我是使用Apache Flink的新手。我从Apache Kafka源读取数据并需要转换DataStream
。
在最后一步中,我尝试应用WindowFunction
:
DataStream<Tuple8<Double, Double, String, Double, Double, Double, Double, Double>> dataStream =
env
.addSource(new FlinkKafkaConsumer08<>(
parameterTool.getRequired("topic"),
new SimpleStringSchema(),
parameterTool.getProperties()))
.flatMap(new SplitIntoRecordsString())
.flatMap(new SplitIntoTuples())
.keyBy(1)
.countWindow(5)
.apply(new windowApplyFunction());
public class windowApplyFunction implements WindowFunction<
Tuple8<Double, Double, String, Double, Double, Double, Double, Double>,
String,
Double,
Window>{
public void apply(Double key, Window window,
Iterable<Tuple8<Double, Double, String, Double, Double, Double, Double, Double>> values,
Collector<String> out)
throws Exception {
out.collect("MyResult");
}
}
不幸的是我收到了以下错误,并且不知道如何解决它:
The method apply(WindowFunction<Tuple8<Double,Double,String,Double,Double,Double,Double,Double>,R,Tuple,GlobalWindow>) in the type WindowedStream<Tuple8<Double,Double,String,Double,Double,Double,Double,Double>,Tuple,GlobalWindow> is not applicable for the arguments (FlinkManager.windowApplyFunction)
如果我用预定义的函数替换apply(new windowApplyFunction())
,那么一切正常。 sum(1)
。
答案 0 :(得分:0)
您的WindowFunction
应该是
WindowFunction<
Tuple8<Double, Double, String, Double, Double, Double, Double, Double>,
String,
Double,
GlobalWindow>
countWindow()
会返回 GlobalWindow
类型。
试一试。
答案 1 :(得分:0)
感谢你的暗示vanekjar!在纠正了这个错误后,我改变了另一个小东西,它现在有效! 正确的代码:
public static class windowApplyFunction implements WindowFunction<
Tuple8<Double, Double, String, Double, Double, Double, Double, Double>,
Tuple8<Double, Double, String, Double, Double, Double, Double, Double>,
Tuple,
GlobalWindow>{
public void apply(Tuple key, GlobalWindow window,
Iterable<Tuple8<Double, Double, String, Double, Double, Double, Double, Double>> values,
Collector<Tuple8<Double, Double, String, Double, Double, Double, Double, Double>> out)
throws Exception {
out.collect(new Tuple8<Double, Double, String, Double, Double, Double, Double, Double>());
}
}