/// <summary>Run the map task.</summary>
/// <param name="input">the set of inputs</param>
/// <param name="output">the object to collect the outputs of the map</param>
/// <param name="reporter">the object to update with status</param>
/// <exception cref="System.IO.IOException"/>
public override void Run(RecordReader <K1, V1> input, OutputCollector <K2, V2> output
, Reporter reporter)
{
Application <K1, V1, K2, V2> application = null;
try
{
RecordReader <FloatWritable, NullWritable> fakeInput = (!Submitter.GetIsJavaRecordReader
(job) && !Submitter.GetIsJavaMapper(job)) ? (RecordReader <FloatWritable, NullWritable
>)input : null;
application = new Application <K1, V1, K2, V2>(job, fakeInput, output, reporter, (
Type)job.GetOutputKeyClass(), (Type)job.GetOutputValueClass());
}
catch (Exception ie)
{
throw new RuntimeException("interrupted", ie);
}
DownwardProtocol <K1, V1> downlink = application.GetDownlink();
bool isJavaInput = Submitter.GetIsJavaRecordReader(job);
downlink.RunMap(reporter.GetInputSplit(), job.GetNumReduceTasks(), isJavaInput);
bool skipping = job.GetBoolean(MRJobConfig.SkipRecords, false);
try
{
if (isJavaInput)
{
// allocate key & value instances that are re-used for all entries
K1 key = input.CreateKey();
V1 value = input.CreateValue();
downlink.SetInputTypes(key.GetType().FullName, value.GetType().FullName);
while (input.Next(key, value))
{
// map pair to output
downlink.MapItem(key, value);
if (skipping)
{
//flush the streams on every record input if running in skip mode
//so that we don't buffer other records surrounding a bad record.
downlink.Flush();
}
}
downlink.EndOfInput();
}
application.WaitForFinish();
}
catch (Exception t)
{
application.Abort(t);
}
finally
{
application.Cleanup();
}
}
/// <summary>Wait for authentication response.</summary>
/// <exception cref="System.IO.IOException"/>
/// <exception cref="System.Exception"/>
internal virtual void WaitForAuthentication()
{
downlink.Flush();
Log.Debug("Waiting for authentication response");
handler.WaitForAuthentication();
}
