Hi. I have a KafkaListener like this:

@KafkaListener(topics = "topic", groupId = "groupId")
public void listen(List<Order> orders) {
	// Magic
}

Now if I include swagger4kafka: (Need to exclude logback and springfox-core because I'm not using logback and a more recent version of springfox 2.9.2).

<dependency>
    <groupId>io.github.stavshamir</groupId>
    <artifactId>swagger4kafka</artifactId>
    <version>0.0.1</version>
    <exclusions>
        <exclusion>
            <groupId>ch.qos.logback</groupId>
            <artifactId>logback-classic</artifactId>
        </exclusion>
        <exclusion>
            <groupId>ch.qos.logback</groupId>
            <artifactId>logback-core</artifactId>
        </exclusion>
        <exclusion>
            <groupId>io.springfox</groupId>
            <artifactId>springfox-core</artifactId>
        </exclusion>
    </exclusions>
</dependency>
<dependency>
    <groupId>io.github.stavshamir</groupId>
    <artifactId>swagger4kafka-ui</artifactId>
    <version>0.0.1</version>
</dependency>

When starting up the spring boot application I get the following error:
{"instant":{"epochSecond":1589899103,"nanoOfSecond":976000000},"thread":"main","level":"ERROR","loggerName":"io.github.stavshamir.swagger4kafka.services.ModelsService","message":"Model for List was not found","endOfBatch":false,"loggerFqcn":"org.apache.logging.slf4j.Log4jLogger","threadId":1,"threadPriority":5,"source":{"class":"io.github.stavshamir.swagger4kafka.services.ModelsService","method":"getExample","file":"ModelsService.java","line":110}}

I can contribute on this, but I need a hint in the right direction. Does Swagger not have default models for standard types like lists or sets that could be used here?

Now my generation document on http://localhost:8086/springwolf/docs is:
{ "myproject": { "asyncapi": "2.2.0" "info": { .... } ...... } }
The root node's "myproject" name is from part of AsyncApiConfig:

    @Bean
    public AsyncApiDocket asyncApiDocket() {
        Info info = Info.builder()
                .version("1.0.0")
                .title("myproject")
                .build();

If I set title to null, NullPointerException hapens.

When i try to generate client schema by AsyncApi generator
# ag http://localhost:8086/springwolf/docs @asyncapi/java-spring-template
I get the error:

Something went wrong:
Error: The `asyncapi` field is missing.
    at parse (/usr/local/lib/node_modules/@asyncapi/generator/node_modules/@asyncapi/parser/lib/parser.js:72:13)
    at Generator.generateFromString (/usr/local/lib/node_modules/@asyncapi/generator/lib/generator.js:265:28)
    at Generator.generateFromURL (/usr/local/lib/node_modules/@asyncapi/generator/lib/generator.js:298:17)
    at processTicksAndRejections (node:internal/process/task_queues:96:5)
    at /usr/local/lib/node_modules/@asyncapi/generator/cli.js:156:9

In the code of parser.js exception happens if there is no field "asyncapi" in Json (at first level). How can I generate the asyncapi document without root project title tag?

when tried adding producer data for AsyncApiConfiguration.java for AMQP example I'm getting issue for binding

kindly help with the configuration for producer in AMQP(RMQ)

Add the minimal required types for producing an asyncapi doc for kafka:

• AsyncAPI
• Info
• Server
• Channel
• Operation
• OperationBinding
• KafkaOperationBinding
• Message
• Components

The classes should be created in io.github.stavshamit.swagger4kafka.types.asyncapi.

Further reading:
https://www.asyncapi.com/docs/specifications/2.0.0/

Describe the bug

1. We have a Spring Boot microservice, let's say it is called "test-service"
2. "test-service" has the following consumers:
   2.1 ConsumerA -> consumers "topic-a"
   2.2 ConsumerB -> consumer "topic-b"
3. Now let's say we configure an asyncApiDocket Bean. Everything works fine and I can see both consumers in springwolf ui.
4. Now we add ProducerA -> produces events to "topic-a'.
5. Now springwolf ui display only consumer for topic-b and producer for topic-a. Seems the consumer for topic-a documentation was overwritted.

Dependencies and versions used
springwolf-kafka version 0.6.1. I tried with 0.7.0 as well.

