Java 如何在apache flink中连接两个流？

我正在开始与弗林克和有一个看

据我所知，此练习的目标是在时间属性上连接两个流

任务：

/*
 * Copyright 2017 data Artisans GmbH
 *
 * Licensed under the Apache License, Version 2.0 (the "License");
 * you may not use this file except in compliance with the License.
 * You may obtain a copy of the License at
 *
 *  http://www.apache.org/licenses/LICENSE-2.0
 *
 * Unless required by applicable law or agreed to in writing, software
 * distributed under the License is distributed on an "AS IS" BASIS,
 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 * See the License for the specific language governing permissions and
 * limitations under the License.
 */

package com.dataartisans.flinktraining.solutions.datastream_java.state;

import com.dataartisans.flinktraining.exercises.datastream_java.datatypes.TaxiFare;
import com.dataartisans.flinktraining.exercises.datastream_java.datatypes.TaxiRide;
import com.dataartisans.flinktraining.exercises.datastream_java.sources.TaxiFareSource;
import com.dataartisans.flinktraining.exercises.datastream_java.sources.TaxiRideSource;
import com.dataartisans.flinktraining.exercises.datastream_java.utils.ExerciseBase;
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.api.java.utils.ParameterTool;
import org.apache.flink.configuration.Configuration;
import org.apache.flink.streaming.api.TimeCharacteristic;
import org.apache.flink.streaming.api.datastream.DataStream;
import org.apache.flink.streaming.api.environment.StreamExecutionEnvironment;
import org.apache.flink.streaming.api.functions.co.RichCoFlatMapFunction;
import org.apache.flink.util.Collector;

/**
 * Java reference implementation for the "Stateful Enrichment" exercise of the Flink training
 * (http://training.data-artisans.com).
 *
 * The goal for this exercise is to enrich TaxiRides with fare information.
 *
 * Parameters:
 * -rides path-to-input-file
 * -fares path-to-input-file
 *
 */
public class RidesAndFaresSolution extends ExerciseBase {
    public static void main(String[] args) throws Exception {

        ParameterTool params = ParameterTool.fromArgs(args);
        final String ridesFile = params.get("rides", pathToRideData);
        final String faresFile = params.get("fares", pathToFareData);

        final int delay = 60;                   // at most 60 seconds of delay
        final int servingSpeedFactor = 1800;    // 30 minutes worth of events are served every second

        // set up streaming execution environment
        StreamExecutionEnvironment env = StreamExecutionEnvironment.getExecutionEnvironment();
        env.setStreamTimeCharacteristic(TimeCharacteristic.EventTime);
        env.setParallelism(ExerciseBase.parallelism);

        DataStream<TaxiRide> rides = env
                .addSource(rideSourceOrTest(new TaxiRideSource(ridesFile, delay, servingSpeedFactor)))
                .filter((TaxiRide ride) -> ride.isStart)
                .keyBy("rideId");

        DataStream<TaxiFare> fares = env
                .addSource(fareSourceOrTest(new TaxiFareSource(faresFile, delay, servingSpeedFactor)))
                .keyBy("rideId");

        DataStream<Tuple2<TaxiRide, TaxiFare>> enrichedRides = rides
                .connect(fares)
                .flatMap(new EnrichmentFunction());

        printOrTest(enrichedRides);

        env.execute("Join Rides with Fares (java RichCoFlatMap)");
    }

    public static class EnrichmentFunction extends RichCoFlatMapFunction<TaxiRide, TaxiFare, Tuple2<TaxiRide, TaxiFare>> {
        // keyed, managed state
        private ValueState<TaxiRide> rideState;
        private ValueState<TaxiFare> fareState;

        @Override
        public void open(Configuration config) {
            rideState = getRuntimeContext().getState(new ValueStateDescriptor<>("saved ride", TaxiRide.class));
            fareState = getRuntimeContext().getState(new ValueStateDescriptor<>("saved fare", TaxiFare.class));
        }

        @Override
        public void flatMap1(TaxiRide ride, Collector<Tuple2<TaxiRide, TaxiFare>> out) throws Exception {
            TaxiFare fare = fareState.value();
            if (fare != null) {
                fareState.clear();
                out.collect(new Tuple2(ride, fare));
            } else {
                rideState.update(ride);
            }
        }

        @Override
        public void flatMap2(TaxiFare fare, Collector<Tuple2<TaxiRide, TaxiFare>> out) throws Exception {
            TaxiRide ride = rideState.value();
            if (ride != null) {
                rideState.clear();
                out.collect(new Tuple2(ride, fare));
            } else {
                fareState.update(fare);
            }
        }
    }
}

