<html>
<head>
<title>
HDF5 File Format Specification
</title>
</head>
<body bgcolor="#FFFFFF">
<hr>
<center>
<table border=0 width=98%>
<tr><td valign=top align=left>
<a href="index.html">Other HDF5 documents and links</a> <br>
<a href="H5.intro.html">Introduction to HDF5</a> <br>
</td>
<td> </td>
<td valign=top align=right>
<a href="H5.user.html">HDF5 User Guide</a> <br>
<a href="RM_H5Front.html">HDF5 Reference Manual</a> <br>
</td></tr>
</table>
</center>
<hr>
<center><h1>HDF5 File Format Specification</h1></center>
<center>
<table border=0 width=90%>
<tr>
<td valign=top>
<ol type=I>
<li><a href="#Intro">Introduction</a>
<li><a href="#BootBlock">Disk Format Level 0 - File Signature and Super Block</a>
<li><a href="#Group">Disk Format Level 1 - File Infrastructure</a>
<font size=-2>
<ol type=A>
<li><a href="#Btrees">Disk Format Level 1A - B-link Trees and B-tree Nodes</a>
<li><a href="#SymbolTable">Disk Format Level 1B - Group</a>
<li><a href="#SymbolTableEntry">Disk Format Level 1C - Group Entry</a>
<li><a href="#LocalHeap">Disk Format Level 1D - Local Heaps</a>
<li><a href="#GlobalHeap">Disk Format Level 1E - Global Heap</a>
<li><a href="#FreeSpaceIndex">Disk Format Level 1F - Free-space Index</a>
</ol>
</font>
<li><a href="#DataObject">Disk Format Level 2 - Data Objects</a>
<font size=-2>
<ol type=A>
<li><a href="#ObjectHeader">Disk Format Level 2a - Data Object Headers</a>
<ol type=1>
<li><a href="#NILMessage">Name: NIL</a> <!-- 0x0000 -->
<li><a href="#SimpleDataSpace">Name: Simple Dataspace</a> <!-- 0x0001 -->
<!--
<li><a href="#DataSpaceMessage">Name: Complex Dataspace</a> --> <!-- 0x0002 -->
<li><a href="#DataTypeMessage">Name: Datatype</a> <!-- 0x0003 -->
<li><a href="#FillValueMessage">Name: Data Storage - Fill Value</a> <!-- 0x0004 -->
<li><a href="#ReservedMessage_0005">Name: Reserved - not assigned yet</a> <!-- 0x0005 -->
</ol>
</ol>
</font>
</ol>
</td><td> </td><td valign=top>
<ol type=I>
<li><a href="#DataObject">Disk Format Level 2 - Data Objects</a>
<font size=-2><i>(Continued)</i>
<ol type=A>
<li><a href="#ObjectHeader">Disk Format Level 2a - Data Object Headers</a><i>(Continued)</i>
<ol type=1>
<li><a href="#CompactDataStorageMessage">Name: Data Storage - Compact</a> <!-- 0x0006 -->
<li><a href="#ExternalFileListMessage">Name: Data Storage - External Data Files</a> <!-- 0x0007 -->
<li><a href="#LayoutMessage">Name: Data Storage - Layout</a> <!-- 0x0008 -->
<li><a href="#ReservedMessage_0009">Name: Reserved - not assigned yet</a> <!-- 0x0009 -->
<li><a href="#ReservedMessage_000A">Name: Reserved - not assigned yet</a> <!-- 0x000a -->
<li><a href="#FilterMessage">Name: Data Storage - Filter Pipeline</a> <!-- 0x000b -->
<li><a href="#AttributeMessage">Name: Attribute</a> <!-- 0x000c -->
<li><a href="#NameMessage">Name: Object Name</a> <!-- 0x000d -->
<li><a href="#ModifiedMessage">Name: Object Modification Date and Time</a> <!-- 0x000e -->
<li><a href="#SharedMessage">Name: Shared Object Message</a> <!-- 0x000f -->
<li><a href="#ContinuationMessage">Name: Object Header Continuation</a> <!-- 0x0010 -->
<li><a href="#SymbolTableMessage">Name: Group Message</a> <!-- 0x0011 -->
</ol>
<li><a href="#SharedObjectHeader">Disk Format: Level 2b - Shared Data Object Headers</a>
<li><a href="#DataStorage">Disk Format: Level 2c - Data Object Data Storage</a>
</ol>
</font>
</ol>
</td></tr>
</table>
</center>
<br><br>
<h2>Introduction</h2>
<table align=right width=100>
<tr><td> </td><td align=center>
<hr>
<img src="FF-IH_FileGroup.gif" alt="HDF5 Groups" hspace=15 vspace=15>
</td><td> </td></tr>
<tr><td> </td><td align=center>
<strong>Figure 1:</strong> Relationships among the HDF5 root group, other groups, and objects
<hr>
</td><td> </td></tr>
<tr><td> </td><td align=center>
<img src="FF-IH_FileObject.gif" alt="HDF5 Objects" hspace=15 vspace=15>
</td><td> </td></tr>
<tr><td> </td><td align=center>
<strong>Figure 2:</strong> HDF5 objects -- datasets, datatypes, or dataspaces
<hr>
</td><td> </td></tr>
</table>
<P>The format of an HDF5 file on disk encompasses several
key ideas of the HDF4 and AIO file formats as well as
addressing some shortcomings therein. The new format is
more self-describing than the HDF4 format and is more
uniformly applied to data objects in the file.
<P>An HDF5 file appears to the user as a directed graph.
The nodes of this graph are the higher-level HDF5 objects
that are exposed by the HDF5 APIs:
<ul>
<li>Groups
<li>Datasets
<li>Datatypes
<li>Dataspaces
</ul>
<P>At the lowest level, as information is actually written to the disk,
an HDF5 file is made up of the following objects:
<ul>
<li>A super block
<li>B-tree nodes (containing either symbol nodes or raw data chunks)
<li>Object headers
<li>Collections
<li>Local heaps
<li>Free space
</ul>
The HDF5 library uses these lower-level objects to represent the
higher-level objects that are then presented to the user or
to applications through the APIs.
For instance, a group is an object header that contains a message that
points to a local heap and to a B-tree which points to symbol nodes.
A dataset is an object header that contains messages that describe
datatype, space, layout, filters, external files, fill value, etc
with the layout message pointing to either a raw data chunk or to a
B-tree that points to raw data chunks.
<h3>This Document</h3>
<p>This document describes the lower-level data objects;
the higher-level objects and their properties are described
in the <a href="H5.user.html"><cite>HDF5 User's Guide</cite></a>.
<!--
<blockquote>
<pre>
Elena> NOTE: give reference to the detailed discussion of the B-trees
Elena> when needed. Right now we do not have specification (only general one)
Elena> for the Symbol Table B-trees and B-trees used to manage chunked datasets.
Elena> B-trees
Elena> General Discussion
Elena> Object related discussions
Elena> Symbol Tables
Elena> Global heap
Elena> "Free-space object"
</pre>
</blockquote>
-->
<P>Three levels of information comprise the file format.
Level 0 contains basic information for identifying and
defining information about the file. Level 1 information contains
the group information (stored as a B-tree) and is used as the
index for all the objects in the file. Level 2 is the rest
of the file and contains all of the data objects, with each object
partitioned into header information, also known as
<em>meta information</em>, and data.
<p>The sizes of various fields in the following layout tables are
determined by looking at the number of columns the field spans
in the table. There are three exceptions: (1) The size may be
overridden by specifying a size in parentheses, (2) the size of
addresses is determined by the <em>Size of Offsets</em> field
in the super block, and (3) the size of size fields is determined
by the <em>Size of Lengths</em> field in the super block.
<br><br>
<br><br>
<h2><a name="BootBlock">
Disk Format: Level 0 - File Signature and Super Block</a></h2>
<P>The super block may begin at certain predefined offsets within
the HDF5 file, allowing a block of unspecified content for
users to place additional information at the beginning (and
end) of the HDF5 file without limiting the HDF5 library's
ability to manage the objects within the file itself. This
feature was designed to accommodate wrapping an HDF5 file in
another file format or adding descriptive information to the
file without requiring the modification of the actual file's
information. The super block is located by searching for the
HDF5 file signature at byte offset 0, byte offset 512 and at
successive locations in the file, each a multiple of two of
the previous location, i.e. 0, 512, 1024, 2048, etc.
<P>The super block is composed of a file signature, followed by
super block and group version numbers, information
about the sizes of offset and length values used to describe
items within the file, the size of each group page,
and a group entry for the root object in the file.
<p>
<center>
<table border align=center cellpadding=4 width="80%">
<caption align=top>
<B>HDF5 Super Block Layout</B>
</caption>
<tr align=center>
<th width="25%">byte</th>
<th width="25%">byte</th>
<th width="25%">byte</th>
<th width="25%">byte</th>
</tr>
<tr align=center>
<td colspan=4><br>HDF5 File Signature (8 bytes)<br><br></td>
</tr>
<tr align=center>
<td>Version # of Super Block</td>
<td>Version # of Global Free-space Storage</td>
<td>Version # of Group</td>
<td>Reserved</td>
</tr>
<tr align=center>
<td>Version # of Shared Header Message Format</td>
<td>Size of Offsets</td>
<td>Size of Lengths</td>
<td>Reserved (zero)</td>
</tr>
<tr align=center>
<td colspan=2>Group Leaf Node K</td>
<td colspan=2>Group Internal Node K</td>
</tr>
<tr align=center>
<td colspan=4>File Consistency Flags</td>
</tr>
<tr align=center>
<td colspan=4>Base Address</td>
</tr>
<tr align=center>
<td colspan=4>Address of Global Free-space Heap</td>
</tr>
<tr align=center>
<td colspan=4>End of File Address</td>
</tr>
<tr align=center>
<td colspan=4>Driver Information Block Address</td>
</tr>
<tr align=center>
<td colspan=4><br>Root Group Address<br><br></td>
</tr>
</table>
</center>
<p>
<center>
<table width="80%">
<tr>
<th width="30%">Field Name</th>
<th width="70%">Description</th>
</tr>
<tr valign=top>
<td>File Signature</td>
<td>This field contains a constant value and can be used to
quickly identify a file as being an HDF5 file. The
constant value is designed to allow easy identification of
an HDF5 file and to allow certain types of data corruption
to be detected. The file signature of an HDF5 file always
contains the following values:
<br><br><center>
<table border align=center cellpadding=4 width="100%">
<tr align=center>
<td>decimal</td>
<td width="8%">137</td>
<td width="8%">72</td>
<td width="8%">68</td>
<td width="8%">70</td>
<td width="8%">13</td>
<td width="8%">10</td>
<td width="8%">26</td>
<td width="8%">10</td>
</tr>
<tr align=center>
<td>hexadecimal</td>
<td width="8%">89</td>
<td width="8%">48</td>
<td width="8%">44</td>
<td width="8%">46</td>
<td width="8%">0d</td>
<td width="8%">0a</td>
<td width="8%">1a</td>
<td width="8%">0a</td>
</tr>
<tr align=center>
<td>ASCII C Notation</td>
<td width="8%">\211</td>
<td width="8%">H</td>
<td width="8%">D</td>
<td width="8%">F</td>
<td width="8%">\r</td>
<td width="8%">\n</td>
<td width="8%">\032</td>
<td width="8%">\n</td>
</tr>
</table>
</center>
<br>
This signature both identifies the file as an HDF5 file
and provides for immediate detection of common
file-transfer problems. The first two bytes distinguish
HDF5 files on systems that expect the first two bytes to
identify the file type uniquely. The first byte is
chosen as a non-ASCII value to reduce the probability
that a text file may be misrecognized as an HDF5 file;
also, it catches bad file transfers that clear bit
7. Bytes two through four name the format. The CR-LF
sequence catches bad file transfers that alter newline
sequences. The control-Z character stops file display
under MS-DOS. The final line feed checks for the inverse
of the CR-LF translation problem. (This is a direct
descendent of the PNG file signature.)</td>
</tr>
<tr valign=top>
<td>Version Number of the Super Block</td>
<td>This value is used to determine the format of the
information in the super block. When the format of the
information in the super block is changed, the version number
is incremented to the next integer and can be used to
determine how the information in the super block is
formatted.</td>
</tr>
<tr valign=top>
<td>Version Number of the Global Free-space Heap</td>
<td>This value is used to determine the format of the
information in the Global Free-space Heap.</td>
</tr>
<tr valign=top>
<td>Version Number of the Group</td>
<td>This value is used to determine the format of the
information in the Group. When the format of
the information in the Group is changed, the
version number is incremented to the next integer and can be
used to determine how the information in the Group
is formatted.</td>
</tr>
<tr valign=top>
<td>Version Number of the Shared Header Message Format</td>
<td>This value is used to determine the format of the
information in a shared object header message, which is
stored in the global small-data heap. Since the format
of the shared header messages differs from the private
header messages, a version number is used to identify changes
in the format.</td>
</tr>
<tr valign=top>
<td>Size of Offsets</td>
<td>This value contains the number of bytes used to store
addresses in the file. The values for the addresses of
objects in the file are offsets relative to a base address,
usually the address of the super block signature. This
allows a wrapper to be added after the file is created
without invalidating the internal offset locations.</td>
</tr>
<tr valign=top>
<td>Size of Lengths</td>
<td>This value contains the number of bytes used to store
the size of an object.</td>
</tr>
<tr valign=top>
<td>Group Leaf Node K</td>
<td>Each leaf node of a group B-tree will have at
least this many entries but not more than twice this
many. If a group has a single leaf node then it
may have fewer entries.</td>
</tr>
<tr valign=top>
<td>Group Internal Node K</td>
<td>Each internal node of a group B-tree will have
at least K pointers to other nodes but not more than 2K
pointers. If the group has only one internal
node then it might have fewer than K pointers.</td>
</tr>
<tr valign=top>
<td>Bytes per B-tree Page</td>
<td>This value contains the number of bytes used for symbol
pairs per page of the B-trees used in the file. All
B-tree pages will have the same size per page.
