfiguration patternBasedPolicy() {
  return new HalConfiguration() //
      .withRenderSingleLinksFor( //
          "http*", RenderSingleLinks.AS_ARRAY); (1)
1 Render all link relations that start with http as an array.
The pattern-based approach uses Spring’s AntPathMatcher.

All of these HalConfiguration withers can be combined to form one comprehensive policy. Be sure to test your API extensively to avoid surprises.

4.1.3. Link title internationalization

HAL defines a title attribute for its link objects. These titles can be populated by using Spring’s resource bundle abstraction and a resource bundle named rest-messages so that clients can use them in their UIs directly. This bundle will be set up automatically and is used during HAL link serialization.

To define a title for a link, use the key template _links.$relationName.title as follows:

Example 30. A sample
_links.cancel.title=Cancel order
_links.payment.title=Proceed to checkout

This will result in the following HAL representation:

Example 31. A sample HAL document with link titles defined
  "_links" : {
    "cancel" : {
      "href" : "…"
      "title" : "Cancel order"
    "payment" : {
      "href" : "…"
      "title" : "Proceed to checkout"

4.1.4. Using the CurieProvider API

The Web Linking RFC describes registered and extension link relation types. Registered rels are well-known strings registered with the IANA registry of link relation types. Extension rel URIs can be used by applications that do not wish to register a relation type. Each one is a URI that uniquely identifies the relation type. The rel URI can be serialized as a compact URI or Curie. For example, a curie of ex:persons stands for the link relation type if ex is defined as{rel}. If curies are used, the base URI must be present in the response scope.

The rel values created by the default RelProvider are extension relation types and, as a result, must be URIs, which can cause a lot of overhead. The CurieProvider API takes care of that: It lets you define a base URI as a URI template and a prefix that stands for that base URI. If a CurieProvider is present, the RelProvider prepends all rel values with the curie prefix. Furthermore a curies link is automatically added to the HAL resource.

The following configuration defines a default curie provider:

@EnableHypermediaSupport(type= {HypermediaType.HAL})
public class Config {

  public CurieProvider curieProvider() {
    return new DefaultCurieProvider("ex", new UriTemplate("{rel}"));

Note that now the ex: prefix automatically appears before all rel values that are not registered with IANA, as in ex:orders. Clients can use the curies link to resolve a curie to its full form. The following example shows how to do so:

  "_links": {
    "self": {
      "href": "https://myhost/person/1"
    "curies": {
      "name": "ex",
      "href": "{rel}",
      "templated": true
    "ex:orders": {
      "href": "https://myhost/person/1/orders"
  "firstname": "Dave",
  "lastname": "Matthews"

Since the purpose of the CurieProvider API is to allow for automatic curie creation, you can define only one CurieProvider bean per application scope.


HAL-FORMS is designed to add runtime FORM support to the HAL media type.

HAL-FORMS "looks like HAL." However, it is important to keep in mind that HAL-FORMS is not the same as HAL — the two should not be thought of as interchangeable in any way.

— Mike Amundsen

To enable this media type, put the following configuration in your code:

Example 32. HAL-FORMS enabled application
@EnableHypermediaSupport(type = HypermediaType.HAL_FORMS)
public class HalFormsApplication {


Anytime a client supplies an Accept header with application/prs.hal-forms+json, you can expect something like this:

Example 33. HAL-FORMS sample document
  "firstName" : "Frodo",
  "lastName" : "Baggins",
  "role" : "ring bearer",
  "_links" : {
    "self" : {
      "href" : "http://localhost:8080/employees/1"
  "_templates" : {
    "default" : {
      "method" : "put",
      "properties" : [ {
        "name" : "firstName",
        "required" : true
      }, {
        "name" : "lastName",
        "required" : true
      }, {
        "name" : "role",
        "required" : true
      } ]
    "partiallyUpdateEmployee" : {
      "method" : "patch",
      "properties" : [ {
        "name" : "firstName",
        "required" : false
      }, {
        "name" : "lastName",
        "required" : false
      }, {
        "name" : "role",
        "required" : false
      } ]

Check out the HAL-FORMS spec to understand the details of the _templates attribute. Read about the Affordances API to augment your controllers with this extra metadata.

As for single-item (EntityModel) and aggregate root collections (CollectionModel), Spring HATEOAS renders them identically to HAL documents.

4.2.1. Defining HAL-FORMS metadata

HAL-FORMS allows to describe criterias for each form field. Spring HATEOAS allows to customize those by shaping the model type for the input and output types and using annotations on them.