这个练习的结果是一个Tuple2记录的数据流，每个记录对应一个不同的rideId。你应该忽略这个问题 结束活动，仅在每次骑行开始时加入活动 其相应的票价数据

结果流应打印到标准输出

问题：EnrichmentFunction如何连接两个流。它如何知道参加哪个游乐项目？我希望它能缓冲多个展会/游乐设施，直到一个即将到来的展会/游乐设施有一个匹配的合作伙伴

据我所知，它只是保存了它看到的每一个骑乘/游乐项目，并将其与下一个最佳骑乘/游乐项目相结合。为什么这是一个合适的连接

提供的解决方案：

/*
 * Copyright 2017 data Artisans GmbH
 *
 * Licensed under the Apache License, Version 2.0 (the "License");
 * you may not use this file except in compliance with the License.
 * You may obtain a copy of the License at
 *
 *  http://www.apache.org/licenses/LICENSE-2.0
 *
 * Unless required by applicable law or agreed to in writing, software
 * distributed under the License is distributed on an "AS IS" BASIS,
 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 * See the License for the specific language governing permissions and
 * limitations under the License.
 */

package com.dataartisans.flinktraining.solutions.datastream_java.state;

import com.dataartisans.flinktraining.exercises.datastream_java.datatypes.TaxiFare;
import com.dataartisans.flinktraining.exercises.datastream_java.datatypes.TaxiRide;
import com.dataartisans.flinktraining.exercises.datastream_java.sources.TaxiFareSource;
import com.dataartisans.flinktraining.exercises.datastream_java.sources.TaxiRideSource;
import com.dataartisans.flinktraining.exercises.datastream_java.utils.ExerciseBase;
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.api.java.utils.ParameterTool;
import org.apache.flink.configuration.Configuration;
import org.apache.flink.streaming.api.TimeCharacteristic;
import org.apache.flink.streaming.api.datastream.DataStream;
import org.apache.flink.streaming.api.environment.StreamExecutionEnvironment;
import org.apache.flink.streaming.api.functions.co.RichCoFlatMapFunction;
import org.apache.flink.util.Collector;

/**
 * Java reference implementation for the "Stateful Enrichment" exercise of the Flink training
 * (http://training.data-artisans.com).
 *
 * The goal for this exercise is to enrich TaxiRides with fare information.
 *
 * Parameters:
 * -rides path-to-input-file
 * -fares path-to-input-file
 *
 */
public class RidesAndFaresSolution extends ExerciseBase {
    public static void main(String[] args) throws Exception {

        ParameterTool params = ParameterTool.fromArgs(args);
        final String ridesFile = params.get("rides", pathToRideData);
        final String faresFile = params.get("fares", pathToFareData);

        final int delay = 60;                   // at most 60 seconds of delay
        final int servingSpeedFactor = 1800;    // 30 minutes worth of events are served every second

        // set up streaming execution environment
        StreamExecutionEnvironment env = StreamExecutionEnvironment.getExecutionEnvironment();
        env.setStreamTimeCharacteristic(TimeCharacteristic.EventTime);
        env.setParallelism(ExerciseBase.parallelism);

        DataStream<TaxiRide> rides = env
                .addSource(rideSourceOrTest(new TaxiRideSource(ridesFile, delay, servingSpeedFactor)))
                .filter((TaxiRide ride) -> ride.isStart)
                .keyBy("rideId");

        DataStream<TaxiFare> fares = env
                .addSource(fareSourceOrTest(new TaxiFareSource(faresFile, delay, servingSpeedFactor)))
                .keyBy("rideId");

        DataStream<Tuple2<TaxiRide, TaxiFare>> enrichedRides = rides
                .connect(fares)
                .flatMap(new EnrichmentFunction());

        printOrTest(enrichedRides);

        env.execute("Join Rides with Fares (java RichCoFlatMap)");
    }

    public static class EnrichmentFunction extends RichCoFlatMapFunction<TaxiRide, TaxiFare, Tuple2<TaxiRide, TaxiFare>> {
        // keyed, managed state
        private ValueState<TaxiRide> rideState;
        private ValueState<TaxiFare> fareState;

        @Override
        public void open(Configuration config) {
            rideState = getRuntimeContext().getState(new ValueStateDescriptor<>("saved ride", TaxiRide.class));
            fareState = getRuntimeContext().getState(new ValueStateDescriptor<>("saved fare", TaxiFare.class));
        }

        @Override
        public void flatMap1(TaxiRide ride, Collector<Tuple2<TaxiRide, TaxiFare>> out) throws Exception {
            TaxiFare fare = fareState.value();
            if (fare != null) {
                fareState.clear();
                out.collect(new Tuple2(ride, fare));
            } else {
                rideState.update(ride);
            }
        }

        @Override
        public void flatMap2(TaxiFare fare, Collector<Tuple2<TaxiRide, TaxiFare>> out) throws Exception {
            TaxiRide ride = rideState.value();
            if (ride != null) {
                rideState.clear();
                out.collect(new Tuple2(ride, fare));
            } else {
                fareState.update(fare);
            }
        }
    }
}

