* Fix gateway proxy for generic interface in XML
Without a `FactoryBean.OBJECT_TYPE_ATTRIBUTE` the application context cannot determine a qualified
with generic an interface injection
* Set back `FactoryBean.OBJECT_TYPE_ATTRIBUTE` in the `GatewayParser`
when not in AOT mode.
Set a `targetType` when used in AOT
* * Remove unused import
* * Fix `GatewayParserTests` for new code base
* Add Java DSL for GraphQL
* GraphQl.outboundChannelAdapter() -> outboundGateway()
* * Add `private` ctor to `GraphQl`
* * Just `GraphQl.gateway()` - there is not going to be any other components
to handle GraphQL operations.
Also, the `outbound` in the name might confuse, where we, essentially, query a GraphQL data
* Fix typo in the doc
Co-authored-by: Gary Russell <grussell@vmware.com>
Co-authored-by: Gary Russell <grussell@vmware.com>
* Add observation to inbound endpoints
* * Improve `Observation` logic in the `MessageProducerSupport` and `MessagingGatewaySupport`
* Clean up Javadocs in observation support classes
* Add test for `MessagingGatewaySupport` instrumentation into `IntegrationObservabilityZipkinTests`
* Add docs for `@EnableIntegrationManagement.observationPatterns()`
* The generated docs for metrics and spans looks OK: have both `Gateway` and `Handler` section now
* * Fix Checkstyle violations
* * Fix `IntegrationMBeanExporter` to filter out a `MessageProducer` from sources.
Marking a `MessageProducerSupport` with an `IntegrationInboundManagement`
causes it to be considered as a source for JMX.
Technically it might have a reason since it is indeed a source, but that's fully different story
* * No observation instrumentation by default
* * Fix "no observation by default" configuration logic
* GH-3926: BeanNameGenerator for @MessagingGateway
Fixes https://github.com/spring-projects/spring-integration/issues/3926
Current approach for generated bean name in the `MessagingGatewayRegistrar`
is to decapitalize simple class name, which is similar to standard `AnnotationBeanNameGenerator`
* Make logic in the `MessagingGatewayRegistrar` based on the provided `BeanNameGenerator`
* Expose an `@IntegrationComponentScan.nameGenerator()` attribute to allow to customize
default bean name generation strategy
* Introduce `IntegrationConfigUtils.annotationBeanNameGenerator()` to
take a provided `AnnotationConfigUtils.CONFIGURATION_BEAN_NAME_GENERATOR` singleton
or fallback to the `AnnotationBeanNameGenerator.INSTANCE`
* Use this utility in the `IntegrationComponentScanRegistrar` if no custom strategy is provided
in the `@IntegrationComponentScan`
* Use same util from the `GatewayParser` since there is no custom naming strategy configuration
* Some other current Java level refactoring in the `IntegrationComponentScanRegistrar`
and `MessagingGatewayRegistrar`
* * Meta-annotate `@MessagingGateway` with a `@MessageEndpoint`
* Alias `@MessageEndpoint.value()` with a `@Component.value()`
* Alias `@MessagingGateway.name()` with a `@MessageEndpoint.value()`
* * Remove unused imports
* * Replace `MessagingGatewayRegistrar` parsing logic for annotation configuration
directly by the `GatewayProxyInstantiationPostProcessor` and `AnnotationGatewayProxyFactoryBean`.
* The `MessagingGatewayRegistrar` logic has been migrated back to the `GatewayParser`,
but only with an XML-relevant parts
* Change the logic of the `IntegrationComponentScanRegistrar` to rely on a `ClassPathBeanDefinitionScanner`
and its `scan()` functionality since this is all what we need to trigger a `GatewayProxyInstantiationPostProcessor`
for scanned components
* Fix `GatewayProxyInstantiationPostProcessor` to call an `afterPropertiesSet()` as well
* Add a `value()` alias attribute for the `@MessagingGateway` to satisfy a name resolution from a `@Component`
* Replace annotation chain resolution logic by new `MessagingAnnotationUtils.resolveMergedAttribute()` API
* Use `MergedAnnotations` API in the `AnnotationGatewayProxyFactoryBean` to preserve a logic for attribute
resolution from the annotation hierarchy
* Add expression resolution for attribute values in the `AnnotationGatewayProxyFactoryBean`
* Make a `GatewayInterfaceTests.CustomBeanNameGenerator` as an `AnnotationBeanNameGenerator`
extension to satisfy the test logic expectations: no custom name if explicit is present
* Clean up some doc typos after merge conflict
The fix in this commit essentially resolves some old JIRA ticket: https://jira.spring.io/browse/INT-4558
* * Remove redundant `MessagingAnnotationUtils.resolveMergedAttribute()`
* Optimize the logic in the `AnnotationGatewayProxyFactoryBean` around
annotation attributes to plain annotation - no adaptation to maps
* GH-3923: Add @MessagingGateway @Import support
Fixes https://github.com/spring-projects/spring-integration/issues/3923
Currently, the `@MessagingGateway` interfaces can be scanned or created
explicitly as `@Bean` definition for `AnnotationGatewayProxyFactoryBean`
* Implement a `GatewayProxyBeanDefinitionPostProcessor` which is invoked
before instantiation attempt on the `BeanDefinition`
* Verify `@Import` for `@MessagingGateway` interface in the `GatewayInterfaceTests`
* Remove supplier for `GatewayProxyInstantiationPostProcessor` bean definition
Co-authored-by: Gary Russell <grussell@vmware.com>
* * Fix bean definition signature
Co-authored-by: Gary Russell <grussell@vmware.com>
* Fix double start for `AbstractEndpoint`
When we use POJO methods in the `IntegrationFlowAdapter`,
the `IntegrationFlowAdapter` is set as a `target` for the `MessagingMethodInvokerHelper`.
When endpoint is started by the application context, such a `start()` is propagated
down to the `MessagingMethodInvokerHelper`.
And in our case back into an `IntegrationFlowAdapter` instance.
This one, in turn, starts its `IntegrationFlow` internally which leads to the
start of the mentioned endpoint in the beginning.
Therefore, we cause a recursive `start()` call on this endpoint from itself.
The `running` flag is set when we are already done with the `doStart()` logic.
Therefore a recursive `start()` call leads to two concurrent polling tasks
in the `AbstractPollingEndpoint`.
* Check also for the `active` flag in the `AbstractEndpoint.start()`
and reset it in case of exception in the `doStart()`
**Cherry-pick to `5.5.x`**
* * Change assert for message timestamps to `isCloseTo()` with percentage
* * Change `catch` in the `AbstractEndpoint.start()` to `RuntimeException`:
the `doStart()` cannot throw unchecked exceptions by definition
* * Fix imports in `AbstractEndpoint`
* Add generic arg to the `GatewayProxyFactoryBean`
Related to https://github.com/spring-projects/spring-integration/issues/3923
When gateway proxy is declared manually, a `GatewayProxyFactoryBean` is used as a `@Bean`.
In this case the info about target interface is not available on a `BeanDefinition`,
unlike with a programmatic registration via `MessagingGatewayRegistrar`.
* Expose `<T>` on a `GatewayProxyFactoryBean` to make end-user to specify the type
this gateway is going to be based on.
This allows Spring container to determine the type of the `FactoryBean` bean definition
properly
* Migrate a programmatic bean definition registration from the
`FactoryBean.OBJECT_TYPE_ATTRIBUTE` to the `targetType` property
of the `BeanDefinition` based on a `ResolvableType.forClassWithGenerics()`
to simulate generic arg for the application context
* Remove `ComponentsRegistration` from the `GatewayProxySpec` since it us out of use
* Fix affected tests for newly added generic arg on the `GatewayProxyFactoryBean`
* Fix another typo in gateway.adoc
Co-authored-by: Gary Russell <grussell@vmware.com>
Co-authored-by: Gary Russell <grussell@vmware.com>
* GH-3920: Support @Primary on @MessagingGateway
Fixes https://github.com/spring-projects/spring-integration/issues/3920
Any Spring `@Component` can also be marked with a `@Primary`.
The `@MessagingGateway` for proxying interfaces has no difference with
a standard `@Component` definition
* Modify `MessagingGatewayRegistrar` to obtain a `@Primary` marked
from the `importingClassMetadata` and pass it to the target `BeanDefinition`
* Verify the `@Primary` effect in the `GatewayInterfaceTests`
* Document this improvement
* * Fix Checkstyle violations
* Remove the `cassandra.yaml` from test resources since it does not
have any effect on the clients.