Each template will get the following attributes defined:

Table 1. Template attributes
Attribute Description


The media type expected to be received by the server. Only included if the controller method pointed to exposes a @RequestMapping(consumes = "…") attribute, or the media type was defined explicitly when setting up the affordance.


The HTTP method to use when submitting the template.


The target URI to submit the form to. Will only be rendered if the affordance target is different than the link it was declared on.


The human readable title when displaying the template.


All properties to be submitted with the form (see below).

Each property will get the following attributes defined:

Table 2. Property attributes
Attribute Description


Set to true if there’s no setter method for the property. If that is present, use Jackson’s @JsonProperty(Access.READ_ONLY) on the accessors or field explicitly. Not rendered by default, thus defaulting to false.


Can be customized by using JSR-303’s @Pattern annotation either on the field or a type. In case of the latter the pattern will be used for every property declared as that particular type. Not rendered by default.


Can be customized by using JSR-303’s @NotNull. Not rendered by default and thus defaulting to false. Templates using PATCH as method will automatically have all properties set to not required.


The maximum value allowed for the property. Derived from Hibernate Validator’s @Range or JSR-303’s @Max annotations.


The maximum length value allowed for the property. Derived from Hibernate Validator’s @Length annotation.


The minimum value allowed for the property. Derived from Hibernate Validator’s @Range or JSR-303’s @Min annotations.


The minimum length value allowed for the property. Derived from Hibernate Validator’s @Length annotation.


The options to select a value from when submitting the form. For details, see Defining HAL-FORMS options for a property.


The user readable prompt to use when rendering the form input. For details, see Property prompts.


A user readable placeholder, to give an example for a format expected. The way of defining those follows Property prompts but uses the suffix _placeholder.


The HTML input type derived from the explicit @InputType annotation, JSR-303 validation annotations or the property’s type.

For types that you cannot annotate manually, you can register a custom pattern via a HalFormsConfiguration bean present in the application context.

class CustomConfiguration {

  HalFormsConfiguration halFormsConfiguration() {

    HalFormsConfiguration configuration = new HalFormsConfiguration();
    configuration.registerPatternFor(CreditCardNumber.class, "[0-9]{16}");

This setup will cause the HAL-FORMS template properties for representation model properties of type CreditCardNumber to declare a regex field with value [0-9]{16}.

Defining HAL-FORMS options for a property

For properties whose value is supposed to match a certain superset of values, HAL-FORMS defines the options sub-document within a property definition. Options available for a certain property can be described via HalFormsConfiguration's withOptions(…) taking a pointer to a type’s property and a creator function to turn a PropertyMetadata into a HalFormsOptions instance.

class CustomConfiguration {

  HalFormsConfiguration halFormsConfiguration() {

    HalFormsConfiguration configuration = new HalFormsConfiguration();
    configuration.withOptions(Order.class, "shippingMethod" metadata ->
      HalFormsOptions.inline("FedEx", "DHL"));

See how we set up the option values FedEx and DHL as the options to select from for the Order.shippingMethod property. Alternatively, HalFormsOptions.remote(…) can point to a remote resource providing values dynamically. Fore more constraints on options settings, refer to the spec or the Javadoc of HalFormsOptions.

4.2.2. Internationalization of form attributes

HAL-FORMS contains attributes that are intended for human interpretation, like a template’s title or property prompts. These can be defined and internationalized using Spring’s resource bundle support and the rest-messages resource bundle configured by Spring HATEOAS by default.

Template titles

To define a template title use the following pattern: _templates.$affordanceName.title. Note that in HAL-FORMS, the name of a template is default if it is the only one. This means that you’ll usually have to qualify the key with the local or fully qualified input type name that affordance describes.

Example 34. Defining HAL-FORMS template titles
_templates.default.title=Some title (1)
_templates.putEmployee.title=Create employee (2)
Employee._templates.default.title=Create employee (3)
com.acme.Employee._templates.default.title=Create employee (4)
1 A global definition for the title using default as key.
2 A global definition for the title using the actual affordance name as key. Unless defined explicitly when creating the affordance, this defaults to the name of the method that has been pointed to when creating the affordance.
3 A locally defined title to be applied to all types named Employee.
4 A title definition using the fully-qualified type name.
Keys using the actual affordance name enjoy preference over the defaulted ones.
Property prompts

Property prompts can also be resolved via the rest-messages resource bundle automatically configured by Spring HATEOAS. The keys can be defined globally, locally or fully-qualified and need an ._prompt concatenated to the actual property key:

Example 35. Defining prompts for an email property
firstName._prompt=Firstname (1)
Employee.firstName._prompt=Firstname (2)
com.acme.Employee.firstName._prompt=Firstname (3)
1 All properties named firstName will get "Firstname" rendered, independent of the type they’re declared in.
2 The firstName property in types named Employee will be prompted "Firstname".
3 The firstName property of com.acme.Employee will get a prompt of "Firstname" assigned.

