So far, we handed the plain instance deserialized from the request body to the repository to persist it. That caused issues in PUT for create scenarios where the URI contains the identifier to be used for the aggregate to create and identifier generation being used in the backend. In that case the identifier submitted was never considered and subsequent requests would've created new instances, effectively breaking the idempotent nature of PUT.
We now make sure the backend identifier derived from the resource is set on the aggregate instance about to be created, so that backend can either accept that situation (new entity + manually defined identifier) or reject it (in case it insists on identifier generation).
We previously erroneously looked for a ~ in a JSON Pointer to indicate collection append in e.g. an add operation. We now also support the actually correct - keeping the original behavior for now to not break clients that currently make use of it.
Repositories in RepositoryCorsConfigurationAccessor may be null now. findCorsConfiguration returns null when no repositories are provided.
Original pull request: #257.
We now explicitly merge associations skipping the linkable ones. We also try to reuse the existing collections and maps if they're mutable falling back to a completely new one if not.
Extracted PropertyHandler and AssociationHandler implementations.
When merging collections on PATCH we now don't use the first collection item's type for all elements but inspect the values for each existing element found. When it comes to appending elements to the collection, wen now just stick to the declared component type as type hint for reading the provided value.
When copying the transient properties of an aggregate, we now try field based access first and fall back to accessor based copying in case both a setter and getter are exposed on the type.
Previously we always expected a field to be present which doesn't necessarily has to be the case.
Related ticket: DATAREST-986.
Jacksons ObjectMapper.readerForUpdate(…) unfortunately doesn't handle nested objects properly. We already have a manual merge process in place for PATCH requests but tweaking that to also handle PUT requests gracefully caused more complexity than anticipated.
We now switched to an object based merge so that we can read in the source JSON structure into a new object and then merge the objects.
Related pull request: #247.
We now use a Map property's generic type information to make sure we convert both the key and the value into the declared types. Previously we just used Object if the source value to map was null. Object is still used as fallback for raw maps though.
We now allow sorting by properties of embedded objects. An embedded object is not linkable to a root object but embedded in the resource itself. If any part of the sort property path points to a linkable association, the whole sort property path is discarded silently and not used for sorting any further.
Original pull request: #251.
Skip all merge logic if the source value is null. That frees all nested logic from handling with that case and us falling back to plain Jackson reading.
The array handling now also opts out if the source value is not a collection or array in the first place as it means we need to let Jackson override the value with the collection given to be deserialized.
Original pull request: #246.
We now interpret If-None-Match and If-Modified-Since headers on requests to resources backed by query methods returning a single instance only. This allows clients to optimize GET requests to those resources to save bandwidth.
Some tiny refactorings in DomainObjectReader. We're now using TypeInformation instead of Class to preserve more generics information when it comes to deeper nesting.
Moved some code around in the unit tests.
Original pull request: #245.
We now support nested Sort properties considering Jackson mapping. Sort translation is optional and skipped if the domain class is not resolvable. Translation in the scope of a domain class maps property paths to apply sorting using embedded properties.
A sort string `nested-name` resolves to a property path `anotherWrap.embedded.name`.
class Aggregate {
@JsonUnwrapped
public UnwrapEmbedded anotherWrap;
}
class UnwrapEmbedded {
@JsonUnwrapped(prefix = "nested-")
public Embedded embedded;
}
class Embedded {
public String name;
}
Original pull request: #232.
We now don't prematurely drop fields that don't have a persistent property exposed in DomainObjectReader. Doing so dropped values for transient fields as the latter are not exposed as persistent property in the first place. We still skip any nested merging though.
Original pull request: #240.
We now explicitly manually merge array nodes that contain complex objects. Previously arrays would've been skipped and the subsequent Jackson update would wipe out all properties not contained in the original document even on PATCH requests.
Removed RepositoryRestConfiguration.addCorsMapping(…) as we currently don't have any other shortcut methods for configuration like this.
Tweaked the setup of (now Repository)CorsConfigurationAccessor to be created earlier so that we avoid recreation for every lookup. Introduced a NoOpStringValueResolver to be used by default so that we don't need to deal with the case of the resolver being null at the end of the call chain. Replaced constructor of RepositoryCorsConfigurationAccessor with corresponding Lombok annotation.