Code example

@SpringBootApplication(scanBasePackages = {})

public class ExampleApplication {

    public static void main(String[] args) {

        SpringApplication.run(ExampleApplication.class, args);
    }
}

@Component
@RequiredArgsConstructor
@Slf4j
@KafkaListener(
    batch = "true",
    topics = "topic-a"
)
public class ConsumerA {
    @KafkaHandler
    public void consume(Object event) {
    }
}

@Component
@RequiredArgsConstructor
@Slf4j
@KafkaListener(
    batch = "true",
    topics = "topic-b"
)
public class ConsumerB {
    @KafkaHandler
    public void consume(Object event) {
    }
}

@EnableAsyncApi
@Configuration
public class AsyncApiConfigurationTest {

    private static final String LISTENERS_SCAN_PACKAGE =
        "example.consumer";

    @Bean
    public AsyncApiDocket asyncApiDocket(
        @Value("${service.name}") String appName,
        @Value("${service.version}") String appVersion,
        @Value("${spring.kafka.bootstrap-servers}") String bootstrapServers) {

        Info info = Info.builder()
            .version(appVersion)
            .title(appName)
            .description("description")
            .build();

        String producerTopicName = "topic-a";
        ProducerData preparedDataProducer = ProducerData.builder()
            .channelName(producerTopicName)
            .operationBinding(ImmutableMap.of("kafka", new KafkaOperationBinding()))
            .payloadType(FetchUserBalances.class)
            .build();

        return AsyncApiDocket.builder()
            .basePackage(LISTENERS_SCAN_PACKAGE)
            .server("kafka", Server.builder().protocol("kafka").url(bootstrapServers).build())
            .producer(preparedDataProducer)
            .info(info)
            .build();
    }

Stack trace and error logs
No exceptions in log

The Producers should also include the "oneOf" in the "message" property of the generated Async api file.

Currently adding two producer on the same topic in the AsyncApiDocket configuration, results to one overriding the other and only the last one is included.

Here's the configuration that results in such an issue. You can see that the searchProducer and the retrieveProducer variables have the same topic.

@Bean
 public AsyncApiDocket asyncApiDocket(KafkaProperties kafkaProperties) {
        Info info = Info.builder()
                .version(asyncApiVersion)
                .title("Party Adapter Async API")
                .build();

        Server kafkaServer = Server.builder()
                .protocol("kafka")
                .url(String.join(",", kafkaProperties.getBootstrapServers()))
                .build();

        ProducerData searchProducer = ProducerData.builder()
                .channelName(replyTopic)
                .payloadType(PartySearchResultPayload.class)
                .binding(ImmutableMap.of("kafka", new KafkaOperationBinding()))
                .build();

        ProducerData retrieveProducer = ProducerData.builder()
                .channelName(replyTopic)
                .payloadType(PartyRetrieveResultPayload.class)
                .binding(ImmutableMap.of("kafka", new KafkaOperationBinding()))
                .build();

        return AsyncApiDocket.builder()
                .basePackage(BASE_PACKAGE_SCAN)
                .info(info)
                .producers(List.of(searchProducer, retrieveProducer))
                .server("kafka", kafkaServer)
                .build();
}

In the previous release you supported the @KafkaHandler that results in having "oneOf" property in the generated json. The producer also needs that.

Thank you

when trying to publish from the example tab in the asyncui.html getting error stating publish failed please find the attached screen shot

After adding AsyncAPI configuration to the RMQ application I'm getting the below error

Motivation

Currently, the project uses in-house implementation of the various asyncapi types (such as Channel, Operation etc.).
This in-house implementation is problematic due to:

1. There are many bindings missing for most of the protocols. Actually only kafka bindings were implemented. For more plugins to be created, we need the bindings for them.
2. It repeats the great job already done by @Pakisan in jasyncapi.

There for, it would be a great step to replace all those in-house implementations (from asyncapi/types package) with the types provided by jasyncapi-core.

Tasks

• Add jasyncapi-core dependency
• Replace all in-house implementations with provided types from jasyncapi. Many have the same names so the switch should be rather smooth, but I did rely on some custom builders, constructors and json serializations that may require additional attention.
• Fix the tests and verify all pass
• Remove local asyncapi/tests package (except for ProducerData, which may not belong there, but its future location may be decided later on).
• Release local snapshot version, and update the kafka plugin. After it is updated, I will release an actual version.

