Example usage for org.apache.hadoop.mapred JobConf getMapperClass

Introduction

In this page you can find the example usage for org.apache.hadoop.mapred JobConf getMapperClass.

Prototype

public Class<? extends Mapper> getMapperClass() 

Source Link

Document

Get the Mapper class for the job.

Usage

From source file:org.archive.mapred.ARCMapRunner.java

License:LGPL

public void configure(JobConf job) {
    this.mapper = (ARCRecordMapper) ReflectionUtils.newInstance(job.getMapperClass(), job);
    // Value is in minutes.
    this.maxtime = job.getLong("wax.index.timeout", 60) * 60 * 1000;
}

---

From source file:org.gridgain.grid.kernal.processors.hadoop.v1.GridHadoopV1MapTask.java

License:Open Source License

/** {@inheritDoc} */
@SuppressWarnings("unchecked")
@Override/*from w  ww  . j  a v  a 2s.c  o m*/
public void run(GridHadoopTaskContext taskCtx) throws GridException {
    GridHadoopJob job = taskCtx.job();

    GridHadoopV2TaskContext ctx = (GridHadoopV2TaskContext) taskCtx;

    JobConf jobConf = ctx.jobConf();

    InputFormat inFormat = jobConf.getInputFormat();

    GridHadoopInputSplit split = info().inputSplit();

    InputSplit nativeSplit;

    if (split instanceof GridHadoopFileBlock) {
        GridHadoopFileBlock block = (GridHadoopFileBlock) split;

        nativeSplit = new FileSplit(new Path(block.file().toString()), block.start(), block.length(),
                EMPTY_HOSTS);
    } else
        nativeSplit = (InputSplit) ctx.getNativeSplit(split);

    assert nativeSplit != null;

    Reporter reporter = new GridHadoopV1Reporter(taskCtx);

    GridHadoopV1OutputCollector collector = null;

    try {
        collector = collector(jobConf, ctx, !job.info().hasCombiner() && !job.info().hasReducer(), fileName(),
                ctx.attemptId());

        RecordReader reader = inFormat.getRecordReader(nativeSplit, jobConf, reporter);

        Mapper mapper = ReflectionUtils.newInstance(jobConf.getMapperClass(), jobConf);

        Object key = reader.createKey();
        Object val = reader.createValue();

        assert mapper != null;

        try {
            try {
                while (reader.next(key, val)) {
                    if (isCancelled())
                        throw new GridHadoopTaskCancelledException("Map task cancelled.");

                    mapper.map(key, val, collector, reporter);
                }
            } finally {
                mapper.close();
            }
        } finally {
            collector.closeWriter();
        }

        collector.commit();
    } catch (Exception e) {
        if (collector != null)
            collector.abort();

        throw new GridException(e);
    }
}

---

• «
• 1
• 2
• »