Implement BytesWritable in DataWritableWriter Class in parquet

Implement BytesWritable in org.apache.hadoop.hive.ql.io.parquet.write.DataWritableWriter Class in parquet .
writePremitive() api of this class has major Writable but not BytesWritable (Note - it has ByteWritable which is just 1 byte) . So we have to use the binaryWritable class . However binaryWritable does not have the set function which means for each record we have to call new (no of times string data type occurs ).

**Reporter**: [Suresh Bahuguna](https://issues.apache.org/jira/secure/ViewProfile.jspa?name=subahugu)

_{**Note**: *This issue was originally created as [PARQUET-795](https://issues.apache.org/jira/browse/PARQUET-795). Please see the [migration documentation](https://issues.apache.org/jira/browse/PARQUET-2502) for further details.*}

Commencez dans org.apache.hadoop.hive.ql.io.parquet.write.DataWritableWriter et examinez writePremitive() ainsi que sa gestion existante de Writable. Vérifiez comment binaryWritable est utilisé pour les données de type string, puis vérifiez que les valeurs BytesWritable sont gérées sans nécessiter un nouvel objet pour chaque enregistrement ; l’issue est terminée lorsque cette prise en charge est couverte par les tests concernés.

Rédigé par le modèle d'indexation à partir du texte de l'issue.