Use `datastax-java-driver.basic.request.timeout` system property instead
with `10 seconds` value to allow Cassandra client to wait for query results
a little more than 2 seconds
* Expose just an IP address for contact endpoint avoiding overhead
with attempts to connect to all addresses returned by `localhost` resolution
* Bring back `log4j-slf4j-impl` dependency since it is required for other libs
It is wrong to try to load an application context before test class in the `TestExecutionListener`
* Change `beforeTestClass()` phase for the `SpringIntegrationTestExecutionListener` to
the `prepareTestInstance()` when other `@BeforeAll` are already executed
**Cherry-pick to 5.5.x**
* Removed overridden `getMethodValue()` in the `CookieTests` which is removed from SF already
* Re-enable JDBC tests related to complex named params because of the fix in SF
* `--add-opens` for Tomcat warnings
* Upgrade to `log4j-slf4j2-impl`
* Add `cassandra.yaml` resource with increased timeouts to wait more in busy environment from Cassandra cluster
The `AbstractMethodAnnotationPostProcessor` makes a decision to use a `MessageHandler`
bean definition as is without wrapping by the type of `MessageHandler`.
There are some of them can be configured via `AbstractSimpleMessageHandlerFactoryBean`,
e.g. an `AggregatorFactoryBean`
* Check for the `AbstractSimpleMessageHandlerFactoryBean` to return as is
* Widen the replying producer by the `AbstractMessageProducingHandler` type
* Add an `AggregatorFactoryBean` with a `@ServiceActivator` configuration to test coverage
* Revert `getMethodValue()` for `CookieTests` since it is removed in the SF SNAPSHOT
* Add `@Suppress("UNCHECKED_CAST")` to `FunctionsTests.kt` to suppress compilation warning
* GH-3897: Deprecate `ChannelSecurityInterceptor`
Fixes https://github.com/spring-projects/spring-integration/issues/3897
Spring Security has deprecated `AccessDecisionManager` and all its infrastructure
in favor of `AuthorizationManager`
* Deprecate and AOP `ChannelSecurityInterceptor` and all its infrastructure,
including `@SecuredChannel` and respective XML configuration.
The `AuthorizationChannelInterceptor` added to respective channels for security
or configured as a global channel interceptor fully covers the previous AOP configuration
* Fix deprecation warnings in other tests with security
* Fix language in docs
Co-authored-by: Gary Russell <grussell@vmware.com>
* * Remove `forRemoval` attr from `@Deprecated` markers for Security classes:
looks like to mark `@Deprecated` and even `@SuppressWarnings("deprecation")`
don't silence warnings on compilation
Co-authored-by: Gary Russell <grussell@vmware.com>
* GH-3902: Add Kotlin Coroutines Support
Fixes https://github.com/spring-projects/spring-integration/issues/3902
* Add `isAsync()` propagation from the `MessagingMethodInvokerHelper`
to the `AbstractMessageProducingHandler` to set into its `async` property.
The logic is based on a `CompletableFuture`, `Publisher` or Kotlin `suspend`
return types of the POJO method
* Introduce `IntegrationMessageHandlerMethodFactory` and `IntegrationInvocableHandlerMethod`
to extend the logic to newly introduced `ContinuationHandlerMethodArgumentResolver`
and call for Kotlin suspend functions.
