flink的带状态的RichFlatMapFunction函数使用

背景

使用RichFlatMapFunction可以带状态来决定如何对数据流进行转换，而且这种用法非常常见，根据之前遇到过的某个key的状态来决定再次遇到同样的key时要如何进行数据转换，本文就来简单举个例子说明下RichFlatMapFunction的使用方法

RichFlatMapFunction使用示例

下面的例子的输入是不用name下的count数量值，当本次name的数量和前一次name的数量相差超过配置的阈值100时，打印出来一条告警日志，详细代码如下：

package wikiedits.func.state;

import java.util.Objects;

import org.apache.flink.api.common.functions.RichFlatMapFunction;
import org.apache.flink.api.common.state.ValueState;
import org.apache.flink.api.common.state.ValueStateDescriptor;
import org.apache.flink.api.java.tuple.Tuple2;
import org.apache.flink.configuration.Configuration;
import org.apache.flink.util.Collector;

/**
 * Tuple2<String, Integer> 是输入的数据类型 String 是监控到异常值后的输出数据类型
 */
public class MyRichFlatMapFunction extends RichFlatMapFunction<Tuple2<String, Integer>, String> {

    // 键值分区状态,对应每个name一个值
    ValueState<StateEntity> nameState;

    @Override
    public void open(Configuration parameters) throws Exception {
        // 创建一个键值分区状态
        ValueStateDescriptor<StateEntity> state = new ValueStateDescriptor<>("nameState", StateEntity.class);
        nameState = getRuntimeContext().getState(state);
    }

    @Override
    public void flatMap(Tuple2<String, Integer> input, Collector<String> collector) throws Exception {
        // 判断状态值是否为空（状态默认值是空）
        if (Objects.isNull(nameState.value())) {
            StateEntity sFalg = new StateEntity(input.f0, input.f1);
            nameState.update(sFalg);
            return;
        }
        // 和上一次的状态值比较
        StateEntity value = nameState.value();
        if (Math.abs(value.count - input.f1) > 100) {
            collector.collect(new String("监控到异常值,名称: " + input.f0 + " 上次的值:" + value + " 本次的值：" + input));
        }
        value.setName(input.f0);
        value.setCount(input.f1);
        // 更新状态值
        nameState.update(value);
    }


}

package wikiedits.func.state;

import java.text.SimpleDateFormat;
import java.util.Date;

import org.apache.commons.lang3.RandomUtils;
import org.apache.flink.api.java.tuple.Tuple2;
import org.apache.flink.streaming.api.datastream.DataStream;
import org.apache.flink.streaming.api.environment.StreamExecutionEnvironment;
import org.apache.flink.streaming.api.functions.source.SourceFunction;

public class RichFlatMapFunctionTest {

    public static void main(String[] args) throws Exception {
        StreamExecutionEnvironment env = StreamExecutionEnvironment.getExecutionEnvironment();
        // 设置数据源，一共三个元素
        DataStream<Tuple2<String, Integer>> dataStream = env.addSource(new SourceFunction<Tuple2<String, Integer>>() {
            @Override
            public void run(SourceContext<Tuple2<String, Integer>> ctx) throws Exception {
                for (int i = 1; i < Integer.MAX_VALUE; i++) {
                    // 只有XXX，YYY，ZZZ三种name
                    String name = (0 == i % 3) ? "XXX" : ((i % 3 == 1) ? "YYY" : "ZZZ");
                    int count = RandomUtils.nextInt(0, 1000);
                    // 使用当前时间作为时间戳
                    long timeStamp = System.currentTimeMillis();
                    // 发射一个元素，并且戴上了时间戳
                    ctx.collectWithTimestamp(new Tuple2<String, Integer>(name, count), timeStamp);
                    // 每发射一次就延时1秒
                    Thread.sleep(5000);
                }
            }

            @Override
            public void cancel() {}
        });
        dataStream.keyBy((f) -> {
            return f.f0;
        }).flatMap(new MyRichFlatMapFunction()).print();

        env.execute();
    }

    public static String time(long timeStamp) {
        return new SimpleDateFormat("yyyy-MM-dd hh:mm:ss").format(new Date(timeStamp));
    }

}