The attribute was previously removed only before exception resolution
in the DispatcherServlet in order to allow error rendering to make an
independent choice on content negotation.
However, Boot rendering happens later in an ERROR dispatch which could
also be a nested dispatch on some servers. So the attribute must also
generally be removed prior to mapping.
We also move the methods where this is done to the base
RequestMappingInfoHandlerMapping class which also deals with the
produces condition and where the producible attribute is added in the
first place.
Closes gh-24466
Even thought Tomcat and Jetty, which commit the response more lazily,
were not impacted by this issue, it still makes sense for them to
complete the WebFlux response (and pre-commit actions) at the same time
as for other servers for consistent behavior.
See gh-24475
Prior to this commit, some WebSocket `RequestUpgradeStrategy` reactive
implementations would prevent the application from writing HTTP headers
and cookies to the response.
For Reactor Netty and Undertow, handling the upgrade and starting the
WebSocket communication marks the response status and headers as sent
and the application cannot update HTTP response headers after that.
This commit ensures that the `RequestUpgradeStrategy` implementations
mark the responses as "complete", so that headers are written before we
delegate to the server implementation.
Fixes gh-24475
This commit updates CORS support in order to check Origin header
in CorsUtils#isPreFlightRequest which does not change how Spring
MVC or WebFlux process CORS request but is more correct in term
of behavior since it is a public API potentially used in another
contexts.
It also removes an unnecessary check in
AbstractHandlerMethodMapping#hasCorsConfigurationSource and processes
every preflight request with PreFlightHandler.
Closes gh-24327
When a request is mapped through a producible condition on an
@RequestMapping, then a failure to find a converter/decoder should be
a 500 because the return type + media type pair were declared by the
controller and that should be possible to render.
Closes gh-23287
Prior to this commit, when WebFlux handlers added `"Content-*"` response
headers and an error happened while handling the request, all those
headers would not be cleared from the response before error handling.
This commit clears those headers from the response in two places:
* when invoking the handler and adapting the response
* when writing the response body
Not removing those headers might break HTTP clients since they're given
wrong information about how to interpret the HTTP response body: the
error response body might be very different from the original one.
Fixes gh-24238
Prior to this commit, WebFlux application would keep the quality
parameter from the "Accept" request header when selecting a media type
for the response. It would then echo it back to the client.
While strictly not wrong, this is unnecessary and can confuse HTTP
clients. This commit aligns WebFlux's behavior with Spring MVC.
Fixes gh-24239
Using Consumer<ClientCodecConfigurer> instead of
Consumer<ExchangeStrategies> eliminates one level of nesting that is
also unnecessary since codecs are the only strategy at present.
Closes gh-24124
The new register methods replace the now deprecated
encoder, decoder, reader, and writer methods, and also offer a choice
to opt into default properties such maxInMemorySize, if configured.
See gh-24124
Prior to this commit, developers could configure their WebClient to use
their custom `ExchangeStrategies`, by providing it in the
`WebClient.Builder` chain.
Once created, an `ExchangeStrategies` instance is not mutable, which
makes it hard for further customizations by other components. In the
case of the reported issue, other components would override the default
configuration for the codecs maxInMemorySize.
This commit makes the `ExchangeStrategies` mutable and uses that fact to
further customize them with a new `WebClient.Builder#exchangeStrategies`
`Consumer` variant. This commit is also deprecating those mutating
variants in favor of a new `WebClient.Builder#exchangeStrategies` that
takes a `ExchangeStrategies#Builder` directly and avoids mutation issues
altogether.
Closes gh-23961
Prior to this commit, developers could configure their WebClient to use
their custom `ExchangeStrategies`, by providing it in the
`WebClient.Builder` chain.
Once created, an `ExchangeStrategies` instance is not mutable, which
makes it hard for further customizations by other components. In the
case of the reported issue, other components would override the default
configuration for the codecs maxInMemorySize.
This commit makes the `ExchangeStrategies` mutable and uses that fact to
further customize them with a new `WebClient.Builder#exchangeStrategies`
`Consumer` variant. This commit is also deprecating those mutating
variants in favor of a new `WebClient.Builder#exchangeStrategies` that
takes a `ExchangeStrategies#Builder` directly and avoids mutation issues
altogether.
Closes gh-23961
This commit refines Coroutines annotated controller support
by considering Kotlin Unit as Java void and using the right
ReactiveAdapter to support all use cases, including suspending
functions that return Flow (usual when using APIs like WebClient).
It also fixes RSocket fire and forget handling and adds related tests
for that use case.