* Remove `MessageHandlerMethodFactoryCreatingFactoryBean` since its logic now is covered with the
`IntegrationMessageHandlerMethodFactory`
* Kotlin suspend functions are essentially reactive, so use `CoroutinesUtils.invokeSuspendingFunction()`
and existing logic in the `AbstractMessageProducingHandler` to deal with `Publisher` reply
* Fix `GroovySplitterTests` for the current code base
* Add `kotlinx.coroutines.flow.Flow` support
The `Flow` is essentially a multi-value reactive `Publisher`,
so use `ReactiveAdapterRegistry` to convert any custom reactive streams result to `Flux` and `Mono`
which we already support as reply types
* Add docs for `Kotlin Coroutines`
Rearrange the doc a bit extracting Kotlin support to individual `kotlin-functions.adoc` file
* Fix missed link to `reactive-streams.adoc` from the `index-single.adoc`
* Fix unintended Javadocs formatting in the `AbstractMessageProducingHandler`
* Add suspend functions support for Messaging Gateway
* Add convenient `CoroutinesUtils` for Coroutines types and `Continuation` argument fulfilling via `Mono`
* Treat `suspend fun` in the `GatewayProxyFactoryBean` as a `Mono` return
* Convert `Mono` to the `Continuation` resuming in the end of gateway call
* Document `suspend fun` for `@MessagingGateway`
* * Make `async` implicitly only for `suspend fun`
* * Remove unused imports
* * Verify sync and async `Flow` processing
* Mention default sync behavior in the docs
* * Improve reflection in the `CoroutinesUtils`
* Fix language in docs
Co-authored-by: Gary Russell <grussell@vmware.com>
* * Rebase and revert blank lines around `include` in docs
Co-authored-by: Gary Russell <grussell@vmware.com>
Fixes https://github.com/spring-projects/spring-integration/issues/3421
Remove `throws Exception;` from production code to honor
the rule `Generic exceptions should never be thrown` which is enabled on SonarQube
* Rework affected usages to `try..catch` with throwing respective runtime exception
or just logging
* Some other refactoring for the affected classes
* Migrate Cassandra extension project
* Add `spring-integration-cassandra` module based on the extension project
* Add Java DSL for Cassandra module
* Add documentation
* Add `CassandraContainerTest` based on a Testcontainers
* Fix language in docs
Co-authored-by: Gary Russell <grussell@vmware.com>
* * Fix `reactive-streams.adoc` for a proper link to
the new `spring-integration-cassandra` module
Co-authored-by: Gary Russell <grussell@vmware.com>
* Migrate Groovy DSL extension project
* Upgrade to Groovy `4.0.5`
* Apply `groovy` plugin for Groovy module to be able to compile Groovy classes for DSL
* Document new feature
* * Restore `lambdaWrapper` for the `GroovyIntegrationFlowDefinition.handle()`
to be able to preserve a generic argument type
* * Migrate more Spock test methods to JUnit style
* * Add missed `assert` to Groovy tests
* Fix language in docs
Co-authored-by: Gary Russell <grussell@vmware.com>
* * Fix Copyright and `@since` in new classes
* Remove deprecated `IntegrationFlows` class mentions in the `groovy-dsl.adoc`
Co-authored-by: Gary Russell <grussell@vmware.com>
The `xmppDomain` has two setters on the `XMPPTCPConnectionConfiguration.Builder`,
so the Java Bean specification may choose a wrong one by name and the
provided value may not fit into expectations
* Rework `XmppConnectionFactoryBeanTests-context.xml` to avoid ambiguity
with target setters presence
The `FixedBackOffPolicy` and `ExponentialBackOffPolicy` have now overloaded
setters for `Supplier` of value.
The XML parser relies on the Java Bean specification and cannot determine
the proper setter by type.
Looks like it is just resolved by name and there is no guarantee which one
will win
* Add inner `FixedBackOffPolicy` and `ExponentialBackOffPolicy` adapters
to expose `*Simple` setters for the values we get from the XML configuration
The plain `SocketAddress.toString()` may resolve
to `hostName/IP` pair which is not parsed properly via `URI.getHost()`
downstream leaving `/IP` part for `uri.getPath()` request.
In the end this may lead to wrong file name to be used in the SFTP logic
* Fix `SftpSession.getHostPort()` to use `SshdSocketAddress` utility
to resolve host and port properly from the provided `SocketAddress`
* INT-3333: Return empty list as is from DB gateway
Fixes https://jira.spring.io/browse/INT-3333
In `JdbcOutboundGateway` and `JpaOutboundGateway` the empty result list
is treated as "no reply" and therefore `null` is returned cause
the flow to stop at this point.
It is better to return such a result as is and the target application to
decided what to do with it, e.g. a `discardChannel` on a downstream splitter
configuration.
NOTE: the `MongoDbOutboundGateway` doesn't treat an empty result as a `null`
* * Fix language in docs
Co-authored-by: Gary Russell <grussell@vmware.com>
Co-authored-by: Gary Russell <grussell@vmware.com>
* GH-3664: Re-enable JavaScript support via GraalVM
Fixes https://github.com/spring-projects/spring-integration/issues/3664
The JavaScript JSR223 engine has been removed from Java.
