0xh3x / parquet-mr

Java implementation to use with Map-Reduce

Parquet MR

Parquet-mr is the java implementation of the Parquet format to be used in Hadoop. It uses the record shredding and assembly algorithm described in the Dremel paper. Integration with Pig and Map/Reduce are provided.

Apache Pig integration

A Loader and a Storer are provided to read and write Parquet files with Apache Pig

Map/Reduce integration

Thrift

Thrift mapping to the parquet schema is provided using a TBase extending class. You can read and write parquet files using Thrift generated classes.

Create your own objects

• The ParquetOutputFormat can be provided a WriteSupport to write your own objects to an event based RecordConsumer.
• the ParquetInputFormat can be provided a ReadSupport to materialize your own POJOs by implementing a RecordMaterializer

See the APIs:

• Record convertion API
• Hadoop API

Build

to run the unit tests: mvn test

to build the jars: mvn package

The build runs in Travis CI:

Authors and contributors

• Julien Le Dem http://twitter.com/J_
• Jonathan Coveney http://twitter.com/jco

Discussions

• google group https://groups.google.com/d/forum/parquet-dev
• the group email address: [e-mail]

License

Copyright 2012 Twitter, Inc.

Licensed under the Apache License, Version 2.0: http://www.apache.org/licenses/LICENSE-2.0