<br>
For 32-bit file offsets, 340 objects is the maximum
per 4KB page; for 64-bit file offset, 254 objects will fit
per 4KB page. In general, the equation is:
<br>
<code> <<i>number of objects</i>> =
<br>
FLOOR((<<i>page size</i>> - <<i>offset size</i>>) /
<br>
(<<i>Symbol size</i>> + <<i>offset size</i>>))
- 1 </code></td>
</tr>
<tr valign=top>
<td>File Consistency Flags</td>
<td>This value contains flags to indicate information
about the consistency of the information contained
within the file. Currently, the following bit flags are
defined:
<ul>
<li>Bit 0 set indicates that the file is opened for
write-access.
<li>Bit 1 set indicates that the file has
been verified for consistency and is guaranteed to be
consistent with the format defined in this document.
<li>Bits 2-31 are reserved for future use.
</ul>
Bit 0 should be
set as the first action when a file is opened for write
access and should be cleared only as the final action
when closing a file. Bit 1 should be cleared during
normal access to a file and only set after the file's
consistency is guaranteed by the library or a
consistency utility.</td>
</tr>
<tr valign=top>
<td>Base Address</td>
<td>This is the absolute file address of the first byte of
the HDF5 data within the file. The library currently
constrains this value to be the absolute file address
of the super block itself when creating new files;
future versions of the library may provide greater
flexibility. Unless otherwise noted,
all other file addresses are relative to this base
address.</td>
</tr>
<tr valign=top>
<td>Address of Global Free-space Heap</td>
<td>Free-space management is not yet defined in the HDF5
file format and is not handled by the library.
Currently this field always contains the
undefined address <code>0xfff...ff</code>.
<!--
<td>This value contains the relative address of the B-tree
used to manage the blocks of data which are unused in the
file currently. The free-space heap is used to manage the
blocks of bytes at the file-level which become unused when
objects are moved within the file.</td>
-->
</tr>
<tr valign=top>
<td>End of File Address</td>
<td>This is the relative file address of the first byte past
the end of all HDF5 data. It is used to determine whether a
file has been accidently truncated and as an address where
file data allocation can occur if the free list is not
used.</td>
</tr>
<tr valign=top>
<td>Driver Information Block Address</td>
<td>This is the relative file address of the file driver
information block which contains driver-specific
information needed to reopen the file. If there is no
driver information block then this entry should be the
undefined address (all bits set).</td>
</tr>
<tr valign=top>
<td>Root Group Address</td>
<td>This is the address of the root group (described later
in this document), which serves as the entry point into
the group graph.</td>
</tr>
</table>
</center>
<p>The <em>file driver information block</em> is an optional region of the
file which contains information needed by the file driver in
order to reopen a file. The format of the file driver information
block is:
<p>
<center>
<table border align=center cellpadding=4 width="80%">
<caption align=top>
<B>Driver Information Block</B>
</caption>
<tr align=center>
<th width="25%">byte</th>
<th width="25%">byte</th>
<th width="25%">byte</th>
<th width="25%">byte</th>
</tr>
<tr align=center>
<td>Version</td>
<td colspan=3>Reserved (zero)</td>
</tr>
<tr align=center>
<td colspan=4>Driver Information Size (4 bytes)</td>
</tr>
<tr align=center>
<td colspan=4><br>Driver Identification (8 bytes)<br><br></td>
</tr>
<tr align=center>
<td colspan=4><br><br>Driver Information<br><br><br></td>
</tr>
</table>
</center>
<p>
<center>
<table align=center width="80%">
<tr>
<th width="30%">Field Name</th>
<th width="70%">Description</th>
</tr>
<tr valign=top>
<td>Version</td>
<td>The version number of the driver information block. The
file format documented here is version zero.</td>
</tr>
<tr valign=top>
<td>Driver Information Size</td>
<td>The size in bytes of the Driver Information part of this
structure.</td>
</tr>
<tr valign=top>
<td>Driver Identification</td>
<td>This is an eight-byte ASCII string without null
termination which identifies the driver and version number
of the Driver Information block. The predefined drivers
supplied with the HDF5 library are identified by the
letters <code>NCSA</code> followed by the first four characters of
the driver name. If the Driver Information block is not
the original version then the last letter(s) of the
identification will be replaced by a version number in
ASCII.
For example, the various versions of the <em>family driver</em>
will be identified by <code>NCSAfami</code>, <code>NCSAfam0</code>,
<code>NCSAfam1</code>, etc.
(<code>NCSAfami</code> is simply <code>NCSAfamily</code> truncated
to eight characters. Subsequent identifiers will be created by
substituting sequential numerical values for the final character,
starting with zero.)
<p>
Identification for user-defined drivers
is arbitrary but should be unique.</td>
</tr>
<tr valign=top>
<td>Driver Information</td>
<td>Driver information is stored in a format defined by the
file driver and encoded/decoded by the driver callbacks
invoked from the <code>H5FD_sb_encode</code> and
<code>H5FD_sb_decode</code> functions.</td>
</tr>
</table>
</center>
<br><br>
<br><br>
<h2><a name="Group">
Disk Format: Level 1 - File Infrastructure</a></h2>
<h3><a name="Btrees">Disk Format: Level 1A - B-link Trees and B-tree Nodes</a></h3>
<p>B-link trees allow flexible storage for objects which tend to grow
in ways that cause the object to be stored discontiguously. B-trees
are described in various algorithms books including "Introduction to
Algorithms" by Thomas H. Cormen, Charles E. Leiserson, and Ronald
L. Rivest. The B-link tree, in which the sibling nodes at a
particular level in the tree are stored in a doubly-linked list,
is described in the "Efficient Locking for Concurrent Operations
on B-trees" paper by Phillip Lehman and S. Bing Yao as published
in the <em>ACM Transactions on Database Systems</em>, Vol. 6,
No. 4, December 1981.
<p>The B-link trees implemented by the file format contain one more
key than the number of children. In other words, each child
pointer out of a B-tree node has a left key and a right key.
The pointers out of internal nodes point to sub-trees while
the pointers out of leaf nodes point to symbol nodes and
raw data chunks.
Aside from that difference, internal nodes and leaf nodes
are identical.
<p>
<center>
<table border cellpadding=4 width="80%">
<caption align=top>
<B>B-tree Nodes</B>
</caption>
<tr align=center>
<th width="25%">byte</th>
<th width="25%">byte</th>
<th width="25%">byte</th>
<th width="25%">byte</th>
<tr align=center>
<td colspan=4>Node Signature</td>
<tr align=center>
<td>Node Type</td>
<td>Node Level</td>
<td colspan=2>Entries Used</td>
<tr align=center>
<td colspan=4>Address of Left Sibling</td>
<tr align=center>
<td colspan=4>Address of Right Sibling</td>
<tr align=center>
<td colspan=4>Key 0 (variable size)</td>
<tr align=center>
<td colspan=4>Address of Child 0</td>
<tr align=center>
<td colspan=4>Key 1 (variable size)</td>
<tr align=center>
<td colspan=4>Address of Child 1</td>
<tr align=center>
<td colspan=4>...</td>
<tr align=center>
<td colspan=4>Key 2<em>K</em> (variable size)</td>
<tr align=center>
<td colspan=4>Address of Child 2<em>K</em></td>
<tr align=center>
<td colspan=4>Key 2<em>K</em>+1 (variable size)</td>
</table>
</center>
<p>
<center>
<table align=center width="80%">
<tr>
<th width="30%">Field Name</th>
<th width="70%">Description</th>
</tr>
<tr valign=top>
<td>Node Signature</td>
<td>The ASCII character string <code>TREE</code> is
used to indicate the
beginning of a B-link tree node. This gives file
consistency checking utilities a better chance of
reconstructing a damaged file.</td>
</tr>
<tr valign=top>
<td>Node Type</td>
<td>Each B-link tree points to a particular type of data.
This field indicates the type of data as well as
implying the maximum degree <em>K</em> of the tree and
the size of each Key field.
<br>
<dl compact>
<dt>0
<dd>This tree points to group nodes.
<dt>1
<dd>This tree points to a new data chunk.
</dl>
</td>
</tr>
<tr valign=top>
<td>Node Level</td>
<td>The node level indicates the level at which this node
appears in the tree (leaf nodes are at level zero). Not
only does the level indicate whether child pointers
point to sub-trees or to data, but it can also be used
to help file consistency checking utilities reconstruct
damanged trees.</td>
</tr>
<tr valign=top>
<td>Entries Used</td>
<td>This determines the number of children to which this
node points. All nodes of a particular type of tree
have the same maximum degree, but most nodes will point
to less than that number of children. The valid child
pointers and keys appear at the beginning of the node
and the unused pointers and keys appear at the end of
the node. The unused pointers and keys have undefined
values.</td>
</tr>
<tr valign=top>
<td>Address of Left Sibling</td>
<td>This is the file address of the left sibling of the
current node relative to the super block. If the current
node is the left-most node at this level then this field
is the undefined address (all bits set).</td>
</tr>
<tr valign=top>
<td>Address of Right Sibling</td>
<td>This is the file address of the right sibling of the
current node relative to the super block. If the current
node is the right-most node at this level then this
field is the undefined address (all bits set).</td>
</tr>
<tr valign=top>
<td>Keys and Child Pointers</td>
<td>Each tree has 2<em>K</em>+1 keys with 2<em>K</em>
child pointers interleaved between the keys. The number
of keys and child pointers actually containing valid
values is determined by the <em>Entries Used</em> field. If
that field is <em>N</em> then the B-link tree contains
<em>N</em> child pointers and <em>N</em>+1 keys.</td>
</tr>
<tr valign=top>
<td>Key</td>
<td>The format and size of the key values is determined by
the type of data to which this tree points. The keys are
ordered and are boundaries for the contents of the child
pointer; that is, the key values represented by child
<em>N</em> fall between Key <em>N</em> and Key
<em>N</em>+1. Whether the interval is open or closed on
each end is determined by the type of data to which the
tree points.
<p>
The format of the key depends on the node type.
For nodes of node type 1, the key is formatted as follows:
<center>
<table>
<tr valign=top align=left>
<td width=40%>Bytes 1-4</td>
<td>Size of chunk in bytes.</td>
<tr valign=top align=left></tr>
<td>Bytes 4-8</td>
<td>Filter mask, a 32-bit bitfield indicating which
filters have been applied to that chunk.</td>
</tr><tr valign=top align=left>
<td><i>N</i> fields of 8 bytes each</td>
<td>A 64-bit index indicating the offset of the
chunk within the dataset where <i>N</i> is the number
of dimensions of the dataset. For example, if
a chunk in a 3-dimensional dataset begins at the
position <code>[5,5,5]</code>, there will be three
such 8-bit indices, each with the value of
<code>5</code>.</td>
</tr>
</table>
</center>
<p>
For nodes of node type 0, the key is formatted as follows:
<center>
<table>
<tr valign=top align=left>
<td width=40%>A single field of <i>Size of Lengths</i>
bytes</td>
<td>Indicates the byte offset into the local heap
for the first object name in the subtree which
that key describes.</td>
</tr>
</table>
</center>
</td>
</tr>
<tr valign=top>
<td>Child Pointers</td>
<td>The tree node contains file addresses of subtrees or
data depending on the node level. Nodes at Level 0 point
to data addresses, either data chunk or group nodes.
Nodes at non-zero levels point to other nodes of the
same B-tree.</td>
</tr>
</table>
</center>
<p>
Each B-tree node looks like this:
<center>
<table>
<tr valign=top align=center>
<td>key[0]</td><td> </td>
<td>child[0]</td><td> </td>
<td>key[1]</td><td> </td>
<td>child[1]</td><td> </td>
<td>key[2]</td><td> </td>
<td>...</td><td> </td>
<td>...</td><td> </td>
<td>key[<i>N</i>-1]</td><td> </td>
<td>child[<i>N</i>-1]</td><td> </td>
<td>key[<i>N</i>]</td>
</tr>
</table>
</center>
where child[<i>i</i>] is a pointer to a sub-tree (at a level
above Level 0) or to data (at Level 0).
Each key[<i>i</i>] describes an <i>item</i> stored by the B-tree
(a chunk or an object of a group node). The range of values
represented by child[<i>i</i>] are indicated by key[<i>i</i>]
and key[<i>i</i>+1].
<p>The following question must next be answered:
"Is the value described by key[<i>i</i>] contained in
child[<i>i</i>-1] or in child[<i>i</i>]?"
The answer depends on the type of tree.
In trees for groups (node type 0) the object described by
key[<i>i</i>] is the greatest object contained in
child[<i>i</i>-1] while in chunk trees (node type 1) the
chunk described by key[<i>i</i>] is the least chunk in
child[<i>i</i>].
<p>That means that key[0] for group trees is sometimes unused;
it points to offset zero in the heap, which is always the
empty string and compares as "less-than" any valid object name.
<p>And key[<i>N</i>] for chunk trees is sometimes unused;
it contains a chunk offset which compares as "greater-than"
any other chunk offset and has a chunk byte size of zero
to indicate that it is not actually allocated.
<h3><a name="SymbolTable">Disk Format: Level 1B - Group and Symbol Nodes</a></h3>
<p>A group is an object internal to the file that allows
arbitrary nesting of objects (including other groups).
A group maps a set of names to a set of file
address relative to the base address. Certain meta data
for an object to which the group points can be duplicated
in the group symbol table in addition to the object header.
<p>An HDF5 object name space can be stored hierarchically by
partitioning the name into components and storing each
component in a group. The group entry for a
non-ultimate component points to the group containing
the next component. The group entry for the last
component points to the object being named.
<p>A group is a collection of group nodes pointed
to by a B-link tree. Each group node contains entries
for one or more symbols. If an attempt is made to add a
symbol to an already full group node containing
2<em>K</em> entries, then the node is split and one node
contains <em>K</em> symbols and the other contains
<em>K</em>+1 symbols.
<p>
<center>
<table border cellpadding=4 width="80%">
<caption align=top>
<B>Group Node (A Leaf of a B-tree)</B>
</caption>
<tr align=center>
<th width="25%">byte</th>
<th width="25%">byte</th>
<th width="25%">byte</th>
<th width="25%">byte</th>
<tr align=center>
<td colspan=4>Node Signature</td>
<tr align=center>
<td>Version Number</td>
<td>Reserved for Future Use</td>
<td colspan=2>Number of Symbols</td>
<tr align=center>
<td colspan=4><br><br>Group Entries<br><br><br></td>
</table>
</center>
<p>
<center>
<table align=center width="80%">
<tr>
<th width="30%">Field Name</th>
<th width="70%">Description</th>
</tr>
<tr valign=top>
<td>Node Signature</td>
<td>The ASCII character string <code>SNOD</code> is
used to indicate the
beginning of a group node. This gives file
consistency checking utilities a better chance of
reconstructing a damaged file.</td>
</tr>
<tr valign=top>
<td>Version Number</td>
<td>The version number for the group node. This
document describes version 1.</td>
</tr>
<tr valign=top>
<td>Number of Symbols</td>
<td>Although all group nodes have the same length,
most contain fewer than the maximum possible number of
symbol entries. This field indicates how many entries
contain valid data. The valid entries are packed at the
beginning of the group node while the remaining
entries contain undefined values.</td>
</tr>
<tr valign=top>
<td>Group Entries</td>
<td>Each symbol has an entry in the group node.
The format of the entry is described below.</td>
</tr>
</table>
</center>
<h3><a name="SymbolTableEntry">
Disk Format: Level 1C - Group Entry </a></h3>
<p>Each group entry in a group node is designed
to allow for very fast browsing of stored objects.
Toward that design goal, the group entries
include space for caching certain constant meta data from the
object header.
<p>
<center>
<table border cellpadding=4 width="80%">
<caption align=top>
<B>Group Entry</B>
</caption>
<tr align=center>
<th width="25%">byte</th>
<th width="25%">byte</th>
<th width="25%">byte</th>
<th width="25%">byte</th>
</tr>
<tr align=center>
<td colspan=4>Name Offset (<size> bytes)</td>
</tr>
<tr align=center>
<td colspan=4>Object Header Address</td>
</tr>
<tr align=center>
<td colspan=4>Cache Type</td>
</tr>
<tr align=center>
<td colspan=4>Reserved</td>
</tr>
<tr align=center>
<td colspan=4><br><br>Scratch-pad Space (16 bytes)<br><br><br></td>
</tr>
</table>
</center>
<p>
<center>
<table align=center width="80%">
<tr>
<th width="30%">Field Name</th>
<th width="70%">Description</th>
</tr>
<tr valign=top>
<td>Name Offset</td>
<td>This is the byte offset into the group local
heap for the name of the object. The name is null
terminated.</td>
</tr>
<tr valign=top>
<td>Object Header Address</td>
<td>Every object has an object header which serves as a
permanent location for the object's meta data. In addition
to appearing in the object header, some meta data can be
cached in the scratch-pad space.</td>
</tr>
<tr valign=top>
<td>Cache Type</td>
<td>The cache type is determined from the object header.
It also determines the format for the scratch-pad space.
<br>
<dl compact>
<dt>0
<dd>No data is cached by the group entry. This
is guaranteed to be the case when an object header
has a link count greater than one.
<dt>1
<dd>Object header meta data is cached in the group
entry. This implies that the group
entry refers to another group.
<dt>2
<dd>The entry is a symbolic link. The first four bytes
of the scratch-pad space are the offset into the local
heap for the link value. The object header address
will be undefined.
<dt><em>N</em>
<dd>Other cache values can be defined later and
libraries that do not understand the new values will
still work properly.
</dl>
</td>
</tr>
<tr valign=top>
<td>Reserved</td>
<td>These four bytes are present so that the scratch-pad
space is aligned on an eight-byte boundary. They are
always set to zero.</td>
</tr>
<tr valign=top>
<td>Scratch-pad Space</td>
<td>This space is used for different purposes, depending
on the value of the Cache Type field. Any meta-data
about a dataset object represented in the scratch-pad
space is duplicated in the object header for that
dataset. This meta data can include the datatype
and the size of the dataspace for a dataset whose datatype
is atomic and whose dataspace is fixed and less than
four dimensions.
Furthermore, no data is cached in the group
entry scratch-pad space if the object header for
the group entry has a link count greater than
one.</td>
</tr>
</table>
</center>
<h4>Format of the Scratch-pad Space</h4>
<p>The group entry scratch-pad space is formatted
according to the value in the Cache Type field.
<p>If the Cache Type field contains the value zero
(<code>0</code>) then no information is
stored in the scratch-pad space.
<p>If the Cache Type field contains the value one
(<code>1</code>), then the scratch-pad space
contains cached meta data for another object header
in the following format:
<p>
<center>
<table border cellpadding=4 width="80%">
<caption align=top>
<B>Object Header Scratch-pad Format</B>
</caption>
<tr align=center>
<th width="25%">byte</th>
<th width="25%">byte</th>
<th width="25%">byte</th>
<th width="25%">byte</th>
<tr align=center>
<td colspan=4>Address of B-tree</td>
<tr align=center>
<td colspan=4>Address of Name Heap</td>
</table>
</center>
<p>
<center>
<table align=center width="80%">
<tr>
<th width="30%">Field Name</th>
<th width="70%">Description</th>
</tr>
<tr valign=top>
<td>Address of B-tree</td>
<td>This is the file address for the root of the
group's B-tree.</td>
</tr>
<tr valign=top>
<td>Address of Name Heap</td>
<td>This is the file address for the group's local
heap, in which are stored the symbol names.</td>
</tr>
</table>
</center>
<p>If the Cache Type field contains the value two
(<code>2</code>), then the scratch-pad space
contains cached meta data for another symbolic link
in the following format:
<p>
<center>
<table border cellpadding=4 width="80%">
<caption align=top>
<B>Symbolic Link Scratch-pad Format</B>
</caption>
<tr align=center>
<th width="25%">byte</th>
<th width="25%">byte</th>
<th width="25%">byte</th>
<th width="25%">byte</th>
</tr>
<tr align=center>
<td colspan=4>Offset to Link Value</td>
</tr>
</table>
</center>
<p>
<center>
<table align=center width="80%">
<tr>
<th width="30%">Field Name</th>
<th width="70%">Description</th>
</tr>
<tr valign=top>
<td>Offset to Link Value</td>
<td>The value of a symbolic link (that is, the name of the
thing to which it points) is stored in the local heap.
This field is the 4-byte offset into the local heap for
the start of the link value, which is null terminated.</td>
</tr>
</table>
</center>
<h3><a name="LocalHeap">Disk Format: Level 1D - Local Heaps</a></h3>
<p>A heap is a collection of small heap objects. Objects can be
inserted and removed from the heap at any time.
The address of a heap does not change once the heap is created.
References to objects are stored in the group table;
the names of those objects are stored in the local heap.
<p>
<center>
<table border cellpadding=4 width="80%">
<caption align=top>
<b>Local Heaps</b>
</caption>
<tr align=center>
<th width="25%">byte</th>
<th width="25%">byte</th>
<th width="25%">byte</th>
<th width="25%">byte</th>
</tr>
<tr align=center>
<td colspan=4>Heap Signature</td>
</tr>
<tr align=center>
<td colspan=4>Reserved (zero)</td>
</tr>
<tr align=center>
<td colspan=4>Data Segment Size</td>
</tr>
<tr align=center>
<td colspan=4>Offset to Head of Free-list (<size> bytes)</td>
</tr>
<tr align=center>
<td colspan=4>Address of Data Segment</td>
</tr>
</table>
</center>
<p>
<center>
<table align=center width="80%">
<tr>
<th width="30%">Field Name</th>
<th width="70%">Description</th>
</tr>
<tr valign=top>
<td>Heap Signature</td>
<td>The ASCII character string <code>HEAP</code>
is used to indicate the
beginning of a heap. This gives file consistency
checking utilities a better chance of reconstructing a
damaged file.</td>
</tr>
<tr valign=top>
<td>Data Segment Size</td>
<td>The total amount of disk memory allocated for the heap
data. This may be larger than the amount of space
required by the object stored in the heap. The extra
unused space holds a linked list of free blocks.</td>
</tr>
<tr valign=top>
<td>Offset to Head of Free-list</td>
<td>This is the offset within the heap data segment of the
first free block (or all 0xff bytes if there is no free
block). The free block contains <size> bytes that
are the offset of the next free chunk (or all 0xff bytes
if this is the last free chunk) followed by <size>
bytes that store the size of this free chunk.</td>
</tr>
<tr valign=top>
<td>Address of Data Segment</td>
<td>The data segment originally starts immediately after
the heap header, but if the data segment must grow as a
result of adding more objects, then the data segment may
be relocated, in its entirety, to another part of the
file.</td>
</tr>
</table>
</center>
<p>Objects within the heap should be aligned on an 8-byte boundary.
<h3><a name="GlobalHeap">Disk Format: Level 1E - Global Heap</a></h3>
<p>Each HDF5 file has a global heap which stores various types of
information which is typically shared between datasets. The
global heap was designed to satisfy these goals:
<ol type="A">
<li>Repeated access to a heap object must be efficient without
resulting in repeated file I/O requests. Since global heap
objects will typically be shared among several datasets, it is
probable that the object will be accessed repeatedly.
<br><br>
<li>Collections of related global heap objects should result in
fewer and larger I/O requests. For instance, a dataset of
void pointers will have a global heap object for each
pointer. Reading the entire set of void pointer objects
should result in a few large I/O requests instead of one small
I/O request for each object.
<br><br>
<li>It should be possible to remove objects from the global heap
and the resulting file hole should be eligible to be reclaimed
for other uses.
<br><br>
</ol>
<p>The implementation of the heap makes use of the memory
management already available at the file level and combines that
with a new top-level object called a <em>collection</em> to
achieve Goal B. The global heap is the set of all collections.
Each global heap object belongs to exactly one collection and
each collection contains one or more global heap objects. For
the purposes of disk I/O and caching, a collection is treated as
an atomic object.
