vchoi_fork

/* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *

 * Copyright by The HDF Group.                                               *

 * All rights reserved.                                                      *

 *                                                                           *

 * This file is part of HDF5.  The full HDF5 copyright notice, including     *

 * terms governing use, modification, and redistribution, is contained in    *

 * the COPYING file, which can be found at the root of the source code       *

 * distribution tree, or in https://support.hdfgroup.org/ftp/HDF5/releases.  *

 * If you do not have access to either file, you may request a copy from     *

 * help@hdfgroup.org.                                                        *

 * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * */

​

/* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *

 * Use Case 1.7 Appending a single chunk

 * Description:

 *     Appending a single chunk of raw data to a dataset along an unlimited

 *     dimension within a pre-created file and reading the new data back.

 * Goal:

 *     Read data appended by the Writer to a pre-existing dataset in a

 *     file. The dataset has one or more unlimited dimensions. The data is

 *     appended by a hyperslab that is contained in one chunk (for example,

 *     appending 2-dim planes along the slowest changing dimension in the

 *     3-dim dataset).

 * Level:

 *     User Level

 * Guarantees:

 *     o    Readers will see the modified dimension sizes after the Writer

 *  finishes HDF5 metadata updates and issues H5Fflush or H5Oflush calls.

 *     o    Readers will see newly appended data after the Writer finishes

 *  the flush operation.

 * 

 * Preconditions:

 *     o    Readers are not allowed to modify the file.  o   All datasets

 *  that are modified by the Writer exist when the Writer opens the file.

 *     o    All datasets that are modified by the Writer exist when a Reader

 *  opens the file.  o   Data is written by a hyperslab contained in

 *  one chunk.

 * 

 * Main Success Scenario:

 *     1.   An application creates a file with required objects (groups,

 *  datasets, and attributes).

 *     2.   The Writer application opens the file and datasets in the file

 *  and starts adding data along the unlimited dimension using a hyperslab

 *  selection that corresponds to an HDF5 chunk.

 *     3.   A Reader opens the file and a dataset in a file, and queries

 *  the sizes of the dataset; if the extent of the dataset has changed,

 *  reads the appended data back.

 * 

 * Discussion points:

 *     1.   Since the new data is written to the file, and metadata update

 *  operation of adding pointer to the newly written chunk is atomic and