JIRA: https://jira.spring.io/browse/INT-3420
New documentation implied that if `expire-groups-upon-timeout`
was true, the group would NOT be removed if the group
was released by the release strategy in `forceComplete()`.
"...if it does so, then expiration is controlled by..."
In fact, `afterRelease()` (which is where `expire-groups-upon-completion`
is only invoked if `expire-groups-upon-timeout` is false.
Clarify that if the release strategy releases the group during timeout
then the group will always be expired if `expire-groups-upon-timeout` is
true (default).
JIRA: https://jira.spring.io/browse/INT-3104
Add phase and remove unnecessary auto-starup
INT-3104: Remove reduntant code and format
Polishing and fixing parser tests
JIRA: https://jira.spring.io/browse/INT-3464
Previously, if an IO exeption occurred while synchronizing
files, and an AcceptOnceFileListFilter is being used, files that were
not transferred would not be fetched next time.
Add strategy `ReversibleFileListFilter` that can rollback previously
accepted files.
Implement this interface on `AcceptOnceFileListFilter` and
`AbstractPersistentAcceptOnceFileListFilter`.
Add test cases.
Polishing
JIRA: https://jira.spring.io/browse/INT-3474
When the MethodInvokingMessageHandler skips ineligible methods while
searching it emits a stack trace under DEBUG logging. This is not
needed, a simple log message is enough to convey the information.
Some POJOs gain ineligible methods when wrapped in a proxy so this
is no fault of the developer.
Other reasons for skipping methods (improperly set annotations etc)
continue to emit a stack trace.
INT-3474: Polishing
Rethrow `IneligibleMethodException` as `IllegalArgumentException` in case of a single `method` for processor
JIRA: https://jira.spring.io/browse/INT-3471
- Make final fields protected so they are available to subclasses
- Add property `removeOnGet` allowing the map entry to be removed immediately when it is used
- Add `time-to-live-expression` allowing override of the reaper delay
Polishing
Conflicts:
src/reference/docbook/whats-new.xml
JIRA: https://jira.spring.io/browse/INT-3465
Add support for adding/removing individual recipients to the RecipientListRouter
Modify documentation in what's new and spring-integration-4.1.xsd
Polishing
`AbstractRemoteFileOutboundGateway`: close `outputStream` before `file.delete()` to release exclusive file-lock
JIRA: https://jira.spring.io/browse/INT-3420
Add option to allow the empty group to remain after timeout
so late arriving messages can be discarded.
INT-3420 Add Callout Hyperlinks
The aggregator configuration documentation had hyperlinks
from the attribute descriptions to the attribute in the
XML, but not vice-versa. For a large number of attributes
such as this, bi-directional hyperlinks are useful.
INT-3420 Doc Polishing
JIRA: https://jira.spring.io/browse/INT-3475
Change the phase for `SourcePollingChannelAdapterFactoryBean`, `AbstractPollingEndpoint`, `MessageProducerSupport`, `JmsMessageDrivenEndpoint`
to the `Integer.MAX_VALUE / 2`
JIRA: https://jira.spring.io/browse/INT-2438
If a get operation fails in the outbound gateway,
clean up (remove) the temporary file if present.
Test Polishing
JIRA: https://jira.spring.io/browse/INT-3473
A simple fix to prevent `Selector closed` error loging message on `TcpNioServerConnectionFactory#stop()`
Polishing
JIRA: https://jira.spring.io/browse/INT-3470
* Apply the last `ChannelInterceptor` changes - just `extends ChannelInterceptorAdapter`
* Change some tests to check content according to the new changes to the `GenericMessage#toString()`
**Cherry-pick to 4.0.x**
JIRA: https://jira.spring.io/browse/INT-3467
Add methods and managed operations to the inbound
channel adapter to allow the subscribed topics to be
changed at runtime, programmatically, or using a
control-bus.
JIRA: https://jira.spring.io/browse/INT-3468
Provide an option to not block when sending and emit
events for sends and delivery confirmations.
Also use the async client for the inbound adapter; while
it doesn't make any performance difference, it does
allow us to timeout the disconnect, which we have seen
to cause hangs on the CI servers.
INT-3468 Polishing - Use Events; Add Docs
INT-3468 Doc Polishing
INT-3468 More Polishing - PR Comments
- Only emit delivered event if async
- Add clientId and a new instance counter to events
INT-3468 Polishing
- Pull client instance up to the abstract class and remove references to the Paho implementation from the events
- Improve tests to include a second client
INT-3468 Fix Test Case
Incorrect classname meant the default location for the application
context config was not found on case-sensitive file systems.
Also, don't auto-start the adapters in the context, in case
the broker is not running.
INT-3468: Polishing
INT-3468 Fix Package Tangle; Add 'async-events'
Add an option (default false) to emit events when async is true.
JIRA: https://jira.spring.io/browse/INT-3469
Previously, the callback was set immediately after
subscribing which caused a small possibility of
message loss.
Set the callback before subscribing.
JIRA: https://jira.spring.io/browse/INT-3434
Previously `PointToPointSubscribableAmqpChannel` declared Queue always.
Change the logic to use `admin.declareQueue` only if `queueName` isn't specified.