<p>
<center>
<table border cellpadding=4 width="80%">
<caption align=top>
<B>A Global Heap Collection</B>
</caption>
<tr align=center>
<th width="25%">byte</th>
<th width="25%">byte</th>
<th width="25%">byte</th>
<th width="25%">byte</th>
</tr>
<tr align=center>
<td colspan=4>Magic Number</td>
</tr>
<tr align=center>
<td>Version</td>
<td colspan=3>Reserved</td>
</td>
<tr align=center>
<td colspan=4>Collection Size</td>
</tr>
<tr align=center>
<td colspan=4><br>Global Heap Object 1
<i>(described below)</i><br><br></td>
</tr>
<tr align=center>
<td colspan=4><br>Global Heap Object 2<br><br></td>
</tr>
<tr align=center>
<td colspan=4><br>...<br><br></td>
</tr>
<tr align=center>
<td colspan=4><br>Global Heap Object <em>N</em><br><br></td>
</tr>
<tr align=center>
<td colspan=4><br>Global Heap Object 0 (free space)<br><br></td>
</tr>
</table>
</center>
<p>
<center>
<table align=center width="80%">
<tr>
<th width="30%">Field Name</th>
<th width="70%">Description</th>
</tr>
<tr valign=top>
<td>Magic Number</td>
<td>The magic number for global heap collections are the
four bytes <code>G</code>, <code>C</code>, <code>O</code>,
and <code>L</code>.</td>
</tr>
<tr valign=top>
<td>Version</td>
<td>Each collection has its own version number so that new
collections can be added to old files. This document
describes version zero of the collections.
</tr>
<tr valign=top>
<td>Collection Data Size</td>
<td>This is the size in bytes of the entire collection
including this field. The default (and minimum)
collection size is 4096 bytes which is a typical file
system block size and which allows for 170 16-byte heap
objects plus their overhead.</td>
</tr>
<tr valign=top>
<td>Object 1 through <em>N</em></td>
<td>The objects are stored in any order with no
intervening unused space.</td>
</tr>
<tr valign=top>
<td>Object 0</td>
<td>Object 0 (zero), when present, represents the free space in
the collection. Free space always appears at the end of
the collection. If the free space is too small to store
the header for Object 0 (described below) then the
header is implied and the collection contains no free space.
</table>
</center>
<p>
<center>
<table border cellpadding=4 width="80%">
<caption align=top>
<B>Global Heap Object</B>
</caption>
<tr align=center>
<th width="25%">byte</th>
<th width="25%">byte</th>
<th width="25%">byte</th>
<th width="25%">byte</th>
</tr>
<tr align=center>
<td colspan=2>Object ID</td>
<td colspan=2>Reference Count</td>
</tr>
<tr align=center>
<td colspan=4>Reserved</td>
</tr>
<tr align=center>
<td colspan=4>Object Data Size</td>
</tr>
<tr align=center>
<td colspan=4><br>Object Data<br><br></td>
</tr>
</table>
</center>
<p>
<center>
<table align=center width="80%">
<tr>
<th width="30%">Field Name</th>
<th width="70%">Description</th>
</tr>
<tr valign=top>
<td>Object ID</td>
<td>Each object has a unique identification number within a
collection. The identification numbers are chosen so that
new objects have the smallest value possible with the
exception that the identifier <code>0</code> always refers to the
object which represents all free space within the
collection.</td>
</tr>
<tr valign=top>
<td>Reference Count</td>
<td>All heap objects have a reference count field. An
object which is referenced from some other part of the
file will have a positive reference count. The reference
count for Object 0 is always zero.</td>
</tr>
<tr valign=top>
<td>Reserved</td>
<td>Zero padding to align next field on an 8-byte
boundary.</td>
</tr>
<tr valign=top>
<td>Object Size</td> <td>This is the size of the the fields
above plus the object data stored for the object. The
actual storage size is rounded up to a multiple of
eight.</td>
</tr>
<tr valign=top>
<td>Object Data</td>
<td>The object data is treated as a one-dimensional array
of bytes to be interpreted by the caller.</td>
</tr>
</table>
</center>
<h3><a name="FreeSpaceIndex">Disk Format: Level 1F - Free-space Heap</a></h3>
<p>The Free-space Index is a collection of blocks of data,
dispersed throughout the file, which are currently not used by
any file objects.
<p>The super block contains a pointer to root of the free-space description;
that pointer is currently (i.e., in HDF5 Release 1.2) required
to be the undefined address <code>0xfff...ff</code>.
<p>The free-sapce index is not otherwise publicly defined at this time.
<!--
<p>The Free-space Index is a collection of blocks of data,
dispersed throughout the file, which are currently not used by
any file objects. The blocks of data are indexed by a B-tree of
their length within the file.
<p>Each B-tree page is composed of the following entries and
B-tree management information, organized as follows:
<p>
<center>
<table border cellpadding=4 width="80%">
<caption align=bottom>
<B>HDF5 Free-space Heap Page</B>
</caption>
<tr align=center>
<th width="25%">byte</th>
<th width="25%">byte</th>
<th width="25%">byte</th>
<th width="25%">byte</th>
<tr align=center>
<td colspan=4>Free-space Heap Signature</td>
<tr align=center>
<td colspan=4>B-tree Left-link Offset</td>
<tr align=center>
<td colspan=4><br>Length of Free-block #1<br> <br></td>
<tr align=center>
<td colspan=4><br>Offset of Free-block #1<br> <br></td>
<tr align=center>
<td colspan=4>.<br>.<br>.<br></td>
<tr align=center>
<td colspan=4><br>Length of Free-block #n<br> <br></td>
<tr align=center>
<td colspan=4><br>Offset of Free-block #n<br> <br></td>
<tr align=center>
<td colspan=4>"High" Offset</td>
<tr align=center>
<td colspan=4>Right-link Offset</td>
</table>
</center>
<p>
<dl>
<dt> The elements of the free-space heap page are described below:
<dd>
<dl>
<dt>Free-space Heap Signature: (4 bytes)
<dd>The ASCII character string <code>FREE</code>
is used to indicate the
beginning of a free-space heap B-tree page. This gives
file consistency checking utilities a better chance of
reconstructing a damaged file.
<dt>B-tree Left-link Offset: (<offset> bytes)
<dd>This value is used to indicate the offset of all offsets
in the B-link-tree which are smaller than the value of the
offset in entry #1. This value is also used to indicate a
leaf node in the B-link-tree by being set to all ones.
<dt>Length of Free-block #n: (<length> bytes)
<dd>This value indicates the length of an unused block in
the file.
<dt>Offset of Free-block #n: (<offset> bytes)
<dd>This value indicates the offset in the file of an
unused block in the file.
<dt>"High" Offset: (4-bytes)
<dd>This offset is used as the upper bound on offsets
contained within a page when the page has been split.
<dt>Right-link Offset: (<offset> bytes)
<dd>This value is used to indicate the offset of the next
child to the right of the parent of this group
page. When there is no node to the right, this value is
all zeros.
</dl>
</dl>
<p>The algorithms for searching and inserting objects in the
B-tree pages are described fully in the Lehman and Yao paper,
which should be read to provide a full description of the
B-tree's usage.
-->
<br><br>
<br><br>
<h2><a name="DataObject">Disk Format: Level 2 - Data Objects </a></h2>
<p>Data objects contain the real information in the file. These
objects compose the scientific data and other information which
are generally thought of as "data" by the end-user. All the
other information in the file is provided as a framework for
these data objects.
<p>A data object is composed of header information and data
information. The header information contains the information
needed to interpret the data information for the data object as
well as additional "meta-data" or pointers to additional
"meta-data" used to describe or annotate each data object.
<h3><a name="ObjectHeader">
Disk Format: Level 2a - Data Object Headers</a></h3>
<p>The header information of an object is designed to encompass
all the information about an object which would be desired to be
known, except for the data itself. This information includes
the dimensionality, number-type, information about how the data
is stored on disk (in external files, compressed, broken up in
blocks, etc.), as well as other information used by the library
to speed up access to the data objects or maintain a file's
integrity. The header of each object is not necessarily located
immediately prior to the object's data in the file and in fact
may be located in any position in the file.
<p>
<center>
<table border cellpadding=4 width="80%">
<caption align=top>
<B>Object Headers</B>
</caption>
<tr align=center>
<th width="25%">byte</th>
<th width="25%">byte</th>
<th width="25%">byte</th>
<th width="25%">byte</th>
</tr>
<tr align=center>
<td colspan=1 width="25%">Version # of Object Header</td>
<td colspan=1 width="25%">Reserved</td>
<td colspan=2 width="50%">Number of Header Messages</td>
</tr>
<tr align=center>
<td colspan=4>Object Reference Count</td>
</tr>
<tr align=center>
<td colspan=4><br>Total Object Header Size<br><br></td>
</tr>
<tr align=center>
<td colspan=2>Header Message Type #1</td>
<td colspan=2>Size of Header Message Data #1</td>
</tr>
<tr align=center>
<td>Flags</td>
<td colspan=3>Reserved</td>
</tr>
<tr align=center>
<td colspan=4><br>Header Message Data #1<br><br></td>
</tr>
<tr align=center>
<td colspan=4>.<br>.<br>.<br></td>
</tr>
<tr align=center>
<td colspan=2>Header Message Type #n</td>
<td colspan=2>Size of Header Message Data #n</td>
</tr>
<tr align=center>
<td>Flags</td>
<td colspan=3>Reserved</td>
</tr>
<tr align=center>
<td colspan=4><br>Header Message Data #n<br><br></td>
</tr>
</table>
</center>
<p>
<center>
<table align=center width="80%">
<tr>
<th width="30%">Field Name</th>
<th width="70%">Description</th>
</tr>
<tr valign=top>
<td>Version number of the object header</td>
<td>This value is used to determine the format of the
information in the object header. When the format of the
information in the object header is changed, the version number
is incremented and can be used to determine how the
information in the object header is formatted.</td>
</tr>
<tr valign=top>
<td>Reserved</td>
<td>Always set to zero.</td>
</tr>
<tr valign=top>
<td>Number of header messages</td>
<td>This value determines the number of messages listed in
this object header. This provides a fast way for software
to prepare storage for the messages in the header.</td>
</tr>
<tr valign=top>
<td>Object Reference Count</td>
<td>This value specifies the number of references to this
object within the current file. References to the
data object from external files are not tracked.</td>
</tr>
<tr valign=top>
<td>Total Object Header Size</td>
<td>This value specifies the total number of bytes of header
message data following this length field for the current
message as well as any continuation data located elsewhere
in the file.</td>
</tr>
<tr valign=top>
<td>Header Message Type</td>
<td>The header message type specifies the type of
information included in the header message data following
the type along with a small amount of other information.
Bit 15 of the message type is set if the message is
constant (constant messages cannot be changed since they
may be cached in group entries throughout the
file). The header message types for the pre-defined
header messages will be included in further discussion
below.</td>
</tr>
<tr valign=top>
<td>Size of Header Message Data</td>
<td>This value specifies the number of bytes of header
message data following the header message type and length
information for the current message. The size includes
padding bytes to make the message a multiple of eight
bytes.</td>
</tr>
<tr valign=top>
<td>Flags</td>
<td>This is a bit field with the following definition:
<dl>
<dt><code>0</code>
<dd>If set, the message data is constant. This is used
for messages like the datatype message of a dataset.
<dt><code>1</code>
<dd>If set, the message is stored in the global heap and
the Header Message Data field contains a Shared Object
message and the Size of Header Message Data field
contains the size of that Shared Object message.
<dt><code>2-7</code>
<dd>Reserved
</dl>
</td>
<tr valign=top>
<td>Header Message Data</td>
<td>The format and length of this field is determined by the
header message type and size respectively. Some header
message types do not require any data and this information
can be eliminated by setting the length of the message to
zero. The data is padded with enough zeros to make the
size a multiple of eight.</td>
</tr>
</table>
</center>
<p>The header message types and the message data associated with
them compose the critical "meta-data" about each object. Some
header messages are required for each object while others are
optional. Some optional header messages may also be repeated
several times in the header itself, the requirements and number
of times allowed in the header will be noted in each header
message description below.
<P>The following is a list of currently defined header messages:
<hr>
<h4><a name="NILMessage">Name: NIL</a></h4>
<b>Type: </b>0x0000<br>
<b>Length:</b> varies<br>
<b>Status:</b> Optional, may be repeated.<br>
<b>Purpose and Description:</b> The NIL message is used to
indicate a message
which is to be ignored when reading the header messages for a data object.
[Probably one which has been deleted for some reason.]<br>
<b>Format of Data:</b> Unspecified.<br>
<!-- Delete examples throughout doc
<b>Examples:</b> None.
-->
<hr>
<h4><a name="SimpleDataSpace">Name: Simple Dataspace</a></h4>
<b>Type: </b>0x0001<br>
<b>Length:</b> Varies according to the number of dimensions,
as described in the following table<br>
<b>Status:</b> The <em>Simple Dataspace</em> message is required
and may not be repeated. This message is currently used with
datasets and named dataspaces.<br>
<p>The <em>Simple Dataspace</em> message describes the number
of dimensions and size of each dimension that the data object
has. This message is only used for datasets which have a
simple, rectilinear grid layout; datasets requiring a more
complex layout (irregularly structured or unstructured grids, etc.)
must use the <em>Complex Dataspace</em> message for expressing
the space the dataset inhabits.