Updated reference documentation accordingly.
Original pull request: #233.
We now support CORS configuration mechanisms introduced in Spring 4.2. CORS can be configured on multiple levels: Repository interface, Repository REST controller and global level. Spring Data REST CORS configuration is isolated so Spring Web MVC'S CORS configuration does not apply to Spring Data REST resources.
Multiple configuration sources are merged so different aspects of CORS can be configured in separate locations.
@CrossOrigin
interface PersonRepository extends CrudRepository<Person, Long> {}
@RepositoryRestController
@RequestMapping("/person")
public class PersonController {
@CrossOrigin(maxAge = 3600)
@RequestMapping(method = RequestMethod.GET, "/xml/{id}", produces = MediaType.APPLICATION_XML_VALUE)
public Person retrieve(@PathVariable Long id) {
// ...
}
}
@Component
public class SpringDataRestCustomization extends RepositoryRestConfigurerAdapter {
@Override
public void configureRepositoryRestConfiguration(RepositoryRestConfiguration config) {
config.addCorsMapping("/person/**")
.allowedOrigins("http://domain2.com")
.allowedMethods("PUT", "DELETE")
.allowedHeaders("header1", "header2", "header3")
.exposedHeaders("header1", "header2")
.allowCredentials(false).maxAge(3600);
}
}
Previously, MappedProperties didn't handle Jackson properties correctly, that do not expose a PersistentProperty, e.g. transient ones. That led to potential nullPointerExceptions in clients as the guarding hasPersistentPropertyForField(…) still answered true, as the backing cache contained an entry with a null value. We now skip those properties completely.
We now resolve default configured Pageable without dropping them if their Sort is empty. Previously, any default Pageable was dropped if its Sort was empty which caused null being handed to controller methods of @RepositoryRestController instances.
Original pull request: #231.
Switched to use @RequiredArgsConstructor where possible. Slightly rearranged test cases. Inlined JacksonMappingAwareSortTranslator to not expose it as bean unless necessary. Use static Jackson BeanClassIntrospector to avoid unnecessary recreation.
Original pull request: #222.
We now consider Jackson field names when resolving Sort arguments. Domain model properties annotated with @JsonProperty("sales") can be specified by their Jackson-mapped field name in sort arguments. Field names are mapped to their according persistent property names to be used in repository query method sorting. Unknown field names are silently dropped.
Original pull request: #222.
The previously static ObjectMapper used by JsonPatchPatchHandler now has to be provided by clients instantiating the handler.
Original pull request: #224.
PUT requests are supposed to replace the state of the resource with the request payload. However, Spring Data REST already handles a couple of domain object propoerties in a special way as theri values map to dedicated HTTP features: identifiers (URIs), last-modified dates (header) etc.
For users, it might be worthwhile to exclude other properties from being set by applying the payload, like properties that are completely under the control of the server, e.g. creation user and date, last modifying user etc. So far, users didn't have any means to exclude those properties as the handling of PUT requests treated every missing property of the payload as null value.
DomainObjectMerger now checks whether a property is actually writable before applying the implicit null value. The application can be disabled by annotation a property with @ReadOnlyProperty.
Spring 4.3's HeaderContentNegotiationStrategy switched from looking up the Accept header via the method returning a single (potentially comma separated) one to the method returning multiple headers in the first place. We now added an override of HttpServletRequestWrapper.getHeaders(…) to out custom adapter defaulting the media type to make sure the defaulting is visible and thus the right handler methods are looked up.
That previously missing caused the DelegatingHandlerMapping selecting the redirect to the HAL browser in case a request to the API root was sent without an accept header as the defaulting of the header to application/hal+json wasn't properly exposed anymore and the redirect to the browser — declaring a produces clause of text/html — was not causing any media type mismatch anymore.
Deprecated the ResourceProcessor invoking infrastructure in place here in favor of the types moved to Spring HATEOAS. Refactored our codebase to make use of these newly introduced types right away.
Related tickets: spring-projects/spring-hateoas#362.
A refactoring in the course of the 2.5 development dropped the application of explicit projection in case of excerpts being requested. This particularly applies to collection resources with explicit projection requests which need to get the explicit projection applied over the default excerpt projection potentially registered.
This is now fixed by the explicitly selected projection (if existing) always trumping the excerpt one.
