如何使用属性值在 Scala 中声明 KafkaListener 的主题

Question

I have a simple Kafka/Scala project that creates a producer. Now I am trying to create the consumer, however, when I use the following code...

@Service
class KafkaService @Autowired()(producer: KafkaTemplate[String, Array[Byte]]){

  @Value("${spring.kafka.topic}") val topic : String = null

  def sendMessage(msg: String): Unit = {
    System.out.println(s"Writing the message $msg to the topic ${this.topic}")
    producer.send(topic, msg.getBytes());
  }

  @KafkaListener(id="test", topics="${this.topic}")
  def consume(record: ConsumerRecord[String, String]): Unit = {
    System.out.println(s"Consumed Strinsg Message : ${record.value()}")
  }

}

I get the following error...

[ERROR] ...\service\KafkaService.scala:26: error: type mismatch;
[ERROR]  found   : String("${this.topic}")
[ERROR]  required: Array[String]
[ERROR]   @KafkaListener(id="test", topics="${this.topic}")

What am I missing?

I also tried the following...

@Configuration
public class CommonConfiguration{
    ...
    @Value("${spring.kafka.topic}")
    public String topic;
    ...
}
@Service
class KafkaService @Autowired()(producer: KafkaTemplate[String, Array[Byte]], config: CommonConfiguration){

  def sendMessage(msg: String): Unit = {
    val topics : Array[String] = config.getTopics();
    println(s"Writing the message $msg ${topics.mkString(" ")}")
    producer.send(config.topic, msg.getBytes());
  }

  @KafkaListener(id="test", topics="#{config.topic.split(',')}")
  def consume(record: ConsumerRecord[String, String]): Unit = {
    System.out.println(s"Consumed Strinsg Message : ${record.value()}")
  }

}

Still no luck but the console log on the producer is getting the right value.

Answer 1

The error shows that you need a Array[String] type for topics whereas you have supplied a String type. You would need to convert the topics string into an array.

The annotations looks very similar to Spring where you can do something like

@Value("#{'${kafka.topic}'.split(',')}")

Answer 2

@KafkaListener(id="scala", topics=Array("#{'${spring.kafka.topic}'.split(',')}"))

有关更多信息，请参阅此问题或此问题