<i>(Note: The <em>Complex Dataspace</em> functionality is
not yet implemented (as of HDF5 Release 1.2). It is not described
in this document.)</i>
<p>
<center>
<table border cellpadding=4 width="80%">
<caption align=top>
<b>Simple Dataspace Message</b>
</caption>
<tr align=center>
<th width="25%">byte</th>
<th width="25%">byte</th>
<th width="25%">byte</th>
<th width="25%">byte</th>
</tr>
<tr align=center>
<td>Version</td>
<td>Dimensionality</td>
<td>Flags</td>
<td>Reserved</td>
</tr>
<tr align=center>
<td colspan=4>Reserved</td>
</tr>
<tr align=center>
<td colspan=4>Dimension Size #1 (<size> bytes)</td>
<tr align=center>
<td colspan=4>.<br>.<br>.<br></td>
<tr align=center>
<td colspan=4>Dimension Size #n (<size> bytes)</td>
<tr align=center>
<td colspan=4>Dimension Maximum #1 (<size> bytes)</td>
<tr align=center>
<td colspan=4>.<br>.<br>.<br></td>
<tr align=center>
<td colspan=4>Dimension Maximum #n (<size> bytes)</td>
<tr align=center>
<td colspan=4>Permutation Index #1</td>
<tr align=center>
<td colspan=4>.<br>.<br>.<br></td>
<tr align=center>
<td colspan=4>Permutation Index #n</td>
</table>
</center>
<p>
<center>
<table align=center width="80%">
<tr>
<th width="30%">Field Name</th>
<th width="70%">Description</th>
</tr>
<tr valign=top>
<td>Version </td>
<td>This value is used to determine the format of the
Simple Dataspace Message. When the format of the
information in the message is changed, the version number
is incremented and can be used to determine how the
information in the object header is formatted.</td>
</tr>
<tr valign=top>
<td>Dimensionality</td>
<td>This value is the number of dimensions that the data
object has.</td>
</tr>
<tr valign=top>
<td>Flags</td>
<td>This field is used to store flags to indicate the
presence of parts of this message. Bit 0 (the least
significant bit) is used to indicate that maximum
dimensions are present. Bit 1 is used to indicate that
permutation indices are present for each dimension.</td>
</tr>
<tr valign=top>
<td>Dimension Size #n (<size> bytes)</td>
<td>This value is the current size of the dimension of the
data as stored in the file. The first dimension stored in
the list of dimensions is the slowest changing dimension
and the last dimension stored is the fastest changing
dimension.</td>
</tr>
<tr valign=top>
<td>Dimension Maximum #n (<size> bytes)</td>
<td>This value is the maximum size of the dimension of the
data as stored in the file. This value may be the special
value <UNLIMITED> (all bits set) which indicates
that the data may expand along this dimension
indefinitely. If these values are not stored, the maximum
value of each dimension is assumed to be the same as the
current size value.</td>
</tr>
<tr valign=top>
<td>Permutation Index #n (4 bytes)</td>
<td>This value is the index permutation used to map
each dimension from the canonical representation to an
alternate axis for each dimension. If these values are
not stored, the first dimension stored in the list of
dimensions is the slowest changing dimension and the last
dimension stored is the fastest changing dimension.</td>
</tr>
</table>
</center>
<!-- Delete examples throughout doc
<h4>Examples</h4>
<dl>
<dt> Example #1
<dd>A sample 640 horizontally by 480 vertically raster image
dimension header. The number of dimensions would be set to 2
and the first dimension's size and maximum would both be set
to 480. The second dimension's size and maximum would both be
set to 640
.
<dt>Example #2
<dd>A sample 4 dimensional scientific dataset which is composed
of 30x24x3 slabs of data being written out in an unlimited
series every several minutes as timestep data (currently there
are five slabs). The number of dimensions is 4. The first
dimension size is 5 and its maximum is <UNLIMITED>. The
second through fourth dimension's size and maximum value are
set to 3, 24, and 30 respectively.
<dt>Example #3
<dd>A sample unlimited length text string, currently of length
83. The number of dimensions is 1, the size of the first
dimension is 83 and the maximum of the first dimension is set
to <UNLIMITED>, allowing further text data to be
appended to the string or possibly the string to be replaced
with another string of a different size. (This could also be
stored as a scalar dataset with number-type set to "string")
</dl>
-->
<!-- DELETE ENTIRE DATASPACE SECTION -->
<!--
<hr>
<h4><a name="DataSpaceMessage">Name: Complex Dataspace (Fiber Bundle?)</a></h4>
<b>Type: </b>0x0002<br>
<b>Length:</b> varies<br>
<b>Status:</b> One of the <em>Simple Dataspace</em> or
<em>Complex Dataspace</em> messages is required (but not both) and may
not be repeated.<br> <b>Purpose and Description:</b> The
<em>Dataspace</em> message describes space that the dataset is
mapped onto in a more comprehensive way than the <em>Simple
Dimensionality</em> message is capable of handling. The
dataspace of a dataset encompasses the type of coordinate system
used to locate the dataset's elements as well as the structure and
regularity of the coordinate system. The dataspace also
describes the number of dimensions which the dataset inhabits as
well as a possible higher dimensional space in which the dataset
is located within.
<br>
<b>Format of Data:</b>
<center>
<table border cellpadding=4 width="80%">
<caption align=bottom>
<B>HDF5 Dataspace Message Layout</B>
</caption>
<tr align=center>
<th width="25%">byte</th>
<th width="25%">byte</th>
<th width="25%">byte</th>
<th width="25%">byte</th>
<tr align=center>
<td colspan=4>Mesh Type</td>
<tr align=center>
<td colspan=4>Logical Dimensionality</td>
</table>
</center>
<p>
<dl>
<dt>The elements of the dimensionality message are described below:
<dd>
<dl>
<dt>Mesh Type: (unsigned 32-bit integer)
<dd>This value indicates whether the grid is
polar/spherical/cartesion,
structured/unstructured and regular/irregular. <br>
The mesh type value is broken up as follows: <br>
<P>
<center>
<table border cellpadding=4 width="80%">
<caption align=bottom>
<B>HDF5 Mesh-type Layout</B>
</caption>
<tr align=center>
<th width="25%">byte</th>
<th width="25%">byte</th>
<th width="25%">byte</th>
<th width="25%">byte</th>
<tr align=center>
<td colspan=1>Mesh Embedding</td>
<td colspan=1>Coordinate System</td>
<td colspan=1>Structure</td>
<td colspan=1>Regularity</td>
</table>
</center>
The following are the definitions of mesh-type bytes:
<dl>
<dt>Mesh Embedding
<dd>This value indicates whether the dataset dataspace
is located within
another dataspace or not:
<dl> <dl>
<dt><STANDALONE>
<dd>The dataset mesh is self-contained and is not
embedded in another mesh.
<dt><EMBEDDED>
<dd>The dataset's dataspace is located within
another dataspace, as
described in information below.
</dl> </dl>
<dt>Coordinate System
<dd>This value defines the type of coordinate system
used for the mesh:
<dl> <dl>
<dt><POLAR>
<dd>The last two dimensions are in polar
coordinates, higher dimensions are
cartesian.
<dt><SPHERICAL>
<dd>The last three dimensions are in spherical
coordinates, higher dimensions
are cartesian.
<dt><CARTESIAN>
<dd>All dimensions are in cartesian coordinates.
</dl> </dl>
<dt>Structure
<dd>This value defines the locations of the grid-points
on the axes:
<dl> <dl>
<dt><STRUCTURED>
<dd>All grid-points are on integral, sequential
locations, starting from 0.
<dt><UNSTRUCTURED>
<dd>Grid-points locations in each dimension are
explicitly defined and
may be of any numeric datatype.
</dl> </dl>
<dt>Regularity
<dd>This value defines the locations of the dataset
points on the grid:
<dl> <dl>
<dt><REGULAR>
<dd>All dataset elements are located at the
grid-points defined.
<dt><IRREGULAR>
<dd>Each dataset element has a particular
grid-location defined.
</dl> </dl>
</dl>
<p>The following grid combinations are currently allowed:
<dl> <dl>
<dt><POLAR-STRUCTURED-REGULAR>
<dt><SPHERICAL-STRUCTURED-REGULAR>
<dt><CARTESIAN-STRUCTURED-REGULAR>
<dt><POLAR-UNSTRUCTURED-REGULAR>
<dt><SPHERICAL-UNSTRUCTURED-REGULAR>
<dt><CARTESIAN-UNSTRUCTURED-REGULAR>
<dt><CARTESIAN-UNSTRUCTURED-IRREGULAR>
</dl> </dl>
All of the above grid types can be embedded within another
dataspace.
<br> <br>
<dt>Logical Dimensionality: (unsigned 32-bit integer)
<dd>This value is the number of dimensions that the dataset occupies.
<P>
<center>
<table border cellpadding=4 width="80%">
<caption align=bottom>
<B>HDF5 Dataspace Embedded Dimensionality Information</B>
</caption>
<tr align=center>
<th width="25%">byte</th>
<th width="25%">byte</th>
<th width="25%">byte</th>
<th width="25%">byte</th>
<tr align=center>
<td colspan=4>Embedded Dimensionality</td>
<tr align=center>
<td colspan=4>Embedded Dimension Size #1</td>
<tr align=center>
<td colspan=4>.<br>.<br>.<br></td>
<tr align=center>
<td colspan=4>Embedded Dimension Size #n</td>
<tr align=center>
<td colspan=4>Embedded Origin Location #1</td>
<tr align=center>
<td colspan=4>.<br>.<br>.<br></td>
<tr align=center>
<td colspan=4>Embedded Origin Location #n</td>
</table>
</center>
<dt>Embedded Dimensionality: (unsigned 32-bit integer)
<dd>This value is the number of dimensions of the space the
dataset is located
within. i.e. a planar dataset located within a 3-D space,
or a 3-D dataset
which is a subset of another 3-D space, etc.
<dt>Embedded Dimension Size: (unsigned 32-bit integer)
<dd>These values are the sizes of the dimensions of the
embedded dataspace
that the dataset is located within.
<dt>Embedded Origin Location: (unsigned 32-bit integer)
<dd>These values comprise the location of the dataset's
origin within the embedded dataspace.
</dl>
</dl>
[Comment: need some way to handle different orientations of the
dataset dataspace
within the embedded dataspace]<br>
<P>
<center>
<table border cellpadding=4 width="80%">
<caption align=bottom>
<B>HDF5 Dataspace Structured/Regular Grid Information</B>
</caption>
<tr align=center>
<th width="25%">byte</th>
<th width="25%">byte</th>
<th width="25%">byte</th>
<th width="25%">byte</th>
<tr align=center>
<td colspan=4>Logical Dimension Size #1</td>
<tr align=center>
<td colspan=4>Logical Dimension Maximum #1</td>
<tr align=center>
<td colspan=4>.<br>.<br>.<br></td>
<tr align=center>
<td colspan=4>Logical Dimension Size #n</td>
<tr align=center>
<td colspan=4>Logical Dimension Maximum #n</td>
</table>
</center>
<p>
<dl>
<dt>The elements of the dimensionality message are described below:
<dd>
<dl>
<dt>Logical Dimension Size #n: (unsigned 32-bit integer)
<dd>This value is the current size of the dimension of the
data as stored in
the file. The first dimension stored in the list of
dimensions is the slowest
changing dimension and the last dimension stored is the
fastest changing
dimension.
<dt>Logical Dimension Maximum #n: (unsigned 32-bit integer)
<dd>This value is the maximum size of the dimension of the
data as stored in
the file. This value may be the special value
<UNLIMITED> which
indicates that the data may expand along this dimension
indefinitely.
</dl>
</dl>
<P>
<center>
<table border cellpadding=4 width="80%">
<caption align=bottom>
<B>HDF5 Dataspace Structured/Irregular Grid Information</B>
</caption>
<tr align=center>
<th width="25%">byte</th>
<th width="25%">byte</th>
<th width="25%">byte</th>
<th width="25%">byte</th>
<tr align=center>
<td colspan=4># of Grid Points in Dimension #1</td>
<tr align=center>
<td colspan=4>.<br>.<br>.<br></td>
<tr align=center>
<td colspan=4># of Grid Points in Dimension #n</td>
<tr align=center>
<td colspan=4>Datatype of Grid Point Locations</td>
<tr align=center>
<td colspan=4>Location of Grid Points in Dimension #1</td>
<tr align=center>
<td colspan=4>.<br>.<br>.<br></td>
<tr align=center>
<td colspan=4>Location of Grid Points in Dimension #n</td>
</table>
</center>
<P>
<center>
<table border cellpadding=4 width="80%">
<caption align=bottom>
<B>HDF5 Dataspace Unstructured Grid Information</B>
</caption>
<tr align=center>
<th width="25%">byte</th>
<th width="25%">byte</th>
<th width="25%">byte</th>
<th width="25%">byte</th>
<tr align=center>
<td colspan=4># of Grid Points</td>
<tr align=center>
<td colspan=4>Datatype of Grid Point Locations</td>
<tr align=center>
<td colspan=4>Grid Point Locations<br>.<br>.<br></td>
</table>
</center>
<h4><a name="DataSpaceExample">Examples:</a></h4>
Need some good examples, this is complex!