/*
*版权所有2017 data Artisans GmbH
*
*根据Apache许可证2.0版（以下简称“许可证”）获得许可；
*除非遵守许可证，否则不得使用此文件。
*您可以通过以下方式获得许可证副本：
*
*  http://www.apache.org/licenses/LICENSE-2.0
*
*除非适用法律要求或书面同意，软件
*根据许可证进行的分发是按“原样”进行分发的，
*无任何明示或暗示的保证或条件。
*请参阅许可证以了解管理权限和权限的特定语言
*许可证下的限制。
*/
包com.dataartisans.flinktraining.solutions.datastream_java.state；
导入com.dataartisans.flinktraining.exerces.datastream_java.datatypes.TaxiFare；
导入com.dataartisans.flinktraining.exerces.datastream_java.datatypes.TaxiRide；
导入com.dataartisans.flinktraining.exerces.datastream_java.sources.TaxiFareSource；
导入com.dataartisans.flinktraining.exerces.datastream_java.sources.TaxiRideSource；
导入com.dataartisans.flinktraining.exerces.datastream_java.utils.ExerciseBase；
导入org.apache.flink.api.common.state.ValueState；
导入org.apache.flink.api.common.state.ValueStateDescriptor；
导入org.apache.flink.api.java.tuple.Tuple2；
导入org.apache.flink.api.java.utils.ParameterTool；
导入org.apache.flink.configuration.configuration；
导入org.apache.flink.streaming.api.TimeCharacteristic；
导入org.apache.flink.streaming.api.datastream.datastream；
导入org.apache.flink.streaming.api.environment.StreamExecutionEnvironment；
导入org.apache.flink.streaming.api.functions.co.RichCoFlatMapFunction；
导入org.apache.flink.util.Collector；
/**
*Flink培训“状态丰富”练习的Java参考实现
* (http://training.data-artisans.com).
*
*这项工作的目标是通过车费信息丰富出租车。
*
*参数：
*-占用输入文件的路径
*-输入文件的路径
*
*/
公共类乘车和票价解决方案扩展了练习库{
公共静态void main（字符串[]args）引发异常{
ParameterTool params=ParameterTool.fromArgs（args）；
最终字符串ridesFile=params.get（“rides”，pathToRideData）；
最终字符串faresFile=params.get（“fares”，pathToFareData）；
final int delay=60；//最多延迟60秒
final int servingSpeedFactor=1800；//每秒提供30分钟的活动
//设置流执行环境
StreamExecutionEnvironment env=StreamExecutionEnvironment.getExecutionEnvironment（）；
环境setStreamTimeCharacteristic（TimeCharacteristic.EventTime）；
环境setParallelism（ExerciseBase.parallelism）；
DataStream rides=env
.addSource（rideSourceOrTest（新的滑行资源（ridesFile、delay、servicingspeedfactor）））
.filter（（滑行行驶）->ride.isStart）
.keyBy（“rideId”）；
数据流票价=环境
.addSource（fareSourceOrTest（新出租车票价源（票价文件、延迟、服务速度系数）））
.keyBy（“rideId”）；
数据流enrichedRides=骑乘
.连接（票价）
.flatMap（新的EnrichmentFunction（））；
printOrTest（enrichedRides）；
env.execute（“以票价加入骑乘（java RichCoFlatMap）”）；
}
公共静态类EnrichmentFunction扩展了RichCoFlatMapFunction{
//键控、管理状态
私人价值观；
私人价值国家票价国家；
@凌驾
公共无效打开（配置）{
rideState=getRuntimeContext（）.getState（新的ValueStateDescriptor（“保存的乘坐”，TaxiRide.class））；
fareState=getRuntimeContext（）.getState（新的ValueStateDescriptor（“保存的票价”，TaxiFare.class））；
}
@凌驾
public void flatMap1（出租车行驶，收集器输出）引发异常{
出租车票价=fareState.value（）；
如果（票价！=null）{
fareState.clear（）；
外出。领取（新的Tuple2（乘车、车费））；
}否则{
骑乘状态更新（骑乘）；
}
}
@凌驾
public void Flatmap 2（出租车车费、收费站外）抛出异常{
滑行行驶=滑行状态。值（）；
如果（行驶！=null）{
清除（）；
外出。领取（新的Tuple2（乘车、车费））；
}否则{
票价状态。更新（票价）；
}
}
}
}

在这个特定的上下文中，rideId的每个值都有三个事件——滑行开始事件、滑行结束事件和滑行票价。本练习的目的是将每个滑行启动事件与具有相同rideId的一个滑行票价事件连接起来，或者换句话说，在知道每个事件只有一个的情况下，将rideId上的乘坐流和票价流连接起来

本练习演示如何设置关键帧状态