In addition it is redundant to get deal with entire `Queue` object - just use the `queueName` for the `SimpleMessageListenerContainer`
Polishing
Declare queue the old way if not already present in broker,
even if the name is supplied.
JIRA: https://jira.spring.io/browse/INT-3448
INT-3448: Addressing PR comments
Late Resolution of @Publisher.defaultChannel
Early resolution of the publisher default channel caused
`@ Configuration` factory beans to be instantiated before
Spring Integration bean post processors had been registered.
Use the channel name instead of channel reference and resolve
it when processing the first publish.
Merge branch 'INT-3448' of https://github.com/garyrussell/spring-integration into INT-3448
Conflicts:
spring-integration-core/src/test/java/org/springframework/integration/configuration/EnableIntegrationTests.java
INT-3448: Polishing
JIRA: https://jira.spring.io/browse/INT-3461
To test it: `gradlew :spring-integration-http:testAll -PspringVersion=4.1.0.BUILD-SNAPSHOT`
INT-3461: Add CP condition for `rometools` into the `HttpRequestHandlingEndpointSupport`
INT-3461: Add CLASSPATH check for `RequestEntity` to determine the SF 4.1
JIRA: https://jira.spring.io/browse/INT-3404
Previously, `AbstractInboundFileSynchronizingMessageSource` implemented both `MessageProducer` and `MessageSource`.
The `MessageProducer` is really redundant and confusing.
Also, it might produce some side-effects when using JavaConfig.
Provide some other simple but important fix for all `AbstractMessageSource` implementation to invoke `super.afterPropertiesSet()`
Polishing
Add mock bean factory to failing tests.
Revert state assertions in AbstractInboundFileSynchronizingMessageSource;
the fields are state, not arguments, at the time they are being
asserted.
JIRA: https://jira.spring.io/browse/INT-3458
* Change Spring AMQP to `1.3.5.RELEASE`
* `HttpRequestHandlingEndpointSupport`: remove `MappingJacksonHttpMessageConverter` registration
* `IntegrationRequestMappingHandlerMapping`: add 'fake' `name()` attribute to the inline `RequestMapping` annotation
* `StoredProcJmxManagedBeanTests`: remove `context.stop();` code, Since `MBeanExporter` deregister MBeans on `stop()` now
* `StoredProcPollingChannelAdapterParserTests`: change deprecated `ParameterizedSingleColumnRowMapper` to the `SingleColumnRowMapper`
* `Jms`: comment out the reflection code to check the value for the `recoveryInterval`, because it is removed already in favor of `backOff`
* `NotificationListeningMessageProducer`: move the start-up listener registration to the `onApplicationEvent`,
because `MBeanExporter` moved `registerBeans()` to the `start()` now.
The same `phase` might cause the issue, that MBeans aren't registered yet for `NotificationListeningMessageProducer`
* `JpaOutboundGatewayTests`: change `@TransactionConfiguration` to the `@Transactional`. Don't know why the first doesn't work now.
**Cherry-pick to 4.0.x**
INT-3458: Addressing PR comments
`NotificationListeningMessageProducer`: defer listener registration until `onApplicationEvent()`
Fixed hung assembler described in INT-3453.
Used CountDownLatch instead of sleep()
INT-3453 Add Test Case
Currently failing for me - assembler thread stuck on CDL.
Remove redundant `if (TcpNioConnection.this.writingLatch != null)` from `ChannelInputStream#write`
JIRA: https://jira.spring.io/browse/INT-3433`
If a read fails due to insufficient threads, delay the read
for (default 100ms) - do not re-enable OP_READ until that
time has elapsed. Avoids spinning the CPU.
INT-3433 More Polishing
If the assembler couldn't execute a new assembler after assembling
the current message (when it detected there is more data), in the
finally block it would "continue" only if the socket was still
open.
The test case closes the socket after sending 4 messages so when
this condition occurred, the assembler failed to continue and
data was left in the buffer.
Remove the `isOpen()` check in the finally block and always continue
if there's not another assembler running and there's data available.
INT-3433 More Polishing
We can still get starvation if the selector is in a long
wait (in select()) when a read is delayed.
Whenever a read is delayed, wake the selector so its next
select will use the readDelay timeout.
INT-3433 Reference Docs
Also remove Thread.yield().
Revert redundant boolean return from `TcpNioConnection#checkForAssembler()`
JIRA: https://jira.spring.io/browse/INT-3442
Previously `RedisQueueMessageDrivenEndpoint.stop()` returned immediately.
It caused an issue when one more `message` might be read and processed to stopped app.
* Introduce `AbstractEndpoint#lifecycleCondition` and wait on it from `RedisQueueMessageDrivenEndpoint.doStop()`
and `signal()` it from `ListenerTask`.
* Since everything is done around `lifecycleLock` the `RedisQueueMessageDrivenEndpoint.stop()` waits for the proper
'last' message process.
* Add tiny `Thread.sleep(1)` to the `popMessageAndSend` cycle to free `lifecycleLock` for other Threads, e.g. `stop()`
INT-3442: Get rid of `lock` from listener cycle
* Introduce `stopTimeout` to minimize the `stop` thread blocking
* If the message is returned after that timeout it moved back to Redis List using `RPUSH`
INT-3442: Addressing PR comments