4.2.3. A complete example

Let’s have a look at some example code that combines all the definition and customization attributes described above. A RepresentationModel for a customer might look something like this:

class CustomerRepresentation
  extends RepresentationModel<CustomerRepresentation> {

  String name;
  LocalDate birthdate; (1)
  @Pattern(regex = "[0-9]{16}") String ccn; (2)
  @Email String email; (3)
1 We define a birthdate property of type LocalDate.
2 We expect ccn to adhere to a regular expression.
3 We define email to be an email using the JSR-303 @Email annotation.

Note that this type is not a domain type. It’s intentionally designed to capture a wide range of potentially invalid input so that potentialy erroneous valies for the fields can be rejected at once.

Let’s continue by having a look at how a controller makes use of that model:

class CustomerController {

  EntityModel<?> createCustomer(@RequestBody CustomerRepresentation payload) { (1)
    // …

  CollectionModel<?> getCustomers() {

    CollectionModel<?> model = …;

    CustomerController controller = methodOn(CustomerController.class);

    model.add(linkTo(controller.getCustomers()).withSelfRel() (2)

    return ResponseEntity.ok(model);
1 A controller method is declared to use the representation model defined above to bind the request body to if a POST is issued to /customers.
2 A GET request to /customers prepares a model, adds a self link to it and additionally declares an affordance on that very link pointing to the controller method mapped to POST. This will cause an affordance model to be built up, which — depending on the media type to be rendered eventually — will be translated into the media type specific format.

Next, let’s add some additional metadata to make the form more accessible to humans:

Additional properties declared in
CustomerRepresentation._template.createCustomer.title=Create customer (1)
CustomerRepresentation.ccn._prompt=Credit card number (2)
CustomerRepresentation.ccn._placeholder=1234123412341234 (2)
1 We define an explicit title for the template created by pointing to the createCustomer(…) method.
2 We explicitly a prompt and placeholder for the ccn property of the CustomerRepresentation model.

If a client now issues a GET request to /customers using an Accept header of application/prs.hal-forms+json, the response HAL document is extended to a HAL-FORMS one to include the following _templates definition:

  "_templates" : {
    "default" : { (1)
      "title" : "Create customer", (2)
      "method" : "post", (3)
      "properties" : [ {
        "name" : "name",
        "required" : true,
        "type" : "text" (4)
      } , {
        "name" : "birthdate",
        "required" : true,
        "type" : "date" (4)
      } , {
        "name" : "ccn",
        "prompt" : "Credit card number", (5)
        "placeholder" : "1234123412341234" (5)
        "required" : true,
        "regex" : "[0-9]{16}", (6)
        "type" : "text"
      } , {
        "name" : "email",
        "prompt" : "Email",
        "required" : true,
        "type" : "email" (7)
      } ]
1 A template named default is exposed. Its name is default as it’s the sole template defined and the spec requires that name to be used. If multiple templates are attached (by declaring additional affordances) they will be each named after the method they’re pointing to.
2 The template title is derived from the value defined in the resource bundle. Note, that depending on the Accept-Language header sent with the request and the availability different values might returned.
3 The method attribute’s value is derived from the mapping of the method the affordance was derived from.
4 The type attribute’s value text is derived from the property’s type String. The same applies to birthdate property, but resulting in date.
5 The prompt and placeholder for the ccn property are derived from the resource bundle as well.
6 The @Pattern declaration for the ccn property is exposed as regex attribute of the template property.
7 The @Email annotation on the email property has been translated into the corresponding type value.

HAL-FORMS templates are considered by e.g. the HAL Explorer, which automatically renders HTML forms from those descriptions.

4.3. HTTP Problem Details

Problem Details for HTTP APIs is a media type to carry machine-readable details of errors in a HTTP response to avoid the need to define new error response formats for HTTP APIs.

HTTP Problem Details defines a set of JSON properties that carry additional information to describe error details to HTTP clients. Find more details about those properties in particular in the relevant section of the RFC document.

