Question

我需要在Mapper和Reducer之间传递一个Enum Bucket类型的列表，我已根据implementation-of-an-arraywritable-for-a-custom-hadoop-type实现了自定义BucketArrayWritable，而Bucket Enum则没有-argument构造函数，但我总是得到错误

java.lang.RuntimeException: java.lang.NoSuchMethodException: Bucket.<init>()
    at org.apache.hadoop.util.ReflectionUtils.newInstance(ReflectionUtils.java:131)
    at org.apache.hadoop.io.WritableFactories.newInstance(WritableFactories.java:58)
    at org.apache.hadoop.io.WritableFactories.newInstance(WritableFactories.java:64)
    at org.apache.hadoop.io.ArrayWritable.readFields(ArrayWritable.java:95)
    at org.apache.hadoop.io.serializer.WritableSerialization$WritableDeserializer.deserialize(WritableSerialization.java:71)
    at org.apache.hadoop.io.serializer.WritableSerialization$WritableDeserializer.deserialize(WritableSerialization.java:42)
    at org.apache.hadoop.mapreduce.task.ReduceContextImpl.nextKeyValue(ReduceContextImpl.java:139)
    at org.apache.hadoop.mapreduce.task.ReduceContextImpl.nextKey(ReduceContextImpl.java:114)
    at org.apache.hadoop.mapreduce.lib.reduce.WrappedReducer$Context.nextKey(WrappedReducer.java:296)
    at org.apache.hadoop.mapreduce.Reducer.run(Reducer.java:163)
    at org.apache.hadoop.mapred.ReduceTask.runNewReducer(ReduceTask.java:610)
    at org.apache.hadoop.mapred.ReduceTask.run(ReduceTask.java:444)
    at org.apache.hadoop.mapred.LocalJobRunner$Job.run(LocalJobRunner.java:449)
Caused by: java.lang.NoSuchMethodException: com.turn.platform.profile.mapreduce.counting.Bucket.<init>()
    at java.lang.Class.getConstructor0(Class.java:3082)
    at java.lang.Class.getDeclaredConstructor(Class.java:2178)
    at org.apache.hadoop.util.ReflectionUtils.newInstance(ReflectionUtils.java:125)

Reducer可能希望对init对象使用反射，但默认情况下Enum构造函数是私有的，

8.9.2 Enum Body Declarations
在枚举声明中，没有访问修饰符的构造函数声明是私有的。

我不知道是不是因为Hadoop ReflectionUtils无法找到私有构造函数

import java.io.DataInput;
import java.io.DataOutput;
import java.io.IOException;
import java.util.function.Predicate;

import org.apache.hadoop.io.LongWritable;
import org.apache.hadoop.io.Writable;


public enum Bucket implements Writable{

    // right bound exclusive
    MAX((count) -> {
        if (count > getValue()) {
            return true;
        }
        return false;
    }, (count, _userId, incre) -> {
        value = count;
        userId = _userId;
    }),
    TOTAL((count) -> {
        return true;
    }),
    BUCKET_0_10((count) -> {
        if (count < 10) {
            return true;
        }
        return false;
    }),
    BUCKET_10_100((count) -> {
        if (count >= 10 && count < 100) {
            return true;
        }
        return false;
    })
    ;

    private static long value = 0;
    private static LongWritable userId = new LongWritable(0);

    private TriConsumer<Long, LongWritable, Long> consumer;
    private Predicate<Long> predicator;

    Bucket() {

    }

    Bucket(Predicate<Long> predicator) {
        this.predicator = predicator;
    }

    Bucket(Predicate<Long> predicator, TriConsumer<Long, LongWritable, Long> consumer) {
        this.consumer = consumer;
        this.predicator = predicator;
    }

    public static long getValue() {
        return value;
    }

    public static void setValue(long newVal) {
        value = newVal;
    }

    public TriConsumer<Long, LongWritable, Long> getConsumer() {
        if (consumer == null) {
            consumer = (count, _userId, incre) -> {
                setValue(getValue() + incre);
                userId = _userId;
            };
        }
        return consumer;
    }

    public Predicate<Long> getPredicator() {
        return predicator;
    }

    @Override
    public void write(DataOutput out) throws IOException {
        out.writeLong(value);
        userId.write(out);
    }

    @Override
    public void readFields(DataInput in) throws IOException {
        value = in.readLong();
        userId.readFields(in);
    }
}

Answer 1

ReflectionUtils类尝试通过反射来初始化枚举，但枚举不能以这种方式实例化，这在Java语言规范中有说明，并且像this answer所说，主要是允许使用==来比较枚举。使用反射，您只能获得现有的参考。

关于enums的使用，你可以将enum'包装'在一个类中，这样你就可以覆盖reducers中通常使用的equals和hashcode方法。同样取决于您正在执行的操作WritableUtils还有方法writeEnum和readEnum。您可以查看有关如何在this answer中实现包装器类的更多信息。

Hadoop ReflectionUtils可以初始化枚举吗？

1 个答案: