Java 卡夫卡流分组和连接

我有一个接收记录的卡夫卡流，我想根据特定字段连接消息

流中的消息如下所示：

Key: 2099
Payload{
  email: tom@emample.com
  eventCode: 2099
}

预期产出：

key: 2099
Payload{
    emails: tom@example, bill@acme.com, jane@example.com
}

我可以让水流正常运行，我只是不确定lamda应该包含什么

这就是我到目前为止所做的。我不确定是否应该使用map、aggregate或reduce或这些操作的组合

final StreamsBuilder builder = new StreamsBuilder();
KStream<String, Payload> inputStream = builder.stream(INPUT_TOPIC);

inputStream
        .groupByKey()
        .windowedBy(TimeWindows.of(TimeUnit.MINUTES.toMillis(300000)))

                                  // Not sure what to do here …..

}).to (OUTPUT_TOPIC );

---

可能是这样的

inputStream.groupByKey().windowedBy(TimeWindows.of(TimeUnit.MINUTES.toMillis(300000)))
.aggregate(PayloadAggr::new, new Aggregator<String, Payload, PayloadAggr>() {
        @Override
        public PayloadAggr apply(String key, Payload newValue, PayloadAggr result) {
            result.setKey(key);
            if(result.getEmails()==null){
                result.setEmails(newValue.getEmail());
            }else{
                result.setEmails(result.getEmails() + "," + newValue.getEmail());
            }
            return result;
        }
    }, .../* You serdes and store */}).toStream().to(OUTPUT_TOPIC);

---

可能是这样的

inputStream.groupByKey().windowedBy(TimeWindows.of(TimeUnit.MINUTES.toMillis(300000)))
.aggregate(PayloadAggr::new, new Aggregator<String, Payload, PayloadAggr>() {
        @Override
        public PayloadAggr apply(String key, Payload newValue, PayloadAggr result) {
            result.setKey(key);
            if(result.getEmails()==null){
                result.setEmails(newValue.getEmail());
            }else{
                result.setEmails(result.getEmails() + "," + newValue.getEmail());
            }
            return result;
        }
    }, .../* You serdes and store */}).toStream().to(OUTPUT_TOPIC);