Closes gh-24057
Closes gh-23866
Previously, spring-webmvc and spring-webflux both contained tests
that would create gzipped files, write them to the filesystem
alongside the project's compiled test classes, and configure them to
be deleted on JVM exit. The output location placed the files on the
classpath, polluting it for every subsequent test that used the same
ClassLoader. The test-sources plugin combined with Gradle's use of
worker JVMs, broadens the scope of this pollution to other, downstream
projects in the same build. For example, the tests for
spring-websocket will have a different classpath depending on whether
or not the tests for spring-webmvc have already been run on the same
worker as part of the current build.
This commit updates the spring-webmvc and spring-webflux modules to
introduce a new JUnit Jupiter extension, GzipSupport. This extension
allows gzipped files to be created via an injectable GzippedFiles
class and automatically deletes each created file in an after-each
callback. This ensures that a gzipped file only exists on the
classpath for the duration of the test that needs it, avoiding the
pollution of the classpath of any subsequent tests.
Closes gh-23970
Previously, the infrastructure provided by WebMvcConfigurationSupport
and WebFluxConfigurationSupport can lead to unexpected results due to
the lack of qualifier for certain dependencies. Those configuration
classes refer to very specific beans, yet their injection points do not
define such qualifiers. As a result, if a candidate exists for the
requested type, the context will inject the existing bean and will
ignore a most specific one as such constraint it not defined. This can
be easily reproduced by having a primary Validator whereas a dedicated
"mvcValidator" is expected. Note that a parameter name is in no way a
constraint as the name is only used as a fallback when a single
candidate cannot be determined.
This commit provides explicit @Qualifier metadata for such injection
points, renaming the parameter name in the process to clarify that it
isn't relevant for the proper bean to be resolved by the context.
Closes gh-23887
When built in an environment where many Gradle build workers are
available, :spring-webflux:test is regularly the only task running at
the end of the build. Therefore, the build's overall execution time
can be reduced by running its tests in parallel, spreading the tests'
execution across the available workers. The configured number of forks
is a maximum with Gradle reducing this as necessary for environments
with low numbers of cores where multiple workers will not improve
build performance.
Closes gh-23701
Prior to this commit, the parameterized DataBufferFactory was never
actually used when setting up the WebClient for each test. This was due
to an oversight when migrating from JUnit 4 to JUnit Jupiter.
See: https://github.com/reactor/reactor-netty/issues/860
This commit fixes this by converting the existing @BeforeEach method to
a local setup method that is invoked from each
@ParameterizedDataBufferAllocatingTest method.
In addition, in order to avoid the 2 second "quiet period" that is
incurred when destroying the ReactorResourceFactory, this commit moves
the setup and destruction of the ReactorResourceFactory to new
@BeforeAll and @AfterAll methods.
The test instance lifecycle has also been switched to PER_CLASS to avoid
static state in the test class.
This commit reverts b2704e1db6 and
configures the ReactorResourceFactory not to use global resources,
thereby allowing all tests in WebClientDataBufferAllocatingTests to
execute within approximately 2 seconds again on Mac OS.
Prior to this commit, Spring Framework would use `Schedulers.elastic()`
in places where we needed to process blocking tasks in a reactive
environment.
With reactor/reactor-core#1804, a new `Schedulers.boundedElastic()`
scheduler is available and achieves the same goal with added security;
it guarantees that resources are bounded.
This commit uses that new scheduler in the standard websocket client,
since the underlying API is blocking for the connection phase and we
need to schedule that off a web server thread.
Closes gh-23661
See gh-23665
Prior to this commit, the isSimpleProperty() and isSimpleValueType()
methods in BeanUtils treated void and Void as simple types; however,
doing so does not make sense in this context, since void implies the
lack of a property or value.
This commit addresses this by explicitly excluding void and Void in the
logic in isSimpleValueType().
This commit also simplifies the implementation of
ViewResolutionResultHandler.supports(HandlerResult) to take advantage
of this change.
Closes gh-23573
Prior to this commit, ClassUtils.isPrimitiveOrWrapper() and
ClassUtils.isPrimitiveWrapper() did not return true for Void.class.
However, ClassUtils.isPrimitiveOrWrapper() did return true for
void.class. This lacking symmetry is inconsistent and can lead to bugs
in reflective code.
See: https://github.com/spring-projects/spring-data-r2dbc/issues/159
This commit addresses this by adding an entry for Void.class -> void.class
in the internal primitiveWrapperTypeMap in ClassUtils.
Closes gh-23572
Rossen Stoyanchev
Brian Clozel
Hyunjin Choi
Jacob Severson
Sébastien Deleuze
Sam Brannen
stsypanov
Arjen Poutsma
Juergen Hoeller
Sam Brannen
Biju Kunjummen
Andy Wilkinson
Johnny Lim
wonwoo
Christoph Dreis
Stephane Nicoll
Phillip Webb