You can create such a JSON response by using the Problem media type domain type in your Spring MVC Controller:

Reporting problem details using Spring HATEOAS' Problem type
class PaymentController {

  ResponseEntity<?> issuePayment(@RequestBody PaymentRequest request) {

    PaymentResult result = payments.issuePayment(request.orderId, request.amount);

    if (result.isSuccess()) {
      return ResponseEntity.ok(result);

    String title = messages.getMessage("payment.out-of-credit");
    String detail = messages.getMessage("payment.out-of-credit.details", //
        new Object[] { result.getBalance(), result.getCost() });

    Problem problem = Problem.create() (1)
        .withType(OUT_OF_CREDIT_URI) //
        .withTitle(title) (2)
        .withDetail(detail) //
        .withInstance(PAYMENT_ERROR_INSTANCE.expand(result.getPaymentId())) //
        .withProperties(map -> { (3)
          map.put("balance", result.getBalance());
          map.put("accounts", Arrays.asList( //
              ACCOUNTS.expand(result.getSourceAccountId()), //
              ACCOUNTS.expand(result.getTargetAccountId()) //

    return ResponseEntity.status(HttpStatus.FORBIDDEN) //
1 You start by creating an instance of Problem using the factory methods exposed.
2 You can define the values for the default properties defined by the media type, e.g. the type URI, the title and details using internationalization features of Spring (see above).
3 Custom properties can be added via a Map or an explicit object (see below).

To use a dedicated object for custom properties, declare a type, create and populate an instance of it and hand this into the Problem instance either via ….withProperties(…) or on instance creation via Problem.create(…).

Using a dedicated type to capture extended problem properties
class AccountDetails {
  int balance;
  List<URI> accounts;


// or


This will result in a response looking like this:

A sample HTTP Problem Details response
  "type": "",
  "title": "You do not have enough credit.",
  "detail": "Your current balance is 30, but that costs 50.",
  "instance": "/account/12345/msgs/abc",
  "balance": 30,
  "accounts": ["/account/12345",

4.4. Collection+JSON

Collection+JSON is a JSON spec registered with IANA-approved media type application/vnd.collection+json.

Collection+JSON is a JSON-based read/write hypermedia-type designed to support management and querying of simple collections.

— Mike Amundsen
Collection+JSON spec

Collection+JSON provides a uniform way to represent both single item resources as well as collections. To enable this media type, put the following configuration in your code:

Example 36. Collection+JSON enabled application
@EnableHypermediaSupport(type = HypermediaType.COLLECTION_JSON)
public class CollectionJsonApplication {


This configuration will make your application respond to requests that have an Accept header of application/vnd.collection+json as shown below.

The following example from the spec shows a single item:

Example 37. Collection+JSON single item example
  "collection": {
    "version": "1.0",
    "href": "", (1)
    "links": [   (2)
        "rel": "feed",
        "href": ""
        "rel": "queries",
        "href": ""
        "rel": "template",
        "href": ""
    "items": [  (3)
        "href": "",
        "data": [  (4)
            "name": "fullname",
            "value": "J. Doe",
            "prompt": "Full Name"
            "name": "email",
            "value": "[email protected]",
            "prompt": "Email"
        "links": [ (5)
            "rel": "blog",
            "href": "",
            "prompt": "Blog"
            "rel": "avatar",
            "href": "",
            "prompt": "Avatar",
            "render": "image"
1 The self link is stored in the document’s href attribute.
2 The document’s top links section contains collection-level links (minus the self link).
3 The items section contains a collection of data. Since this is a single-item document, it only has one entry.
4 The data section contains actual content. It’s made up of properties.
5 The item’s individual links.

The previous fragment was lifted from the spec. When Spring HATEOAS renders an EntityModel, it will:

  • Put the self link into both the document’s href attribute and the item-level href attribute.

  • Put the rest of the model’s links into both the top-level links as well as the item-level links.

  • Extract the properties from the EntityModel and turn them into …​

When rendering a collection of resources, the document is almost the same, except there will be multiple entries inside the items JSON array, one for each entry.

Spring HATEOAS more specifically will:

  • Put the entire collection’s self link into the top-level href attribute.

  • The CollectionModel links (minus self) will be put into the top-level links.

  • Each item-level href will contain the corresponding self link for each entry from the CollectionModel.content collection.

  • Each item-level links will contain all other links for each entry from CollectionModel.content.

4.5. UBER - Uniform Basis for Exchanging Representations

UBER is an experimental JSON spec

The UBER document format is a minimal read/write hypermedia type designed to support simple state transfers and ad-hoc hypermedia-based transitions.

— Mike Amundsen
UBER spec

UBER provides a uniform way to represent both single item resources as well as collections. To enable this media type, put the following configuration in your code:

Example 38. UBER+JSON enabled application
@EnableHypermediaSupport(type = HypermediaType.UBER)
public class UberApplication {


This configuration will make your application respond to requests using the Accept header application/vnd.amundsen-uber+json as show below:

Example 39. UBER sample document
  "uber" : {
    "version" : "1.0",
    "data" : [ {
      "rel" : [ "self" ],
      "url" : "/employees/1"
    }, {
      "name" : "employee",
      "data" : [ {
        "name" : "role",
        "value" : "ring bearer"
      }, {
        "name" : "name",
        "value" : "Frodo"
      } ]
    } ]

This media type is still under development as is the spec itself. Feel free to open a ticket if you run into issues using it.