Is it possible to provide additional description to the fields in a message payload? For example, could I use some annotation to describe the field, provide an example value or define whether it is required or not?

For example, with Swagger annotations, I could do something like

class Payload {
    @ApiModelProperty(value = "my description", example = "my example")
    private final String property;
}

Now that support for @KafkaHandler has been added in 0.4.0, it would be nice to add such an example to the kafka plugin example project.
This would require to:

• Add a class with two @KafkaHandler methods, with a non-primitive payload (can use the existing example DTOs).
• Since this project also serves as integration test, update the expected async api doc with the new channel.
• Verify that publishing from the UI works.

Hi,

I just added the dependencies in a project with spring webflux, and I ran into the following error

Hi everyone who develops this tool!
I've tried to use this tool on my production project to make useful API for Rabbit AMQP that we expose internally.
But found that in fact library covers at most 10% of AsyncAPI specification
Also I found that for some reason it was designed in a way that restricts any extension of existing "Services/Providers/Models", and all used models are not from async-api dependency but own-created and they doesn't hold enough properties to match AsyncAPI Spec.
Also it is not possible to extend Generated Spec with some custom descriptions of operations/parameters (e.g. via annotations).

So, looking at this facts I have single question - is it planned to support this project and make it usable not only for "Hello World" application?

Kind regards,
Pavlo

Describe the bug
The RabbitChannelsScanner seems not to support Queues with multiple RoutingKeys bound, which is funny :D

Dependencies and versions used
springwolf-amqp version 0.3.1. (thows Exception)
springwolf-amqp version 0.2.0. (ignores 2nd Binding)

Code example
One queue, two bindings.

    @Bean
    @Qualifier("featured-list")
    public Queue featuredListQueue(@Value("${featured.items.list.queue}") String queue){
        return QueueBuilder.durable(queue).quorum().build();
    }

    @Bean
    public Binding bindfeaturedImplicit(@Qualifier("channel") Exchange channelExchange, @Qualifier("featured-list") Queue featuredListQueue){
        return BindingBuilder.bind(featuredListQueue).to(channelExchange).with("channel.listMy.rk").noargs();
    }

    @Bean
    public Binding bindfeaturedExplicit(@Qualifier("featured") Exchange featuredExchange,
                                        @Qualifier("featured-list") Queue featuredListQueue,
                                        @Value("${featured.items.list.rk}") String rk){
        return BindingBuilder.bind(featuredListQueue).to(featuredExchange).with(rk).noargs();
    }

Stack trace and error logs
If an exception has been thrown or an error was logged by springwolf.

java.lang.IllegalStateException: Failed to load ApplicationContext
Caused by: org.springframework.beans.factory.BeanCreationException: 
	Error creating bean with name 'rabbitChannelsScanner' defined in URL [jar:file:/C:/Users/.../.m2/repository/io/github/springwolf/springwolf-amqp/0.3.1/springwolf-amqp-0.3.1.jar!/io/github/stavshamir/springwolf/asyncapi/scanners/channels/RabbitChannelsScanner.class]: ...
Caused by: org.springframework.beans.BeanInstantiationException: ...
Caused by: java.lang.IllegalStateException: 
	Duplicate key featured.items.list.queue (attempted merging values 
	Binding [destination=featured.items.list.queue, exchange=channel.exchange, routingKey=channel.listMy.rk, arguments={}] and
	Binding [destination=featured.items.list.queue, exchange=featured.exchange, routingKey=featured.items.list.rk, arguments={}])

Describe the feature request
A clear and concise description of what you want to happen.

The "newer" implementation of rabbit/spring cloud stream listeners makes use of spring cloud functions to consume messages with the binding between queues and the method matching on the method name with properties defined in application properties. This library should be able to automatically find these consumers and produce documentation for them.

Motivation
Please describe the motivation for adding this feature (what do you feel is missing, why do you think this enhancement is a good idea etc.)

This is required for continued support of automatically documenting consumers using the more-modern definition pattern.

Technical details
If possible, add technical details like design, implementation, etc. Any idea is welcome!

"Consumer" methods are defined as beans as follows:
@bean
public Consumer messageReceiptConsumer() {
}

queue/binding information is defined in application.properties in the following way:
spring.cloud.stream.bindings.exampleNewConsumerMethod-in-0.destination=exchangeName
spring.cloud.stream.bindings.exampleNewConsumerMethod-in-0.group=example-new-bindings-queue
spring.cloud.stream.rabbit.bindings.exampleNewConsumerMethod-in-0.consumer.bindingRoutingKey=binding.new.routing.key

note: not all of these properties are required and spring has a way of automatically creating queue/exchange names when not present. I don't know for sure, but I believe at least one of these properties is required for binding to occur.

More information on the spring properties supported can be found here:
https://docs.spring.io/spring-cloud-stream/docs/current/reference/html/spring-cloud-stream-binder-rabbit.html#spring-cloud-stream-binder-rabbit-reference

Describe alternatives you've considered
Supporting manual definition of consumers like producers are defined.

as Asyncapi supports SQS: https://www.asyncapi.com/docs/specifications/v2.3.0, it would be nice, if the tool could handle it as well.

when added producer configuration and trying to publish message from the ui it isn't validating the queue please help me to understand a way in validating the queue

Caused by: org.springframework.beans.BeanInstantiationException: Failed to instantiate [com.stavshamir.swagger4kafka.services.KafkaEndpointsService]: Constructor threw exception; nested exception is java.lang.NullPointerException
	at org.springframework.beans.BeanUtils.instantiateClass(BeanUtils.java:217) ~[spring-beans-5.2.5.RELEASE.jar:5.2.5.RELEASE]
	at org.springframework.beans.factory.support.SimpleInstantiationStrategy.instantiate(SimpleInstantiationStrategy.java:117) ~[spring-beans-5.2.5.RELEASE.jar:5.2.5.RELEASE]
	at org.springframework.beans.factory.support.ConstructorResolver.instantiate(ConstructorResolver.java:310) ~[spring-beans-5.2.5.RELEASE.jar:5.2.5.RELEASE]
	... 19 common frames omitted
Caused by: java.lang.NullPointerException: null
	at com.stavshamir.swagger4kafka.services.ModelsService.getEnumProperties(ModelsService.java:89) ~[swagger4kafka-1.0.0.jar:na]
	at com.stavshamir.swagger4kafka.services.ModelsService.setDeserializableEnumValues(ModelsService.java:84) ~[swagger4kafka-1.0.0.jar:na]
	at com.stavshamir.swagger4kafka.services.ModelsService.setDeserializableEnumValues(ModelsService.java:63) ~[swagger4kafka-1.0.0.jar:na]
	at com.stavshamir.swagger4kafka.services.ModelsService.register(ModelsService.java:51) ~[swagger4kafka-1.0.0.jar:na]
	at com.stavshamir.swagger4kafka.services.KafkaListenersScanner.topicToEndpoint(KafkaListenersScanner.java:65) ~[swagger4kafka-1.0.0.jar:na]
	at com.stavshamir.swagger4kafka.services.KafkaListenersScanner.lambda$createKafkaEndpoints$2(KafkaListenersScanner.java:48) ~[swagger4kafka-1.0.0.jar:na]
	at java.base/java.util.stream.ReferencePipeline$3$1.accept(ReferencePipeline.java:195) ~[na:na]

Version used:

 <dependency>
            <groupId>com.stavshamir</groupId>
            <artifactId>swagger4kafka</artifactId>
            <version>1.0.0</version>
        </dependency>
        <dependency>
            <groupId>com.stavshamir</groupId>
            <artifactId>swagger4kafka-ui</artifactId>
            <version>1.0.0</version>
        </dependency>

Springboot version:

    <parent>
        <groupId>org.springframework.boot</groupId>
        <artifactId>spring-boot-starter-parent</artifactId>
        <version>2.2.6.RELEASE</version>
        <relativePath/> <!-- lookup parent from repository -->
    </parent>

getting this error while running the app, added below dependencies and then also added Async configuration

Caused by: org.springframework.beans.factory.UnsatisfiedDependencyException: Error creating bean with name 'defaultAsyncApiService' defined in URL [jar:file:/home/vcap/app/BOOT-INF/lib/springwolf-core-0.2.0.jar!/io/github/stavshamir/springwolf/asyncapi/DefaultAsyncApiService.class]: Unsatisfied dependency expressed through constructor parameter 1; nested exception is org.springframework.beans.factory.BeanCreationException: Error creating bean with name 'defaultChannelsService': Invocation of init method failed; nested exception is java.lang.IllegalArgumentException: Only single parameter KafkaListener methods are supported

Thanks in Advance

Hello,
maybe it would bring a benefit if we change the GroupId groupId to a more generic type in order to fulfill also other requirements like the clientId.

