Kafka Streams DSL 到 Processor API 有什么区别 [英] What is the difference between Kafka Streams DSL to Processor API

问题描述

从我从这里读到的:https://docs.confluent.io/3.0.0/streams/developer-guide.html#streams-developer-guide-processor-api

与我的理解不同的是，我可以在处理器 API 中更具体地说明哪些记录序列化程序进入流(源)到退出流(接收器)的记录，并且在流 DSL 中它必须相同.

The difference to my understanding is that i can be more specific in the Processor API as to what record serializer enters the stream (source) to what exits the stream (sink) and in the stream DSL it must be the same.

是否可以使用 KStreamBuilder 和 TopologyBuilder 创建一个 Kstream 来生成完全相同的流?如果可以，我可以看一个例子吗?

is it possible to create a Kstream with KStreamBuilder and with TopologyBuilder that will result in the exact same stream? if so can i see an example?

推荐答案

...并且在流 DSL 中它必须相同.

... and in the stream DSL it must be the same.

我不认为这是正确的.您可以从输入流中获取消息，更改值甚至键，然后将它们放在不同的输出流中.

I don't think that is right. You can take messages from an input stream, change the value or even the key, and put them on a different output stream.

final KStream<String, String> inputStream = builder.stream("inputStream");
...
inputStream
    .filter(this::acceptCertainMessages)
    .transform(new MyTransformerSupplier<String, String>("store"), "store")
    .to("outputStream");

这篇关于Kafka Streams DSL 到 Processor API 有什么区别的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！