UBER media type is not associated in any way with Uber Technologies Inc., the ride sharing company.

4.6. ALPS - Application-Level Profile Semantics

ALPS is a media type for providing profile-based metadata about another resource.

An ALPS document can be used as a profile to explain the application semantics of a document with an application- agnostic media type (such as HTML, HAL, Collection+JSON, Siren, etc.). This increases the reusability of profile documents across media types.

— Mike Amundsen
ALPS spec

ALPS requires no special activation. Instead you "build" an Alps record and return it from either a Spring MVC or a Spring WebFlux web method as shown below:

Example 40. Building an Alps record
@GetMapping(value = "/profile", produces = ALPS_JSON_VALUE)
Alps profile() {

  return Alps.alps() //
      .doc(doc() //
          .href("") //
          .value("value goes here") //
          .format(Format.TEXT) //
          .build()) //
      .descriptor(getExposedProperties(Employee.class).stream() //
          .map(property -> Descriptor.builder() //
              .id("class field [" + property.getName() + "]") //
              .name(property.getName()) //
              .type(Type.SEMANTIC) //
              .ext(Ext.builder() //
                  .id("ext [" + property.getName() + "]") //
                  .href("" + property.getName()) //
                  .value("value goes here") //
                  .build()) //
              .rt("rt for [" + property.getName() + "]") //
              .descriptor(Collections.singletonList(Descriptor.builder().id("embedded").build())) //
              .build()) //
  • This example leverages PropertyUtils.getExposedProperties() to extract metadata about the domain object’s attributes.

This fragment has test data plugged in. It yields JSON like this:

Example 41. ALPS JSON
  "version": "1.0",
  "doc": {
    "format": "TEXT",
    "href": "",
    "value": "value goes here"
  "descriptor": [
      "id": "class field [name]",
      "name": "name",
      "type": "SEMANTIC",
      "descriptor": [
          "id": "embedded"
      "ext": {
        "id": "ext [name]",
        "href": "",
        "value": "value goes here"
      "rt": "rt for [name]"
      "id": "class field [role]",
      "name": "role",
      "type": "SEMANTIC",
      "descriptor": [
          "id": "embedded"
      "ext": {
        "id": "ext [role]",
        "href": "",
        "value": "value goes here"
      "rt": "rt for [role]"

Instead of linking each field "automatically" to a domain object’s fields, you can write them by hand if you like. It’s also possible to use Spring Framework’s message bundles and the MessageSource interface. This gives you the ability to delegate these values to locale-specific message bundles and even internationalize the metadata.

4.7. Community-based media types

Thanks to the ability to create your own media type, there are several community-led efforts to build additional media types.

4.7.1. JSON:API

Maven coordinates
    <version>{see project page for current version}</version>
Gradle coordinates
implementation 'com.toedter:spring-hateoas-jsonapi:{see project page for current version}'

Visit the project page for more details if you want snapshot releases.

4.7.2. Siren

Maven coordinates
    <version>{see project page for current version}</version>
Gradle coordinates
implementation 'de.ingogriebsch.hateoas:spring-hateoas-siren:{see project page for current version}'

4.8. Registering a custom media type

Spring HATEOAS allows you to integrate custom media types through an SPI. The building blocks of such an implementation are:

  1. Some form of Jackson ObjectMapper customization. In its most simple case that’s a Jackson Module implementation.

  2. A LinkDiscoverer implementation so that the client-side support is able to detect links in representations.

  3. A small bit of infrastructure configuration that will allow Spring HATEOAS to find the custom implementation and pick it up.

4.8.1. Custom media type configuration

Custom media type implementations are picked up by Spring HATEOAS by scanning the application context for any implementations of the HypermediaMappingInformation interface. Each media type must implement this interface in order to:

To define your own media type could look as simple as this:

public class MyMediaTypeConfiguration implements HypermediaMappingInformation {

  public List<MediaType> getMediaTypes() {
    return Collections.singletonList(MediaType.parseMediaType("application/vnd-acme-media-type")); (1)

  public Module getJacksonModule() {
    return new Jackson2MyMediaTypeModule(); (2)

  MyLinkDiscoverer myLinkDiscoverer() {
    return new MyLinkDiscoverer(); (3)
1 The configuration class returns the media type it supports. This applies to both server-side and client-side scenarios.
2 It overrides getJacksonModule() to provide custom serializers to create the media type specific representations.
3 It also declares a custom LinkDiscoverer implementation for further client-side support.

