Flink java 自定义reduce函数，以wordcount为例

maven项目的文本文件与pom.xml配置请参考：https://blog.csdn.net/weixin_35757704/article/details/120555968
同样以wordcount为例

package transform;

import org.apache.flink.api.common.functions.FlatMapFunction;
import org.apache.flink.api.common.functions.ReduceFunction;
import org.apache.flink.api.java.tuple.Tuple2;
import org.apache.flink.streaming.api.datastream.DataStream;
import org.apache.flink.streaming.api.datastream.KeyedStream;
import org.apache.flink.streaming.api.datastream.SingleOutputStreamOperator;
import org.apache.flink.streaming.api.environment.StreamExecutionEnvironment;
import org.apache.flink.util.Collector;

public class WordCountKeyBy {
    public static void main(String[] args) throws Exception {
        // 1.创建流式执行环境
        StreamExecutionEnvironment env = StreamExecutionEnvironment.getExecutionEnvironment();

        // 2.从文件中读取数据
        DataStream dataStream = env.readTextFile("src/main/resources/hello.txt");
        // 执行环境并行度设置3
        env.setParallelism(3);

        // 3.按照空格分词
        DataStream> sensorStream = dataStream.flatMap(new FlatMapFunction>() {
            @Override
            public void flatMap(String value, Collector> out) throws Exception {
                String[] wordString = value.split(" ");
                for (String wordLine : wordString) {
                    out.collect(new Tuple2<>(wordLine, 1));
                }
            }
        });
        // 4.分组
        KeyedStream, Object> key = sensorStream.keyBy(tuple -> tuple.f0);
        // 5.聚合
        SingleOutputStreamOperator> resultStream = key.reduce(new ReduceFunction>() {
            @Override
            public Tuple2 reduce(Tuple2 value1, Tuple2 value2) throws Exception {
                return new Tuple2(value1.f0, value1.f0.length() + value1.f1);
            }
        });
        resultStream.print();
        //执行
        env.execute();
    }
}

在上面的第5步为自定义的聚合操作，其中：reduce(Tuple2 value1, Tuple2 value2)中value1为旧有的状态，value2为新输入的状态；

上面的代码中Tuple第一个位置返回原有的单词，而Tuple第二个位置每次都加一次当前单词的长度