一、ReducingState的方法
- ReducingState是和ReduceFunction配合使用
- get() 获取状态的值
- add(IN value)方法添加一个元素,触发reduceFunction计算一次
二、ReducingState的描述器
ReducingState的描述器和之前ValueState、ListState不同,它得和一个ReduceFunction配合使用。
所以得先实现一个ReduceFunction,例如下面这个计算最高温度的。
public static class MyMaxTemp implements ReduceFunction<SensorRecord> {
@Override
public SensorRecord reduce(SensorRecord value1, SensorRecord value2) throws Exception {
return value1.getRecord() >= value2.getRecord() ? value1 : value2;
}
}
ReducingStateDescriptor第2个参数传new SensorRecordUtils.MyMaxTemp()。
以后每次来一个数据,都会执行这个MyMaxTemp的reduce()方法。
//用ReducingStateDescriptor定义描述器
ReducingStateDescriptor reducingStateDescriptor = new ReducingStateDescriptor(
"max-temp-state",//id
new SensorRecordUtils.MyMaxTemp(),//ReduceFunction
SensorRecord.class);//状态里的值的类型
//获取ReducingState
reducingState = getRuntimeContext().getReducingState(reducingStateDescriptor);
三、程序主体
public class Test05_ReduceState {
public static void main(String[] args) throws Exception {
StreamExecutionEnvironment env = StreamExecutionEnvironment.getExecutionEnvironment();
//方便测试,设置为1
env.setParallelism(1);
DataStreamSource<String> source = env.socketTextStream(BaseConstant.URL, BaseConstant.PORT);
/*
设置watermark和指定时间属性
*/
SingleOutputStreamOperator<SensorRecord> dataStream = source
.map(new SensorRecordUtils.BeanMap());
dataStream
.keyBy(SensorRecord::getId)
.process(new MyKeyedProcessFunction())
.print();
env.execute();
}
}
四、KeyedProcessFunction处理类
public static class MyKeyedProcessFunction extends KeyedProcessFunction<String, SensorRecord, SensorRecord> {
private transient ReducingState reducingState;
@Override
public void open(Configuration parameters) throws Exception {
super.open(parameters);
//用ReducingStateDescriptor定义描述器
ReducingStateDescriptor reducingStateDescriptor = new ReducingStateDescriptor(
"max-temp-state",//id
new SensorRecordUtils.MyMaxTemp(),//ReduceFunction
SensorRecord.class);//状态里的值的类型
//获取ReducingState
reducingState = getRuntimeContext().getReducingState(reducingStateDescriptor);
}
@Override
public void processElement(SensorRecord value, Context ctx, Collector<SensorRecord> out) throws Exception {
reducingState.add(value);
out.collect((SensorRecord) reducingState.get());
}
}
运行结果: