GH-10083: Apply Nullability to some core support packages #10360
Conversation
...qp/src/main/java/org/springframework/integration/amqp/inbound/AmqpInboundChannelAdapter.java
Outdated
Show resolved
Hide resolved
ngocnhan-tran1996
force-pushed
the
gh-10083-core-support-packages
branch
from
39a4b45 to
a1437de
Compare
...rc/main/java/org/springframework/integration/support/channel/BeanFactoryChannelResolver.java
Outdated
Show resolved
Hide resolved
...ation-core/src/main/java/org/springframework/integration/support/context/NamedComponent.java
Outdated
Show resolved
Hide resolved
|
Th fix for AMQP is in. |
Related to: spring-projects#10083 This commit applies Nullability to `support.channel`, `support.context` and `support.converter` packages Signed-off-by: Tran Ngoc Nhan <[email protected]>
Signed-off-by: Tran Ngoc Nhan <[email protected]>
Signed-off-by: Tran Ngoc Nhan <[email protected]>
ngocnhan-tran1996
force-pushed
the
gh-10083-core-support-packages
branch
from
649983e to
a0ce80d
Compare
Signed-off-by: Tran Ngoc Nhan <[email protected]>
| @Nullable | ||
| public String getComponentType() { | ||
| return null; | ||
| return "integration-object-support"; |
There was a problem hiding this comment.
Not very good.
I believe we are in a situation where we should remove this method implementation here in the IntegrationObjectSupport.
And ensure that it is implemented properly everywhere else.
Signed-off-by: Tran Ngoc Nhan <[email protected]>
|
|
||
| @Override | ||
| public String getComponentType() { | ||
| return "channel"; |
There was a problem hiding this comment.
header-channel-registry.
I can argue that this class is not supposed to be an IntegrationObjectSupport.
But that's different story.
|
|
||
| @Override | ||
| public String getComponentType() { | ||
| return "converter"; |
There was a problem hiding this comment.
Same here: this is not an IntegrationObjectSupport.
But separate story.
|
|
||
| @Override | ||
| public String getComponentType() { | ||
| return "endpoint"; |
There was a problem hiding this comment.
I think this is not correct: every each AbstractEndpoint must provide their proper implementation.
|
|
||
| @Override | ||
| public String getComponentType() { | ||
| return "connection"; |
|
|
||
| @Override | ||
| public String getComponentType() { | ||
| return "processor"; |
There was a problem hiding this comment.
This one also must not be an IntegrationObjectSupport.
Signed-off-by: Tran Ngoc Nhan <[email protected]>
ngocnhan-tran1996
force-pushed
the
gh-10083-core-support-packages
branch
from
44debbc to
70e49b9
Compare
|
|
||
| @Override | ||
| public String getComponentType() { | ||
| return "polling-endpoint"; |
There was a problem hiding this comment.
The SourcePollingChannelAdapter does provide an impl of this method.
So, another one - PollingConsumer - has to do that as well: polling-consumer.
|
|
||
| @Override | ||
| public String getComponentType() { | ||
| return "event-driven"; |
There was a problem hiding this comment.
Well, event-driven-consumer to be precise.
|
|
||
| @Override | ||
| public String getComponentType() { | ||
| return "message-producer"; |
There was a problem hiding this comment.
I think neither of abstract class should provide an impl.
That is really responsibility of the specific one.
Without an implementation here, we may spot a bug somewhere down the hierarchy, e.g. ReactiveMessageSourceProducer.
|
|
||
| @Override | ||
| public String getComponentType() { | ||
| return "reactive-streams"; |
|
Merged as: 17e6a31 |
Related to: #10083
This commit applies Nullability to
support.channel,support.contextandsupport.converterpackages