What do you think about it?

Greetings
Chris

Hi!
Amazing project

I build and reactive api with spring webflux and kafka, I have an issue when I implement springwolf-core to document kafka listeners and WebTestClient for my tests, the log is next

Caused by: org.springframework.beans.factory.NoSuchBeanDefinitionException: No qualifying bean of type 'org.springframework.test.web.reactive.server.WebTestClient' available: expected at least 1 bean which qualifies as autowire candidate. Dependency annotations: {@org.springframework.beans.factory.annotation.Autowired(required=true)}
at org.springframework.beans.factory.support.DefaultListableBeanFactory.raiseNoMatchingBeanFound(DefaultListableBeanFactory.java:1777)
at org.springframework.beans.factory.support.DefaultListableBeanFactory.doResolveDependency(DefaultListableBeanFactory.java:1333)
at org.springframework.beans.factory.support.DefaultListableBeanFactory.resolveDependency(DefaultListableBeanFactory.java:1287)
at org.springframework.beans.factory.annotation.AutowiredAnnotationBeanPostProcessor$AutowiredFieldElement.inject(AutowiredAnnotationBeanPostProcessor.java:640)
... 87 more

The error looks like a conflict with swagger-inflector
This is my test class

@SpringBootTest(classes = [TestConfiguration::class])
@ExtendWith(SpringExtension::class)
@AutoConfigureWebTestClient
class CrudControllerTest {
    @Autowired
    private lateinit var client: WebTestClient
    ...    
}

I have an already existing spring boot application that listens to RabbitMQ queue. It consumes a specific message (wrapped in CloudEvent) under specific topic. Can this generator assist me in exposing it as AsyncApi? Does it respect CloudEvent object?

I tried several times with rabbit listener but channels was not recognized by spring-wolf.
the http://localhost:8081/springwolf/asyncapi-ui.html opens up fine, showing Server and Info, but not Channels.
In RabbitMQ Management console, channels are fine with each queue name

I did with springwolf 0.4.0, 0.3.1 but I found that documentation states that latest support for RabbitMQ is 0.2.0
So, tried with 0.2.0, some implementations being different(i.e. basePackage() doesn't exist in AsyncApiDocket), also not applicable.

additionally, I cannot find any debug log after this:
......i.g.s.s.asyncapi.DefaultAsyncApiService : Building AsyncAPI document

any test being done lately??

Our services typically have used @KafkaListener on the class declaration and @KafkaHandler on the method declarations within those classes. Springwolf-kafka does not seem to pick these up. I modified the code in one of my handler classes to only use @KafkaListener on the method declarations, and springwolf picks them up correctly.

I don't know whether this is considered an enhancement or a bug, but it'd be super nifty if this worked either way.

Create AsyncApiService, retrieved will be responsible to instantiate the AsyncAPI object.

• User input should be received through an AsyncApiDocket bean.
• Kafka Listener data should be converted to asynapi objects and retrieved from KafkaListenersScanner, which should be renamed to KafkaChannelsService.

Description

Currently, listeners are mapped as subscribe channels, and producers are mapped as publish channels. Apparently, it should be the other way around.
See https://www.asyncapi.com/blog/publish-subscribe-semantics.

Thanks to https://github.com/kalarani for notifying me.

Fix

1. The subscribe operation in AbstractChannelScanner::buildChannel should be replace with publish.
2. Similarly, publish should be switched to subscribe in the ProducerChannelScanner
3. Recommended - add a unit test for AbstractChannelScanner and ProducerChannelScanner. For AbstractChannelScanner, this would require creating an inner class in the unit test the implements the abstract methods as simple as possible.
4. After the fix in the core project, I will release a new version, and then the tests in kafka plugin project should be fixed. Also, the UI project should be fixed accordingly.