-->
<hr>
<h4><a name="DataTypeMessage">Name: Datatype</a></h4>
<b>Type:</b> 0x0003<br>
<b>Length:</b> variable<br>
<b>Status:</b> One required per dataset or named datatype<br>
<p>The datatype message defines the datatype for each data point
of a dataset. A datatype can describe an atomic type like a
fixed- or floating-point type or a compound type like a C
struct. A datatype does not, however, describe how data points
are combined to produce a dataset. Datatypes are stored on disk
as a datatype message, which is a list of datatype classes and
their associated properties.
<p>
<center>
<table border cellpadding=4 width="80%">
<caption align=top>
<b>Datatype Message</b>
</caption>
<tr align=center>
<th width="25%">byte</th>
<th width="25%">byte</th>
<th width="25%">byte</th>
<th width="25%">byte</th>
</tr>
<tr align=center>
<td>Type Class and Version</td>
<td colspan=3>Class Bit Field</td>
</tr>
<tr align=center>
<td colspan=4>Size in Bytes (4 bytes)</td>
</tr>
<tr align=center>
<td colspan=4><br><br>Properties<br><br><br></td>
</tr>
</table>
</center>
<p>The Class Bit Field and Properties fields vary depending
on the Type Class, which is the low-order four bits of the Type
Class and Version field (the high-order four byte are the
version which should be set to the value one). The type class
is one of 0 (fixed-point number), 1 (floating-point number),
2 (date and time), 3 (text string), 4 (bit field), 5 (opaque),
6 (compound), 7 (reference), 8 (enumeration), or 9 (variable-length).
The Class Bit Field is zero and the size of the
Properties field is zero except for the cases noted here.
<p>
<center>
<table border cellpadding=4 width="80%">
<caption align=top>
<b>Bit Field for Fixed-point Numbers (Class 0)</b>
</caption>
<tr align=center>
<th width="10%">Bits</th>
<th width="90%">Meaning</th>
</tr>
<tr valign=top>
<td>0</td>
<td><b>Byte Order.</b> If zero, byte order is little-endian;
otherwise, byte order is big endian.</td>
</tr>
<tr valign=top>
<td>1, 2</td>
<td><b>Padding type.</b> Bit 1 is the lo_pad type and bit 2
is the hi_pad type. If a datum has unused bits at either
end, then the lo_pad or hi_pad bit is copied to those
locations.</td>
</tr>
<tr valign=top>
<td>3</td>
<td><b>Signed.</b> If this bit is set then the fixed-point
number is in 2's complement form.</td>
</tr>
<tr valign=top>
<td>4-23</td>
<td>Reserved (zero).</td>
</tr>
</table>
</center>
<p>
<center>
<table border cellpadding=4 width="80%">
<caption align=top>
<b>Properties for Fixed-point Numbers (Class 0)</b>
</caption>
<tr align=center>
<th width="25%">Byte</th>
<th width="25%">Byte</th>
<th width="25%">Byte</th>
<th width="25%">Byte</th>
</tr>
<tr align=center>
<td colspan=2>Bit Offset</td>
<td colspan=2>Bit Precision</td>
</tr>
</table>
</center>
<p>
<center>
<table border cellpadding=4 width="80%">
<caption align=top>
<b>Bit Field for Floating-point Numbers (Class 1)</b>
</caption>
<tr align=center>
<th width="10%">Bits</th>
<th width="90%">Meaning</th>
</tr>
<tr valign=top>
<td>0</td>
<td><b>Byte Order.</b> If zero, byte order is little-endian;
otherwise, byte order is big endian.</td>
</tr>
<tr valign=top>
<td>1, 2, 3</td>
<td><b>Padding type.</b> Bit 1 is the low bits pad type, bit 2
is the high bits pad type, and bit 3 is the internal bits
pad type. If a datum has unused bits at either or between
the sign bit, exponent, or mantissa, then the value of bit
1, 2, or 3 is copied to those locations.</td>
</tr>
<tr valign=top>
<td>4-5</td>
<td><b>Normalization.</b> The value can be 0 if there is no
normalization, 1 if the most significant bit of the
mantissa is always set (except for 0.0), and 2 if the most
signficant bit of the mantissa is not stored but is
implied to be set. The value 3 is reserved and will not
appear in this field.</td>
</tr>
<tr valign=top>
<td>6-7</td>
<td>Reserved (zero).</td>
</tr>
<tr valign=top>
<td>8-15</td>
<td><b>Sign.</b> This is the bit position of the sign
bit.</td>
</tr>
<tr valign=top>
<td>16-23</td>
<td>Reserved (zero).</td>
</tr>
</table>
</center>
<p>
<center>
<table border cellpadding=4 width="80%">
<caption align=top>
<b>Properties for Floating-point Numbers (Class 1)</b>
</caption>
<tr align=center>
<th width="25%">Byte</th>
<th width="25%">Byte</th>
<th width="25%">Byte</th>
<th width="25%">Byte</th>
</tr>
<tr align=center>
<td colspan=2>Bit Offset</td>
<td colspan=2>Bit Precision</td>
</tr>
<tr align=center>
<td>Exponent Location</td>
<td>Exponent Size in Bits</td>
<td>Mantissa Location</td>
<td>Mantissa Size in Bits</td>
</tr>
<tr align=center>
<td colspan=4>Exponent Bias</td>
</tr>
</table>
</center>
<p>
<center>
<table border cellpadding=4 width="80%">
<caption align=top>
<b>Bit Field for Strings (Class 3)</b>
</caption>
<tr align=center>
<th width="10%">Bits</th>
<th width="90%">Meaning</th>
</tr>
<tr valign=top>
<td>0-3</td>
<td><b>Padding type.</b> This four-bit value determines the
type of padding to use for the string. The values are:
<dl>
<dt><code>0</code> Null terminate.
<dd>A zero byte marks the end of the string and is
guaranteed to be present after converting a long
string to a short string. When converting a short
string to a long string the value is padded with
additional null characters as necessary.
<br><br>
<dt><code>1</code> Null pad.
<dd>Null characters are added to the end of the value
during conversions from short values to long values
but conversion in the opposite direction simply
truncates the value.
<br><br>
<dt><code>2</code> Space pad.
<dd>Space characters are added to the end of the value
during conversions from short values to long values
but conversion in the opposite direction simply
truncates the value. This is the Fortran
representation of the string.
<br><br>
<dt><code>3-15</code> Reserved.
<dd>These values are reserved for future use.
</dl>
</tr>
<tr valign=top>
<td>4-7</td>
<td><b>Character Set.</b> The character set to use for
encoding the string. The only character set supported is
the 8-bit ASCII (zero) so no translations have been defined
yet.</td>
</tr>
<tr valign=top>
<td>8-23</td>
<td>Reserved (zero).</td>
</tr>
</table>
</center>
<p>
<center>
<table border cellpadding=4 width="80%">
<caption align=top>
<b>Bit Field for Bitfield types (Class 4)</b>
</caption>
<tr align=center>
<th width="10%">Bits</th>
<th width="90%">Meaning</th>
</tr>
<tr valign=top>
<td>0</td>
<td><b>Byte Order.</b> If zero, byte order is little-endian;
otherwise, byte order is big endian.</td>
</tr>
<tr valign=top>
<td>1, 2</td>
<td><b>Padding type.</b> Bit 1 is the lo_pad type and bit 2
is the hi_pad type. If a datum has unused bits at either
end, then the lo_pad or hi_pad bit is copied to those
locations.</td>
</tr>
<tr valign=top>
<td>3-23</td>
<td>Reserved (zero).</td>
</tr>
</table>
</center>
<p>
<center>
<table border cellpadding=4 width="80%">
<caption align=top>
<b>Properties for Bitfield types (Class 4)</b>
</caption>
<tr align=center>
<th width="25%">Byte</th>
<th width="25%">Byte</th>
<th width="25%">Byte</th>
<th width="25%">Byte</th>
</tr>
<tr align=center>
<td colspan=2>Bit Offset</td>
<td colspan=2>Bit Precision</td>
</tr>
</table>
</center>
<p>
<center>
<table border cellpadding=4 width="80%">
<caption align=top>
<b>Bit Field for Opaque types (Class 5)</b>
</caption>
<tr align=center>
<th width="10%">Bits</th>
<th width="90%">Meaning</th>
</tr>
<tr valign=top>
<td>0-23</td>
<td>Reserved (zero).</td>
</tr>
</table>
</center>
<p>
<center>
<table border cellpadding=4 width="80%">
<caption align=top>
<b>Properties for Opaque types (Class 5)</b>
</caption>
<tr align=center>
<th width="25%">Byte</th>
<th width="25%">Byte</th>
<th width="25%">Byte</th>
<th width="25%">Byte</th>
</tr>
<tr align=center>
<td colspan=4><br>Null-terminated ASCII Tag<br>
(multiple of 8 bytes)<br><br></td>
</tr>
</table>
</center>
<p>
<center>
<table border cellpadding=4 width="80%">
<caption align=top>
<b>Bit Field for Compound Types (Class 6)</b>
</caption>
<tr align=center>
<th width="10%">Bits</th>
<th width="90%">Meaning</th>
</tr>
<tr valign=top>
<td>0-15</td>
<td><b>Number of Members.</b> This field contains the number
of members defined for the compound datatype. The member
definitions are listed in the Properties field of the data
type message.
</tr>
<tr valign=top>
<td>15-23</td>
<td>Reserved (zero).</td>
</tr>
</table>
</center>
<p>The Properties field of a compound datatype is a list of the
member definitions of the compound datatype. The member
definitions appear one after another with no intervening bytes.
The member types are described with a recursive datatype
message.
<p>
<center>
<table border cellpadding=4 width="80%">
<caption align=top>
<b>Properties for Compound Types (Class 6)</b>
</caption>
<tr align=center>
<th width="25%">Byte</th>
<th width="25%">Byte</th>
<th width="25%">Byte</th>
<th width="25%">Byte</th>
</tr>
<tr align=center>
<td colspan=4><br><br>Name (null terminated, multiple of
eight bytes)<br><br><br></td>
</tr>
<tr align=center>
<td colspan=4>Byte Offset of Member in Compound Instance</td>
</tr>
<tr align=center>
<td>Dimensionality</td>
<td colspan=3>reserved</td>
</tr>
<tr align=center>
<td colspan=4>Dimension Permutation</td>
</tr>
<tr align=center>
<td colspan=4>Reserved</td>
</tr>
<tr align=center>
<td colspan=4>Size of Dimension 0 (required)</td>
</tr>
<tr align=center>
<td colspan=4>Size of Dimension 1 (required)</td>
</tr>
<tr align=center>
<td colspan=4>Size of Dimension 2 (required)</td>
</tr>
<tr align=center>
<td colspan=4>Size of Dimension 3 (required)</td>
</tr>
<tr align=center>
<td colspan=4><br><br>Member Type Message<br><br><br></td>
</tr>
</table>
</center>
<p>
<center>
<table border cellpadding=4 width="80%">
<caption align=top>
<b>Bit Field for Enumeration types (Class 8)</b>
</caption>
<tr align=center>
<th width="10%">Bits</th>
<th width="90%">Meaning</th>
</tr>
<tr valign=top>
<td>0-15</td>
<td><b>Number of Members.</b> The number of name/value
pairs defined for the enumeration type.</td>
</tr>
<tr valign=top>
<td>16-23</td>
<td>Reserved (zero).</td>
</tr>
</table>
</center>
<p>
<center>
<table border cellpadding=4 width="80%">
<caption align=top>
<b>Properties for Enumeration types (Class 8)</b>
</caption>
<tr align=center>
<th width="25%">Byte</th>
<th width="25%">Byte</th>
<th width="25%">Byte</th>
<th width="25%">Byte</th>
</tr>
<tr align=center>
<td colspan=4><br>Parent Type<br><br></td>
</tr>
<tr align=center>
<td colspan=4><br>Names<br><br></td>
</tr>
<tr align=center>
<td colspan=4><br>Values<br><br></td>
</tr>
</table>
</center>
<center>
<table border=0 cellpadding=4 width="80%">
<tr align=left valign=top>
<td valign=top width=20%>Parent Type:</td>
<td valign=top>Each enumeration type is based on some parent type,
usually an integer. The information for that parent type is
described recursively by this field.</td>
</tr><tr align=left valign=top>
<td valign=top>Names:</td>
<td valign=top>The name for each name/value pair. Each name is
stored as a null terminated ASCII string in a multiple of
eight bytes. The names are in no particular order.</td>
</tr><tr align=left valign=top>
<td valign=top>Values:</td>
<td valign=top>The list of values in the same order as the names.
The values are packed (no inter-value padding) and the
size of each value is determined by the parent type.</td>
</tr>
</table>
</center>
<!--
<p>Datatype examples are <a href="Datatypes.html">here</a>.