Migrate JavaScript support into GraalVM Polyglot API:
* Implement `PolyglotScriptExecutor` which can also support other GraalVM languages
* Change the `ScriptExecutorFactory` to use a `PolyglotScriptExecutor` when it encounters JavaScript
* Re-enable tests for JavaScript
* Add GraalVM Polyglot support into docs
* * Reinstate `@EnabledIfSystemProperty` for GraalVM JS system property
The latest changes in Spring AMQP about migrating from RabbitMQ Hop
project to WebFlux has lead to dropping a Jackson deps which were
before transitive from Hop
* Add `com.fasterxml.jackson.core:jackson-databind` as tests dependency into AMQP module
Fixes https://jira.spring.io/browse/INT-2086
Some JMS vendors handle differently a `TemporaryTopic` and `TemporaryQueue`,
so, change a `JmsOutboundGateway` to create respective temporary destination
according the `replyPubSubDomain` option set
The `#args` and `#gatewayMethod` SpEL variables have been deprecated for a while
* Remove their population and usage in favor of `MethodArgsHolder` `root` of the evaluation context
This change optimize a gateway mapping logic the way that there is no need in evaluation context
for every call: we can just reuse a global one
* Some other `GatewayMethodInboundMessageMapper` code style refactoring
* Fix effected test classes and their configs
There is just enough to check the value type name against "kotlin.Unit" literal.
This way we don't need extra reflection bits to be exposed into a native image
It looks like `replyContainer()` is not registered as a bean,
so its lifecycle is somehow out of application context control
* Mark `replyContainer()` method as a `@Bean` in the `KafkaDslTests` and `KafkaDslKotlinTests`
to see if this fixes flaky state of the test suite
* Upgrade to Kotlin `1.7.20` and fix `KafkaDslKotlinTests` according to its requirements
* Add documentation for Observability
* Adapt Observation code to the latest dependencies
* Add doc generation tasks for meters and spans
* Document new Observation API features
* Include generated meters and spans docs to a general `metrics.adoc` chapter
* * Adapt `ObservationPropagationChannelInterceptorTests` for the latest `SpansAssert` API
* * Adjust to the latest Micrometer SNAPSHOT
* Make Observation doc generation tasks only as local.
We don't need ambiguous changes to source code on CI
* * Automate metrics/spans docs generation as a part of `reference` build phase
* Replace 'org.springframework.integration' content in the generated files with a 'o.s.i'
to make it easier to read, especially in the tables
* Break `DefaultMessageReceiverObservationConvention <=> IntegrationObservation` classes tangle
using literal for `KeyValues` in the `DefaultMessageReceiverObservationConvention`
instead of nested enums from the `IntegrationObservation`
* Some other minor build script clean up
* Fix indent in `build.gradle` for `micrometerVersion` property code line
* Add new line after observation section in whats-new.adoc
* * Adapt to the latest Micrometer changes
* * Use Reactor `2022.0.0-SNAPSHOT` version
Starting with Spring Data 2022.0.0 there is not going to be GemFire (Geode) support
directly from Spring team
A source code of the `spring-integration-gemfire` is moving to Spring Integration Extensions
from where community may consider to pull it and support in their own manner
* GH-3635: Add Future<Void> & Mono<Void> to gateway
Fixes https://github.com/spring-projects/spring-integration/issues/3635
When `Future<Void>` & `Mono<Void>` is used as a messaging gateway return type,
the application hangs out on this barrier which may lead to the out of memory eventually
* Add support for the `Future<Void>` & `Mono<Void>` messaging gateway return type
and ensure an asynchronous call for the `gateway.send(Message)` operation and
its exception handling.
In case of successful call, the `Future` is fulfilled with `null` and `Mono` is completed as empty
* * Check for `void.class` as well in the `GatewayProxyFactoryBean.isVoidReturnType`
* * Allow `Future<Void>` as a reply type of the gateway request-reply operation
* * Fix Checkstyle violations
* * Resolve `System.err.println()` in the test code
* Add `Thread.currentThread.interrupt()` to the `InterruptedException` block in the test
* Fix new classes and packages tangles
* Move `MessagingAnnotationPostProcessor` and `MethodAnnotationPostProcessor` impls
out of the `config.annotation` package due to usage of the `FactoryBean` configs
* Move `GenericHandler` and `GenericTransformer` to the `core` package to break
a tangle with a `LambdaMessageProcessor`
* Clean up affected tests and docs
* * Fix Checkstyle violation for imports order