Hello, it seems that the multiple producer support does not work, it was fixed on this issue #59

Im attaching the project that does not work with multiple producers.

async-api-project.zip

Calling http://localhost:8080/springwolf/docs will bring the following json, but only one producer is in the json!

{
  "Async-api-test": {
    "asyncapi": "2.0.0",
    "info": {
      "title": "Async-api-test",
      "version": "0.0.1"
    },
    "servers": {
      "kafka": {
        "url": "localhost:9092",
        "protocol": "kafka"
      }
    },
    "channels": {
      "ACCOUNT": {
        "subscribe": {
          "bindings": {
            "kafka": {
              
            }
          },
          "message": {
            "name": "com.example.asyncapi.model.ClassB",
            "title": "ClassB",
            "payload": {
              "$ref": "#/components/schemas/ClassB"
            }
          }
        }
      }
    },
    "components": {
      "schemas": {
        "ClassA": {
          "type": "object"
        },
        "ClassB": {
          "type": "object"
        }
      }
    }
  }
}

Thanks

Describe the bug
When there multiple consumer group-id consuming the same topic with different payloads, ClassLevelKafkaListenerScanner throws an exception while scanning due to duplicate key of topics.

Dependencies and versions used
springwolf-kafka version 0.7.0.

Stack trace and error logs
i.g.s.s.asyncapi.DefaultChannelsService : An error was encountered during channel scanning with io.github.stavshamir.springwolf.asyncapi.scanners.channels.ClassLevelKafkaListenerScanner@1dd76982: Duplicate key CUSTOMER_PRODUCT_ADPT (attempted merging values ChannelItem(description=null, subscribe=null, publish=Operation(operationId=null, summary=null, description=null, tags=null, externalDocs=null, bindings={kafka=KafkaOperationBinding(groupId=GroupA, clientId=null, bindingVersion=null)}, traits=null, message={oneOf=[Message(name=java.lang.Object, title=Object, payload=io.github.stavshamir.springwolf.asyncapi.types.channel.operation.message.PayloadReference@1), Message(name=com.example.PayloadA, title=PayloadA, payload=io.github.stavshamir.springwolf.asyncapi.types.channel.operation.message.PayloadReference@1)]}), parameters=null, bindings={kafka=KafkaChannelBinding()}) and ChannelItem(description=null, subscribe=null, publish=Operation(operationId=null, summary=null, description=null, tags=null, externalDocs=null, bindings={kafka=KafkaOperationBinding(groupId=GroupB, clientId=null, bindingVersion=null)}, traits=null, message={oneOf=[Message(name=java.lang.Object, title=Object, payload=io.github.stavshamir.springwolf.asyncapi.types.channel.operation.message.PayloadReference@1), Message(name=com.example.PayloadB, title=PayloadB, payload=io.github.stavshamir.springwolf.asyncapi.types.channel.operation.message.PayloadReference@1)]}), parameters=null, bindings={kafka=KafkaChannelBinding()}))

is there a way in making the specs independent of application, which means if the application is down can i able to view the specs?

Thank you for awesome work with this great project!
One issue found related to producers ui. Those are not working. Details below.
https://github.com/stavshamir/springwolf/blob/master/springwolf-ui/src/app/shared/asyncapi.service.ts
line:95 const isSubscribe = subscribe !== null;
when "subscribe" is null in back-end, in generated api spec json it is empty.
subscribe !=null, but undefined, so failing further.

There is an IndexOutOfBoundsException when using RabbitListener bindings instead of queues.

Here is the stack trace :

Caused by: java.lang.IndexOutOfBoundsException: Index 0 out of bounds for length 0
	at java.base/jdk.internal.util.Preconditions.outOfBounds(Preconditions.java:64)
	at java.base/jdk.internal.util.Preconditions.outOfBoundsCheckIndex(Preconditions.java:70)
	at java.base/jdk.internal.util.Preconditions.checkIndex(Preconditions.java:248)
	at java.base/java.util.Objects.checkIndex(Objects.java:372)
	at java.base/java.util.ArrayList.get(ArrayList.java:459)
	at io.github.stavshamir.springwolf.asyncapi.scanners.channels.RabbitChannelsScanner.getChannelName(RabbitChannelsScanner.java:45)
	at io.github.stavshamir.springwolf.asyncapi.scanners.channels.RabbitChannelsScanner.getChannelName(RabbitChannelsScanner.java:21)
	at io.github.stavshamir.springwolf.asyncapi.scanners.channels.AbstractChannelScanner.mapMethodToChannel(AbstractChannelScanner.java:81)

Hi!
Thank you for project!
I want to know, can i use it for documenting Spring Websocket functionality?

