"Parquet is a columnar storage format for Hadoop" (for further details read http://parquet.io/)
the main purpouse of this project is to extract parquet parser from impala code. the 2nd purpouse is to port it to zerovm platfrom
the code is heavily under development and is not even compiling.