-->
<hr>
<h4><a name="FillValueMessage">Name: Data Storage - Fill Value</a></h4>
<b>Type:</b> 0x0004<br>
<b>Length:</b> varies<br>
<b>Status:</b> Optional, may not be repeated.<br>
<p>The fill value message stores a single data point value which
is returned to the application when an uninitialized data point
is read from the dataset. The fill value is interpretted with
the same datatype as the dataset. If no fill value message is
present then a fill value of all zero is assumed.
<p>
<center>
<table border cellpadding=4 width="80%">
<caption align=top>
<b>Fill Value Message</b>
</caption>
<tr align=center>
<th width="25%">byte</th>
<th width="25%">byte</th>
<th width="25%">byte</th>
<th width="25%">byte</th>
</tr>
<tr align=center>
<td colspan=4>Size (4 bytes)</td>
</tr>
<tr align=center>
<td colspan=4><br>Fill Value<br><br></td>
</tr>
</table>
</center>
<p>
<center>
<table align=center width="80%">
<tr>
<th width="30%">Field Name</th>
<th width="70%">Description</th>
</tr>
<tr valign=top>
<td>Size (4 bytes)</td>
<td>This is the size of the Fill Value field in bytes.</td>
</tr>
<tr valign=top>
<td>Fill Value</td>
<td>The fill value. The bytes of the fill value are
interpreted using the same datatype as for the dataset.</td>
</tr>
</table>
</center>
<hr>
<h4><a name="ReservedMessage_0005">Name: Reserved - Not Assigned Yet</a></h4>
<b>Type:</b> 0x0005<br>
<b>Length:</b> N/A<br>
<b>Status:</b> N/A<br>
<hr>
<h4><a name="CompactDataStorageMessage">Name: Data Storage - Compact</a></h4>
<b>Type:</b> 0x0006<br>
<b>Length:</b> varies<br>
<b>Status:</b> Optional, may not be repeated.<br>
<p>This message indicates that the data for the data object is
stored within the current HDF file by including the actual
data as the header data for this message. The data is
stored internally in
the <em>normal format</em>, i.e. in one chunk, uncompressed, etc.
<P>Note that one and only one of the <em>Data Storage</em> headers can be
stored for each data object.
<P><b>Format of Data:</b> The message data is actually composed
of dataset data, so the format will be determined by the dataset
format.
<!-- Delete examples throughout doc
<h4><a name="CompactDataStorageExample">Examples:</a></h4>
[very straightforward]
-->
<hr>
<h4><a name="ExternalFileListMessage">Name: Data Storage -
External Data Files</a></h4>
<b>Type:</b> 0x0007<BR>
<b>Length:</b> varies<BR>
<b>Status:</b> Optional, may not be repeated.<BR>
<p><b>Purpose and Description:</b> The external object message
indicates that the data for an object is stored outside the HDF5
file. The filename of the object is stored as a Universal
Resource Location (URL) of the actual filename containing the
data. An external file list record also contains the byte offset
of the start of the data within the file and the amount of space
reserved in the file for that data.
<p>
<center>
<table border cellpadding=4 width="80%">
<caption align=top>
<b>External File List Message</b>
</caption>
<tr align=center>
<th width="25%">byte</th>
<th width="25%">byte</th>
<th width="25%">byte</th>
<th width="25%">byte</th>
</tr>
<tr align=center>
<td>Version</td>
<td colspan=3>Reserved</td>
</tr>
<tr align=center>
<td colspan=2>Allocated Slots</td>
<td colspan=2>Used Slots</td>
</tr>
<tr align=center>
<td colspan=4><br>Heap Address<br><br></td>
</tr>
<tr align=center>
<td colspan=4><br>Slot Definitions...<br><br></td>
</tr>
</table>
</center>
<p>
<center>
<table align=center width="80%">
<tr>
<th width="30%">Field Name</th>
<th width="70%">Description</th>
</tr>
<tr valign=top>
<td>Version </td>
<td>This value is used to determine the format of the
External File List Message. When the format of the
information in the message is changed, the version number
is incremented and can be used to determine how the
information in the object header is formatted.</td>
</tr>
<tr valign=top>
<td>Reserved</td>
<td>This field is reserved for future use.</td>
</tr>
<tr valign=top>
<td>Allocated Slots</td>
<td>The total number of slots allocated in the message. Its
value must be at least as large as the value contained in
the Used Slots field.</td>
</tr>
<tr valign=top>
<td>Used Slots</td>
<td>The number of initial slots which contain valid
information. The remaining slots are zero filled.</td>
</tr>
<tr valign=top>
<td>Heap Address</td>
<td>This is the address of a local name heap which contains
the names for the external files. The name at offset zero
in the heap is always the empty string.</td>
</tr>
<tr valign=top>
<td>Slot Definitions</td>
<td>The slot definitions are stored in order according to
the array addresses they represent. If more slots have
been allocated than what has been used then the defined
slots are all at the beginning of the list.</td>
</tr>
</table>
</center>
<p>
<center>
<table border cellpadding=4 width="80%">
<caption align=top>
<b>External File List Slot</b>
</caption>
<tr align=center>
<th width="25%">byte</th>
<th width="25%">byte</th>
<th width="25%">byte</th>
<th width="25%">byte</th>
</tr>
<tr align=center>
<td colspan=4><br>Name Offset (<size> bytes)<br><br></td>
</tr>
<tr align=center>
<td colspan=4><br>File Offset (<size> bytes)<br><br></td>
</tr>
<tr align=center>
<td colspan=4><br>Size<br><br></td>
</tr>
</table>
</center>
<p>
<center>
<table align=center width="80%">
<tr>
<th width="30%">Field Name</th>
<th width="70%">Description</th>
</tr>
<tr valign=top>
<td>Name Offset (<size> bytes)</td>
<td>The byte offset within the local name heap for the name
of the file. File names are stored as a URL which has a
protocol name, a host name, a port number, and a file
name:
<code><em>protocol</em>:<em>port</em>//<em>host</em>/<em>file</em></code>.
If the protocol is omitted then "file:" is assumed. If
the port number is omitted then a default port for that
protocol is used. If both the protocol and the port
number are omitted then the colon can also be omitted. If
the double slash and host name are omitted then
"localhost" is assumed. The file name is the only
mandatory part, and if the leading slash is missing then
it is relative to the application's current working
directory (the use of relative names is not
recommended).</td>
</tr>
<tr valign=top>
<td>File Offset (<size> bytes)</td>
<td>This is the byte offset to the start of the data in the
specified file. For files that contain data for a single
dataset this will usually be zero.</td>
</tr>
<tr valign=top>
<td>Size</td>
<td>This is the total number of bytes reserved in the
specified file for raw data storage. For a file that
contains exactly one complete dataset which is not
extendable, the size will usually be the exact size of the
dataset. However, by making the size larger one allows
HDF5 to extend the dataset. The size can be set to a value
larger than the entire file since HDF5 will read zeros
past the end of the file without failing.</td>
</tr>
</table>
</center>
<hr>
<h4><a name="LayoutMessage">Name: Data Storage - Layout</a></h4>
<b>Type:</b> 0x0008<BR>
<b>Length:</b> varies<BR>
<b>Status:</b> Required for datasets, may not be repeated.
<p><b>Purpose and Description:</b> Data layout describes how the
elements of a multi-dimensional array are arranged in the linear
address space of the file. Two types of data layout are
supported:
<ol>
<li>The array can be stored in one contiguous area of the file.
The layout requires that the size of the array be constant and
does not permit chunking, compression, checksums, encryption,
etc. The message stores the total size of the array and the
offset of an element from the beginning of the storage area is
computed as in C.
<li>The array domain can be regularly decomposed into chunks and
each chunk is allocated separately. This layout supports
arbitrary element traversals, compression, encryption, and
checksums, and the chunks can be distributed across external
raw data files (these features are described in other
messages). The message stores the size of a chunk instead of
the size of the entire array; the size of the entire array can
be calculated by traversing the B-tree that stores the chunk
addresses.
</ol>
<p>
<center>
<table border cellpadding=4 width="80%">
<caption align=top>
<B>Data Layout Message</B>
</caption>
<tr align=center>
<th width="25%">byte</th>
<th width="25%">byte</th>
<th width="25%">byte</th>
<th width="25%">byte</th>
</tr>
<tr align=center>
<td>Version</td>
<td>Dimensionality</td>
<td>Layout Class</td>
<td>Reserved</td>
</tr>
<tr align=center>
<td colspan=4>Reserved</td>
</tr>
<tr align=center>
<td colspan=4><br>Address<br><br></td>
</tr>
<tr align=center>
<td colspan=4>Dimension 0 (4-bytes)</td>
</tr>
<tr align=center>
<td colspan=4>Dimension 1 (4-bytes)</td>
</tr>
<tr align=center>
<td colspan=4>...</td>
</tr>
</table>
</center>
<p>
<center>
<table align=center width="80%">
<tr>
<th width="30%">Field Name</th>
<th width="70%">Description</th>
</tr>
<tr valign=top>
<td>Version</td>
<td>A version number for the layout message. This
documentation describes version one.</td>
</tr>
<tr valign=top>
<td>Dimensionality</td>
<td>An array has a fixed dimensionality. This field
specifies the number of dimension size fields later in the
message.</td>
</tr>
<tr valign=top>
<td>Layout Class</td>
<td>The layout class specifies how the other fields of the
layout message are to be interpreted. A value of one
indicates contiguous storage while a value of two
indicates chunked storage. Other values will be defined
in the future.</td>
</tr>
<tr valign=top>
<td>Address</td>
<td>For contiguous storage, this is the address of the first
byte of storage. For chunked storage this is the address
of the B-tree that is used to look up the addresses of the
chunks.</td>
</tr>
<tr valign=top>
<td>Dimensions</td>
<td>For contiguous storage the dimensions define the entire
size of the array while for chunked storage they define
the size of a single chunk.</td>
</tr>
</table>
</center>
<hr>
<h4><a name="ReservedMessage_0009">Name: Reserved - Not Assigned Yet</a></h4>
<b>Type:</b> 0x0009<BR>
<b>Length:</b> N/A<BR>
<b>Status:</b> N/A<BR>
<b>Purpose and Description:</b> N/A<BR>
<b>Format of Data:</b> N/A
<hr>
<h4><a name="ReservedMessage_000A">Name: Reserved - Not Assigned Yet</a></h4>
<b>Type:</b> 0x000A<BR>
<b>Length:</b> N/A<BR>
<b>Status:</b> N/A<BR>
<b>Purpose and Description:</b> N/A<BR>
<b>Format of Data:</b> N/A
<hr>
<h4><a name="FilterMessage">Name: Data Storage - Filter Pipeline</a></h4>
<b>Type:</b> 0x000B<BR>
<b>Length:</b> varies<BR>
<b>Status:</b> Optional, may not be repeated.
<p><b>Purpose and Description:</b> This message describes the
filter pipeline which should be applied to the data stream by
providing filter identification numbers, flags, a name, an
client data.
<p>
<center>
<table border align=center cellpadding=4 witdh="80%">
<caption align=top>
<b>Filter Pipeline Message</b>
</caption>
<tr align=center>
<th width="25%">byte</th>
<th width="25%">byte</th>
<th width="25%">byte</th>
<th width="25%">byte</th>
</tr>
<tr align=center>
<td>Version</td>
<td>Number of Filters</td>
<td colspan=2>Reserved</td>
</tr>
<tr align=center>
<td colspan=4>Reserved</td>
</tr>
<tr align=center>
<td colspan=4><br>Filter List<br><br></td>
</tr>
</table>
</center>
<p>
<center>
<table align=center width="80%">
<tr>
<th width="30%">Field Name</th>
<th width="70%">Description</th>
</tr>
<tr valign=top>
<td>Version</td>
<td>The version number for this message. This document
describes version one.</td>
</tr>
<tr valign=top>
<td>Number of Filters</td>
<td>The total number of filters described by this
message. The maximum possible number of filters in a
message is 32.</td>
</tr>
<tr valign=top>
<td>Filter List</td>
<td>A description of each filter. A filter description
appears in the next table.</td>
</tr>
</table>
</center>
<p>
<center>
<table border align=center cellpadding=4 witdh="80%">
<caption align=top>
<b>Filter Pipeline Message</b>
</caption>
<tr align=center>
<th width="25%">byte</th>
<th width="25%">byte</th>
<th width="25%">byte</th>
<th width="25%">byte</th>
</tr>
<tr align=center>
<td colspan=2>Filter Identification</td>
<td colspan=2>Name Length</td>
</tr>
<tr align=center>
<td colspan=2>Flags</td>
<td colspan=2>Client Data Number of Values</td>
</tr>
<tr align=center>
<td colspan=4><br>Name<br><br></td>
</tr>
<tr align=center>
<td colspan=4><br>Client Data<br><br></td>
</tr>
<tr align=center>
<td colspan=4>Padding</td>
</tr>
</table>
</center>
<p>
<center>
<table align=center width="80%">
<tr>
<th width="30%">Field Name</th>
<th width="70%">Description</th>
</tr>
<tr valign=top>
<td>Filter Identification</td>
<td>This is a unique (except in the case of testing)
identifier for the filter. Values from zero through 255
are reserved for filters defined by the NCSA HDF5
library. Values 256 through 511 have been set aside for
use when developing/testing new filters. The remaining
values are allocated to specific filters by contacting the
<a href="mailto:hdf5dev@ncsa.uiuc.edu">HDF5 Development
Team</a>.</td>
</tr>
<tr valign=top>
<td>Name Length</td>
<td>Each filter has an optional null-terminated ASCII name
and this field holds the length of the name including the
null termination padded with nulls to be a multiple of
eight. If the filter has no name then a value of zero is
stored in this field.</td>
</tr>
<tr valign=top>
<td>Flags</td>
<td>The flags indicate certain properties for a filter. The
bit values defined so far are:
<dl>
<dt><code>bit 1</code>
<dd>If set then the filter is an optional filter.