The Jackson module usually declares Serializer and Deserializer implementations for the representation model types RepresentationModel, EntityModel, CollectionModel and PagedModel. In case you need further customization of the Jackson ObjectMapper (like a custom HandlerInstantiator), you can alternatively override configureObjectMapper(…).

Prior versions of reference documentation has mentioned implementing the MediaTypeConfigurationProvider interface and registering it with spring.factories. This is NOT necessary. This SPI is ONLY used for out-of-the-box media types provided by Spring HATEOAS. Merely implementing the HypermediaMappingInformation interface and registering it as a Spring bean is all that’s needed.

4.8.2. Recommendations

The preferred way to implement media type representations is by providing a type hierarchy that matches the expected format and can be serialized by Jackson as is. In the Serializer and Deserializer implementations registered for RepresentationModel, convert the instances into the media type-specific model types and then lookup the Jackson serializer for those.

The media types supported by default use the same configuration mechanism as third-party implementations would do. So it’s worth studying the implementations in the mediatype package. Note, that the built in media type implementations keep their configuration classes package private, as they’re activated via @EnableHypermediaSupport. Custom implementations should probably make those public instead to make sure, users can import those configuration classes from their application packages.

5. Configuration

This section describes how to configure Spring HATEOAS.

5.1. Using @EnableHypermediaSupport

To let the RepresentationModel subtypes be rendered according to the specification of various hypermedia representation types, you can activate support for a particular hypermedia representation format through @EnableHypermediaSupport. The annotation takes a HypermediaType enumeration as its argument. Currently, we support HAL as well as a default rendering. Using the annotation triggers the following:

  • It registers necessary Jackson modules to render EntityModel and CollectionModel in the hypermedia specific format.

  • If JSONPath is on the classpath, it automatically registers a LinkDiscoverer instance to look up links by their rel in plain JSON representations (see Using LinkDiscoverer Instances).

  • By default, it enables entity links and automatically picks up EntityLinks implementations and bundles them into a DelegatingEntityLinks instance that you can autowire.

  • It automatically picks up all RelProvider implementations in the ApplicationContext and bundles them into a DelegatingRelProvider that you can autowire. It registers providers to consider @Relation on domain types as well as Spring MVC controllers. If the EVO inflector is on the classpath, collection rel values are derived by using the pluralizing algorithm implemented in the library (see [spis.rel-provider]).

5.1.1. Explicitly enabling support for dedicated web stacks

By default, @EnableHypermediaSupport will reflectively detect the web application stack you’re using and hook into the Spring components registered for those to enable support for hypermedia representations. However, there are situations in which you’d only explicitly want to activate support for a particular stack. E.g. if your Spring WebMVC based application uses WebFlux' WebClient to make outgoing requests and that one is not supposed to work with hypermedia elements, you can restrict the functionality to be enabled by explicitly declaring WebMVC in the configuration:

Example 42. Explicitly activating hypermedia support for a particular web stack
@EnableHypermediaSupport(…, stacks = WebStack.WEBMVC)
class MyHypermediaConfiguration { … }

6. Client-side Support

This section describes Spring HATEOAS’s support for clients.

6.1. Traverson

Spring HATEOAS provides an API for client-side service traversal. It is inspired by the Traverson JavaScript library. The following example shows how to use it:

Map<String, Object> parameters = new HashMap<>();
parameters.put("user", 27);

Traverson traverson = new Traverson(URI.create("http://localhost:8080/api/"), MediaTypes.HAL_JSON);
String name = traverson
    .follow("movies", "movie", "actor").withTemplateParameters(parameters)

You can set up a Traverson instance by pointing it to a REST server and configuring the media types you want to set as Accept headers. You can then define the relation names you want to discover and follow. Relation names can either be simple names or JSONPath expressions (starting with an $).

The sample then hands a parameter map into the Traverson instance. The parameters are used to expand URIs (which are templated) found during the traversal. The traversal is concluded by accessing the representation of the final traversal. In the preceding example, we evaluate a JSONPath expression to access the actor’s name.

The preceding example is the simplest version of traversal, where the rel values are strings and, at each hop, the same template parameters are applied.

There are more options to customize template parameters at each level. The following example shows these options.

ParameterizedTypeReference<EntityModel<Item>> resourceParameterizedTypeReference = new ParameterizedTypeReference<EntityModel<Item>>() {};

EntityModel<Item> itemResource = traverson.//
    follow(rel("items").withParameter("projection", "noImages")).//

The static rel(…​) function is a convenient way to define a single Hop. Using .withParameter(key, value) makes it simple to specify URI template variables.