For example I use next configuration:

@Configuration
@EnableWebSocketMessageBroker
public class WebSocketConfig implements WebSocketMessageBrokerConfigurer {
	@Override
	public void configureMessageBroker(MessageBrokerRegistry config) {
		config.enableSimpleBroker("/topic", "/queue");
	}

	@Override
	public void registerStompEndpoints(StompEndpointRegistry registry) {
		registry
				.addEndpoint("/custom-endpoint")
				.setAllowedOrigins("*")
				.withSockJS();
	}
}

And next controller method:

@MessageMapping("/custom-path")
@SendToUser("/queue/custom-path")
public CustomModelResponse getStops(@Payload(required = false) CustomFilter filter) {

	// business logic invocation
        return ...
}

Can I describe it with your project to generate documentation and show it in Swagger like format (like Swagger for REST) and give it to client application? I guess, request, response models and url path will be show inside.
Thank you!

I like the idea of generating documentation in the build time.
Adding idea of an option to generate asyncapi json|yaml in the build time using gradle.
Generating the html UI I see only as an optional step, as there are usually already provided solutions that can visualise the provided spec in a fancy way.

1. Support for multiple messages per producers was added in #59 for Kafka. Such support should be also added for AMQP.

2. At least for AMQP multiple messages per consumer should be supported. This is available for Spring Boot with the @RabbitListener together with the @RabbitHandler annotation, see documentation.

When using springwolf-kafka 0.3.0 and springwolf-ui 0.3.1 and trying to declare multiple producers the portal does not show any channels.

return AsyncApiDocket.builder()
                .basePackage("foo.bar")
                .info(info)
                .server("kafka", Server.builder().protocol("kafka").url(BOOTSTRAP_SERVERS).build())
                .producers(List.of(
                        topicProducerData1,
                        topicProducerData2
                ))
                .build();

Having the ability to register custom model converters would be quite useful.

I ran into this issue because we are using MonetaryAmount in our POJOs and it breaks the schema generation. There was a similar issue in springdoc that was solved this way: springdoc/springdoc-openapi#606

As per our great discussion on discord, the best way to approach this is to add the ability to register new model converters in the core package and then to create a module that contains common model converters like the one for Monetary Amount to avoid bloating the core with unnecessary dependencies.

Thank you very much for the great work @stavshamir.

Receive the required output type as a request parameter (either yaml or json).

well, similar to: #55, maven is another popular build tool => adding it as well.

I like the idea of generating documentation in the build time.
Adding idea of an option to generate asyncapi json|yaml in the build time using maven.
Generating the html UI I see only as an optional step, as there are usually already provided solutions that can visualise the provided spec in a fancy way.

Goal

Allow users to specify producers, so that the producers will be added to the asyncapi document.

API

The producers will be set in the AsyncApiDocket, in a new field: producers: List<ProducerData>.
The ProducerData should have a builder and contain the following fields:

• channelName: String, not null
• payloadType: Class<?>, not null
• binding: Map<String, ? extends OperationBinding>, not null, and must contain exactly one item. The key must be the name of the protocol.

In order to make it easier for the users, protocol specific producer can be made where instead of providing the binding as a map, the user only need to provide the OperationBinding instance, and the key is hardcoded per protocol.
For example, KafkaProducerData, inherits ProducerData, and has a different setBinding method that accepts OperationBinding binding, and puts a "kafka": binding in the backing map.

ChannelScanner

A new implementation of ChannelScanner, ProducersChannelScanner should be implemented, taking the producers data from the docker and transform them to Channel.

