CODEWITHSUNDEEP
hadoopapache-pigsequencefile
Aslan986
Aslan986

Reputation: 10324

Fail to write SequenceFile with Pig

I want to store some Pig variable to Hadoop SequenceFile, so as to run external MapReduce jobs.

Assume that my data has the (chararray, int) schema:

(hello,1)
(test,2)
(example,3)

I wrote this storing function:

import java.io.IOException;
import java.util.logging.Level;
import java.util.logging.Logger;
import org.apache.hadoop.fs.Path;
import org.apache.hadoop.io.IntWritable;
import org.apache.hadoop.io.Text;

import org.apache.hadoop.mapreduce.Job;
import org.apache.hadoop.mapreduce.OutputFormat;
import org.apache.hadoop.mapreduce.RecordWriter;
import org.apache.hadoop.mapreduce.lib.output.FileOutputFormat;
import org.apache.hadoop.mapreduce.lib.output.SequenceFileOutputFormat;
import org.apache.hadoop.mapreduce.lib.output.TextOutputFormat;
import org.apache.pig.StoreFunc;
import org.apache.pig.data.Tuple;


public class StoreTest extends StoreFunc {

    private String storeLocation;
    private RecordWriter writer;
    private Job job;

    public StoreTest(){

    }

    @Override
    public OutputFormat getOutputFormat() throws IOException {
        //return new TextOutputFormat();
        return new SequenceFileOutputFormat();
    }

    @Override
    public void setStoreLocation(String location, Job job) throws IOException {
        this.storeLocation = location;
        this.job = job;
        System.out.println("Load location is " + storeLocation);
        FileOutputFormat.setOutputPath(job, new Path(location));        
        System.out.println("Out path " + FileOutputFormat.getOutputPath(job));
    }

    @Override
    public void prepareToWrite(RecordWriter writer) throws IOException {
        this.writer = writer;
    }

    @Override
    public void putNext(Tuple tuple) throws IOException {
        try {
            Text k = new Text(((String)tuple.get(0)));
            IntWritable v = new IntWritable((Integer)tuple.get(1));
            writer.write(k, v);
        } catch (InterruptedException ex) {
            Logger.getLogger(StoreTest.class.getName()).log(Level.SEVERE, null, ex);
        }

    }
}

and this Pig code:

register MyUDFs.jar;
x = load '/user/pinoli/input' as (a:chararray,b:int);
store x into '/user/pinoli/output/' using StoreTest();

However, the storing fails and I get this error:

ERROR org.apache.pig.tools.pigstats.PigStats - ERROR 0: java.io.IOException: wrong key class: org.apache.hadoop.io.Text is not class org.apache.hadoop.io.LongWritable

Is there any way to fix that??

Upvotes: 0

Views: 760

Answers (1)

Lorand Bendig
Lorand Bendig

Reputation: 10650

The problem is that you didn't set the output key/value class. You can do this in the setStoreLocation() method:

@Override
public void setStoreLocation(String location, Job job) throws IOException {
    this.storeLocation = location;
    this.job = job;
    this.job.setOutputKeyClass(Text.class);   // !!!
    this.job.setOutputValueClass(IntWritable.class);  // !!!
    ...

}

I guess you want to use your storer with different key/value types. In that case you can pass their types to the constructor. E.g:

private Class<? extends WritableComparable> keyClass;
private Class<? extends Writable> valueClass;
...

public StoreTest() {...}

@SuppressWarnings({ "unchecked", "rawtypes" })
    public StoreTest(String keyClass, String valueClass) {
        try {
            this.keyClass = (Class<? extends WritableComparable>) Class.forName(keyClass);
            this.valueClass = (Class<? extends Writable>) Class.forName(valueClass);
        }
        catch (Exception e) {
            throw new RuntimeException("Invalid key/value type", e);
        }
    }

...

 @Override
    public void setStoreLocation(String location, Job job) throws IOException {
        this.storeLocation = location;
        this.job = job;
        this.job.setOutputKeyClass(keyClass);
        this.job.setOutputValueClass(valueClass);
        ...
    }

Then, set the correct types in the Pig script:

register MyUDFs.jar;
DEFINE myStorer StoreTest('org.apache.hadoop.io.Text', 'org.apache.hadoop.io.IntWritable');
x = load '/user/pinoli/input' as (a:chararray,b:int);
store x into '/user/pinoli/output/' using myStorer();

Upvotes: 1

Related Questions