是的，当然可以。请尝试以下脚本：

loadPlugin("/yourpath/PluginKafka.txt")

producerCfg = dict(STRING, ANY);
producerCfg["metadata.broker.list"] = "localhost:9092";
producer = kafka::producer(producerCfg);

def sendMsgToKafkaFunc(producer, msg){
    try {
        kafka::produce(producer, "dolphindb-producer-test", 1, msg, true)
    }
    catch(ex) {
        writeLog("[Kafka Plugin] Failed to send msg to kafka with error:" +ex)
    }
}

subscribeTable(tableName="kafkaTest", actionName="sendMsgToKafka", offset=0, handler=sendMsgToKafkaFunc{producer}, msgAsTable=true, reconnect=true)

Kafka所使用的信息：

...
{"id":[1,1,...],"datetime":["2020.01.01T10:55:12","2020.01.01T10:55:13",...],"val":[73,74,...]}
{"id":[1,1,...],"datetime":["2020.01.01T23:55:12","2020.01.01T23:55:13",...],"val":[88,1,...]}
...