Usage Example

@Bean
public AsyncApiDocket asyncApiDocket() {
    ProducerData kafkaProducerData = ProducerData.builder()
            .channelName("produce-topic")
            .payloadType(Foo.class)
            .binding(ImmutableMap.of("kafka", new KafkaOperationBinding())
            .build();

    return AsyncApiDocket.builder()
            .info(...)
            .server(...)
            .producers(Arrays.asList(kafkaProducerData))
            .build();
}

How can we add security to the specs that we've generated while viewing the UI something we have for swagger ui screen

Much thanks in advance..

how can we add example values for the payload to publish via asyncAPI ui?

There should be a support to add Kafka headers. For example the spring TypeId header should be provided in order for the spring to serialise and deserialize payloads. Without this header the publish button on the example tab cannot work as it need a TypeId header to do its job.

Thanks and keep up the good work.

• For users of the Spring ecosystem, the dependency of org.springframework.boot:spring-boot-starter-websocket is usually used to provide STOMP and sock.js protocol encapsulation for Websocket. STOMP is a sub-protocol of websocket, and the behavior and paradigm of STOMP are defined at https://www.asyncapi.com/docs/specifications/v2.4.0 . I hope there will be a corresponding treatment in SpringWolf.
• A simple demo is at https://github.com/spring-tips/controllers/tree/main/websockets , which helps to locate the actual model binding scenario.

Feature request to add yaml format to the /springwolf/docs endpoint. This can be useful for certain tools that work from YAML files to render AsyncAPI documentation.

Would it be possible to make the docs path configurable via properties?

Example application.properties

springwolf.path=/my/docs/path

Hi, while I running project, there occurs a exception :
org.springframework.beans.factory.BeanDefinitionStoreException: Failed to parse configuration class [com.caocao.dic.charge.DicChargeApplication]; nested exception is org.springframework.context.annotation.ConflictingBeanDefinitionException: Annotation-specified bean name 'kafkaProducer' for bean class [io.github.stavshamir.springwolf.producer.KafkaProducer] conflicts with existing, non-compatible bean definition of same name and class [com.caocao.dic.charge.server.mq.producer.KafkaProducer]
If project already exist class KafkaProducer, then will conflict with the class io.github.stavshamir.springwolf.producer.KafkaProducer. How should I solve the problem?

I know JMS is not so cool any more, but quite some java projects still depend on it.
As AsyncAPI supports JMS: https://www.asyncapi.com/docs/specifications/v2.3.0, it should be supported as well.

There's a handful of problems with supporting a case when there are multiple RabbitTemplate's.
Here's a use case, instead of using one rabbit template with a default exchange, and specifying a concrete exchange and routing key when sending messages, there could be multiple rabbit template with set exchanges, so that when sending a message you specify only the routing key.
The issues:

1. The app fails to start when there are multiple instances of RabbitTemplate's as the SpringwolfAmqpProducer requires a single instance (I can mark one of the RabbitTemplate's with Primary annotation, but that brings up the following issue)
2. The send method in the SpringwolfAmqpProducer uses the channelName to send the message, which is ok when we use a default exchange - we could just use the queue name here, but when we're not - we have to supply a routing key instead.

I think it needs to use a List of RabbitTemplate's in the SpringwolfAmqpProducer bean, and choose an appropriate one to send the message.
In order to send the message correctly, can the channel binding and operation binding info be used? https://github.com/asyncapi/bindings/tree/master/amqp#channel-binding-object (Specifically exchange.name/queue.name from channel binding and cc from operation binding)

Here's a starter demo if you want to test it out: https://github.com/DmitriButorchin/amqp-demo/tree/multiple-rabbit-templates

Describe the bug
When @ConditionalOnProperty is present in any listener class, it is discarded, even when Condition is true.
I've only tested with @ConditionalOnProperty, not sure if happening with other Conditional classes.
Only tested with @KafkaListener at method level.

Dependencies and versions used
springwolf-kafka version 0.7.0.

Code example

@Component
@ConditionalOnProperty("kafka.bootstrap-servers")
public class DenormalizationEventListener {
    @Autowired
    private DenormalizationService denormalizationService;

    @KafkaListener(topics = "test-topic",
                   containerFactory = "denormalizationEventListenerContainerFactory")
    public void transformUsuarioEstructuraToMongo(@Payload Evento event) {
        denormalizationService.denormalize(event);
    }
}

More details
As soon as @ConditionalOnProperty is removed, the channel is included in docs.
Property is being used in a lot of places, so I'm sure it's present.