.withParameter() returns a new Hop object that is chainable. You can string together as many .withParameter as you like. The result is a single Hop definition. The following example shows one way to do so:
ParameterizedTypeReference<EntityModel<Item>> resourceParameterizedTypeReference = new ParameterizedTypeReference<EntityModel<Item>>() {};

Map<String, Object> params = Collections.singletonMap("projection", "noImages");

EntityModel<Item> itemResource = traverson.//

You can also load an entire Map of parameters by using .withParameters(Map).

follow() is chainable, meaning you can string together multiple hops, as shown in the preceding examples. You can either put multiple string-based rel values (follow("items", "item")) or a single hop with specific parameters.

6.1.1. EntityModel<T> vs. CollectionModel<T>

The examples shown so far demonstrate how to sidestep Java’s type erasure and convert a single JSON-formatted resource into a EntityModel<Item> object. However, what if you get a collection like an \_embedded HAL collection? You can do so with only one slight tweak, as the following example shows:

CollectionModelType<Item> collectionModelType =
    TypeReferences.CollectionModelType<Item>() {};

CollectionModel<Item> itemResource = traverson.//

Instead of fetching a single resource, this one deserializes a collection into CollectionModel.

When working with hypermedia enabled representations, a common task is to find a link with a particular relation type in it. Spring HATEOAS provides JSONPath-based implementations of the LinkDiscoverer interface for either the default representation rendering or HAL out of the box. When using @EnableHypermediaSupport, we automatically expose an instance supporting the configured hypermedia type as a Spring bean.

Alternatively, you can set up and use an instance as follows:

String content = "{'_links' :  { 'foo' : { 'href' : '/foo/bar' }}}";
LinkDiscoverer discoverer = new HalLinkDiscoverer();
Link link = discoverer.findLinkWithRel("foo", content);

assertThat(link.getRel(), is("foo"));
assertThat(link.getHref(), is("/foo/bar"));

6.3. Configuring WebClient instances

If you need configure a WebClient to speak hypermedia, it’s easy. Get a hold of the HypermediaWebClientConfigurer as shown below:

Example 43. Configuring a WebClient yourself
WebClient.Builder hypermediaWebClient(HypermediaWebClientConfigurer configurer) { (1)
 return configurer.registerHypermediaTypes(WebClient.builder()); (2)
1 Inside your @Configuration class, get a copy of the HypermediaWebClientConfigurer bean Spring HATEOAS registers.
2 After creating a WebClient.Builder, use the configurer to register hypermedia types.
What HypermediaWebClientConfigurer does it register all the right encoders and decoders with a WebClient.Builder. To make use of it, you need to inject the builder somewhere into your application, and run the build() method to produce a WebClient.

If you’re using Spring Boot, there is another way: the WebClientCustomizer.

Example 44. Letting Spring Boot configure things
@Bean (4)
WebClientCustomizer hypermediaWebClientCustomizer(HypermediaWebClientConfigurer configurer) { (1)
    return webClientBuilder -> { (2)
        configurer.registerHypermediaTypes(webClientBuilder); (3)
1 When creating a Spring bean, request a copy of Spring HATEOAS’s HypermediaWebClientConfigurer bean.
2 Use a Java 8 lambda expression to define a WebClientCustomizer.
3 Inside the function call, apply the registerHypermediaTypes method.
4 Return the whole thing as a Spring bean so Spring Boot can pick it up and apply it to its autoconfigured WebClient.Builder bean.

At this stage, whenever you need a concrete WebClient, simply inject WebClient.Builder into your code, and use build(). The WebClient instance will be able to interact using hypermedia.

6.4. Configuring WebTestClient Instances

When working with hypermedia-enabled representations, a common task is to run various tests by using WebTestClient.

To configure an instance of WebTestClient in a test case, check out this example:

Example 45. Configuring WebTestClient when using Spring HATEOAS
@Test // #1225
void webTestClientShouldSupportHypermediaDeserialization() {

  // Configure an application context programmatically.
  AnnotationConfigApplicationContext context = new AnnotationConfigApplicationContext();
  context.register(HalConfig.class); (1)

  // Create an instance of a controller for testing
  WebFluxEmployeeController controller = context.getBean(WebFluxEmployeeController.class);

  // Extract the WebTestClientConfigurer from the app context.
  HypermediaWebTestClientConfigurer configurer = context.getBean(HypermediaWebTestClientConfigurer.class);

  // Create a WebTestClient by binding to the controller and applying the hypermedia configurer.
  WebTestClient client = WebTestClient.bindToApplicationContext(context).build().mutateWith(configurer); (2)

  // Exercise the controller.
  client.get().uri("http://localhost/employees").accept(HAL_JSON) //
      .exchange() //
      .expectStatus().isOk() //
      .expectBody(new TypeReferences.CollectionModelType<EntityModel<Employee>>() {}) (3)
      .consumeWith(result -> {
        CollectionModel<EntityModel<Employee>> model = result.getResponseBody(); (4)

        // Assert against the hypermedia model.
1 Register your configuration class that uses @EnableHypermediaSupport to enable HAL support.
2 Use HypermediaWebTestClientConfigurer to apply hypermedia support.
3 Ask for a response of CollectionModel<EntityModel<Employee>> using Spring HATEOAS’s TypeReferences.CollectionModelType helper.
4 After getting the "body" in Spring HATEOAS format, assert against it!
WebTestClient is an immutable value type, so you can’t alter it in place. HypermediaWebClientConfigurer returns a mutated variant that you must then capture to use it.

If you are using Spring Boot, there are additional options, like this:

Example 46. Configuring WebTestClient when using Spring Boot
@AutoConfigureWebTestClient (1)
class WebClientBasedTests {

    void exampleTest(@Autowired WebTestClient.Builder builder, @Autowired HypermediaWebTestClientConfigurer configurer) { (2)
        client = builder.apply(configurer).build(); (3)

        client.get().uri("/") //
                .exchange() //
                .expectBody(new TypeReferences.EntityModelType<Employee>() {}) (4)
                .consumeWith(result -> {
                    // assert against this EntityModel<Employee>!
1 This is Spring Boot’s test annotation that will configure a WebTestClient.Builder for this test class.
2 Autowire Spring Boot’s WebTestClient.Builder into builder and Spring HATEOAS’s configurer as method parameters.
3 Use HypermediaWebTestClientConfigurer to register support for hypermedia.
4 Signal you want an EntityModel<Employee> returned using TypeReferences.

Again, you can use similar assertions as the earlier example.

There are many other ways to fashion test cases. WebTestClient can be bound to controllers, functions, and URLs. This section isn’t meant to show all that. Instead, this gives you some examples to get started. The important thing is that by applying HypermediaWebTestClientConfigurer, any instance of WebTestClient can be altered to handle hypermedia.

6.5. Configuring RestTemplate instances

If you want to create your own copy of RestTemplate, configured to speak hypermedia, you can use the HypermediaRestTemplateConfigurer:

Example 47. Configuring RestTemplate yourself
 * Use the {@link HypermediaRestTemplateConfigurer} to configure a {@link RestTemplate}.
RestTemplate hypermediaRestTemplate(HypermediaRestTemplateConfigurer configurer) { (1)
	return configurer.registerHypermediaTypes(new RestTemplate()); (2)
1 Inside your @Configuration class, get a copy of the HypermediaRestTemplateConfigurer bean Spring HATEOAS registers.
2 After creating a RestTemplate, use the configurer to apply hypermedia types.

You are free to apply this pattern to any instance of RestTemplate that you need, whether is to create a registered bean, or inside a service you define.

If you’re using Spring Boot, there is another approach.

In general, Spring Boot has moved away from the concept of registering a RestTemplate bean in the application context.

  • When talking to different services, you often need different credentials.

  • When RestTemplate uses an underlying connection pool, you run into additional issues.

  • Users often need different instances rather than a single bean.

To compensate for this, Spring Boot provides a RestTemplateBuilder. This autoconfigured bean lets you define various beans used to fashion a RestTemplate instance. You ask for a RestTemplateBuilder bean, call its build() method, and then apply final settings (such as credentials and other details).

To register hypermedia-based message converters, add the following to your code:

Example 48. Letting Spring Boot configure things
@Bean (4)
RestTemplateCustomizer hypermediaRestTemplateCustomizer(HypermediaRestTemplateConfigurer configurer) { (1)
    return restTemplate -> { (2)
        configurer.registerHypermediaTypes(restTemplate); (3)
1 When creating a Spring bean, request a copy of Spring HATEOAS’s HypermediaRestTemplateConfigurer bean.
2 Use a Java 8 lambda expression to define a RestTemplateCustomizer.
3 Inside the function call, apply the registerHypermediaTypes method.
4 Return the whole thing as a Spring bean so Spring Boot can pick it up and apply it to its autoconfigured RestTemplateBuilder.

At this stage, whenever you need a concrete RestTemplate, simply inject RestTemplateBuilder into your code, and use build(). The RestTemplate instance will be able to interact using hypermedia.