During output, if an optional filter fails it will be
silently removed from the pipeline.
</dl>
</tr>
<tr valign=top>
<td>Client Data Number of Values</td>
<td>Each filter can store a few integer values to control
how the filter operates. The number of entries in the
Client Data array is stored in this field.</td>
</tr>
<tr valign=top>
<td>Name</td>
<td>If the Name Length field is non-zero then it will
contain the size of this field, a multiple of eight. This
field contains a null-terminated, ASCII character
string to serve as a comment/name for the filter.</td>
</tr>
<tr valign=top>
<td>Client Data</td>
<td>This is an array of four-byte integers which will be
passed to the filter function. The Client Data Number of
Values determines the number of elements in the
array.</td>
</tr>
<tr valign=top>
<td>Padding</td>
<td>Four bytes of zeros are added to the message at this
point if the Client Data Number of Values field contains
an odd number.</td>
</tr>
</table>
</center>
<hr>
<h4><a name="AttributeMessage">Name: Attribute</a></h4>
<b>Type:</b> 0x000C<BR>
<b>Length:</b> varies<BR>
<b>Status:</b> Optional, may be repeated.<BR>
<p><b>Purpose and Description:</b> The <em>Attribute</em>
message is used to list objects in the HDF file which are used
as attributes, or "meta-data" about the current object. An
attribute is a small dataset; it has a name, a datatype, a data
space, and raw data. Since attributes are stored in the object
header they must be relatively small (<64kb) and can be
associated with any type of object which has an object header
(groups, datasets, named types and spaces, etc.).
<p>
<center>
<table border align=center cellpadding=4 width="80%">
<caption align=top>
<b>Attribute Message</b>
</caption>
<tr align=center>
<th width="25%">byte</th>
<th width="25%">byte</th>
<th width="25%">byte</th>
<th width="25%">byte</th>
</tr>
<tr align=center>
<td>Version</td>
<td>Reserved</td>
<td colspan=2>Name Size</td>
</tr>
<tr align=center>
<td colspan=2>Type Size</td>
<td colspan=2>Space Size</td>
</tr>
<tr align=center>
<td colspan=4><br>Name<br><br></td>
</tr>
<tr align=center>
<td colspan=4><br>Type<br><br></td>
</tr>
<tr align=center>
<td colspan=4><br>Space<br><br></td>
</tr>
<tr align=center>
<td colspan=4><br>Data<br><br></td>
</tr>
</table>
</center>
<p>
<center>
<table align=center width="80%">
<tr>
<th width="30%">Field Name</th>
<th width="70%">Description</th>
</tr>
<tr valign=top>
<td>Version</td>
<td>Version number for the message. This document describes
version 1 of attribute messages.</td>
</tr>
<tr valign=top>
<td>Reserved</td>
<td>This field is reserved for later use and is set to
zero.</td>
</tr>
<tr valign=top>
<td>Name Size</td>
<td>The length of the attribute name in bytes including the
null terminator. Note that the Name field below may
contain additional padding not represented by this
field.</td>
</tr>
<tr valign=top>
<td>Type Size</td>
<td>The length of the datatype description in the Type
field below. Note that the Type field may contain
additional padding not represented by this field.</td>
</tr>
<tr valign=top>
<td>Space Size</td>
<td>The length of the dataspace description in the Space
field below. Note that the Space field may contain
additional padding not represented by this field.</td>
</tr>
<tr valign=top>
<td>Name</td>
<td>The null-terminated attribute name. This field is
padded with additional null characters to make it a
multiple of eight bytes.</td>
</tr>
<tr valign=top>
<td>Type</td>
<td>The datatype description follows the same format as
described for the datatype object header message. This
field is padded with additional zero bytes to make it a
multiple of eight bytes.</td>
</tr>
<tr valign=top>
<td>Space</td>
<td>The dataspace description follows the same format as
described for the dataspace object header message. This
field is padded with additional zero bytes to make it a
multiple of eight bytes.</td>
</tr>
<tr valign=top>
<td>Data</td>
<td>The raw data for the attribute. The size is determined
from the datatype and dataspace descriptions. This
field is <em>not</em> padded with additional zero
bytes.</td>
</tr>
</table>
</center>
<hr>
<h4><a name="NameMessage">Name: Object Name</a></h4>
<p><b>Type:</b> 0x000D<br>
<b>Length:</b> varies<br>
<b>Status:</b> Optional, may not be repeated.
<p><b>Purpose and Description:</b> The object name or comment is
designed to be a short description of an object. An object name
is a sequence of non-zero (<code>\0</code>) ASCII characters with no other
formatting included by the library.
<p>
<center>
<table border align=center cellpadding=4 width="80%">
<caption align=top>
<b>Name Message</b>
</caption>
<tr align=center>
<th width="25%">byte</th>
<th width="25%">byte</th>
<th width="25%">byte</th>
<th width="25%">byte</th>
</tr>
<tr align=center>
<td colspan=4><br>Name<br><br></td>
</tr>
</table>
</center>
<p>
<center>
<table align=center width="80%">
<tr>
<th width="30%">Field Name</th>
<th width="70%">Description</th>
</tr>
<tr valign=top>
<td>Name</td>
<td>A null terminated ASCII character string.</td>
</tr>
</table>
</center>
<hr>
<h4><a name="ModifiedMessage">Name: Object Modification Date & Time</a></h4>
<p><b>Type:</b> 0x000E<br>
<b>Length:</b> fixed<br>
<b>Status:</b> Optional, may not be repeated.
<p><b>Purpose and Description:</b> The object modification date
and time is a timestamp which indicates (using ISO-8601 date and
time format) the last modification of an object. The time is
updated when any object header message changes according to the
system clock where the change was posted.
<p>
<center>
<table border align=center cellpadding=4 width="80%">
<caption align=top>
<b>Modification Time Message</b>
</caption>
<tr align=center>
<th width="25%">byte</th>
<th width="25%">byte</th>
<th width="25%">byte</th>
<th width="25%">byte</th>
</tr>
<tr align=center>
<td colspan=4>Year</td>
</tr>
<tr align=center>
<td colspan=2>Month</td>
<td colspan=2>Day of Month</td>
</tr>
<tr align=center>
<td colspan=2>Hour</td>
<td colspan=2>Minute</td>
</tr>
<tr align=center>
<td colspan=2>Second</td>
<td colspan=2>Reserved</td>
</tr>
</table>
</center>
<p>
<center>
<table align=center width="80%">
<tr>
<th width="30%">Field Name</th>
<th width="70%">Description</th>
</tr>
<tr valign=top>
<td>Year</td>
<td>The four-digit year as an ASCII string. For example,
<code>1998</code>. All fields of this message should be interpreted
as coordinated universal time (UTC)</td>
</tr>
<tr valign=top>
<td>Month</td>
<td>The month number as a two digit ASCII string where
January is <code>01</code> and December is <code>12</code>.</td>
</tr>
<tr valign=top>
<td>Day of Month</td>
<td>The day number within the month as a two digit ASCII
string. The first day of the month is <code>01</code>.</td>
</tr>
<tr valign=top>
<td>Hour</td>
<td>The hour of the day as a two digit ASCII string where
midnight is <code>00</code> and 11:00pm is <code>23</code>.</td>
</tr>
<tr valign=top>
<td>Minute</td>
<td>The minute of the hour as a two digit ASCII string where
the first minute of the hour is <code>00</code> and
the last is <code>59</code>.</td>
</tr>
<tr valign=top>
<td>Second</td>
<td>The second of the minute as a two digit ASCII string
where the first second of the minute is <code>00</code>
and the last is <code>59</code>.</td>
</tr>
<tr valign=top>
<td>Reserved</td>
<td>This field is reserved and should always be zero.</td>
</tr>
</table>
</center>
<hr>
<h4><a name="SharedMessage">Name: Shared Object Message</a></h4>
<b>Type:</b> 0x000F<br>
<b>Length:</b> 4 Bytes<br>
<b>Status:</b> Optional, may be repeated.
<p>A constant message can be shared among several object headers
by writing that message in the global heap and having the object
headers all point to it. The pointing is accomplished with a
Shared Object message which is understood directly by the object
header layer of the library. It is also possible to have a
message of one object header point to a message in some other
object header, but care must be exercised to prevent cycles.
<p>If a message is shared, then the message appears in the global
heap and its message ID appears in the Header Message Type
field of the object header. Also, the Flags field in the object
header for that message will have bit two set (the
<code>H5O_FLAG_SHARED</code> bit). The message body in the
object header will be that of a Shared Object message defined
here and not that of the pointed-to message.
<p>
<center>
<table border cellpadding=4 width="80%">
<caption align=top>
<b>Shared Message Message</b>
</caption>
<tr align=center>
<th width="25%">byte</td>
<th width="25%">byte</td>
<th width="25%">byte</td>
<th width="25%">byte</td>
</tr>
<tr align=center>
<td>Version</td>
<td>Flags</td>
<td colspan=2>Reserved</td>
</tr>
<tr align=center>
<td colspan=4>Reserved</td>
</tr>
<tr align=center>
<td colspan=4><br>Pointer<br><br></td>
</tr>
</table>
</center>
<p>
<center>
<table align=center width="80%">
<tr>
<th width="30%">Field Name</th>
<th width="70%">Description</th>
</tr>
<tr valign=top>
<td>Version</td>
<td>The version number for the message. This document
describes version one of shared messages.</td>
</tr>
<tr valign=top>
<td>Flags</td>
<td>The Shared Message message points to a message which is
shared among multiple object headers. The Flags field
describes the type of sharing:
<dl>
<dt><code>Bit 0</code>
<dd>If this bit is clear then the actual message is the
first message in some other object header; otherwise
the actual message is stored in the global heap.
<dt><code>Bits 2-7</code>
<dd>Reserved (always zero)
</dl>
</tr>
<tr valign=top>
<td>Pointer</td>
<td>This field points to the actual message. The format of
the pointer depends on the value of the Flags field. If
the actual message is in the global heap then the pointer
is the file address of the global heap collection that
holds the message, and a four-byte index into that
collection. Otherwise the pointer is a group entry
that points to some other object header.</td>
</tr>
</table>
</center>
<hr>
<h4><a name="ContinuationMessage">Name: Object Header Continuation</a></h4>
<b>Type:</b> 0x0010<BR>
<b>Length:</b> fixed<BR>
<b>Status:</b> Optional, may be repeated.<BR>
<b>Purpose and Description:</b> The object header continuation is the location
in the file of more header messages for the current data object. This can be
used when header blocks are large, or likely to change over time.<BR>
<b>Format of Data:</b><p>
The object header continuation is formatted as follows (assuming a 4-byte
length & offset are being used in the current file):
<P>
<center>
<table border cellpadding=4 width=60%>
<caption align=bottom>
<B>HDF5 Object Header Continuation Message Layout</B>
</caption>
<tr align=center>
<th width=25%>byte</th>
<th width=25%>byte</th>
<th width=25%>byte</th>
<th width=25%>byte</th>
<tr align=center>
<td colspan=4>Header Continuation Offset</td>
<tr align=center>
<td colspan=4>Header Continuation Length</td>
</table>
</center>
<P>
<dl>
<dt>The elements of the Header Continuation Message are described below:
<dd>
<dl>
<dt>Header Continuation Offset: (<offset> bytes)
<dd>This value is the offset in bytes from the beginning of the file where the
header continuation information is located.
<dt>Header Continuation Length: (<length> bytes)
<dd>This value is the length in bytes of the header continuation information in
the file.
</dl>
</dl>
<!-- Delete examples throughout doc
<h4><a name="ContinuationExample">Examples:</a></h4>
[straightforward]
-->
<hr>
<h4><a name="SymbolTableMessage">Name: Group Message</a></h4>
<b>Type:</b> 0x0011<BR>
<b>Length:</b> fixed<BR>
<b>Status:</b> Required for groups, may not be repeated.<BR>
<b>Purpose and Description:</b> Each group has a B-tree and a
name heap which are pointed to by this message.<BR>
<b>Format of data:</b>
<p>The group message is formatted as follows:
<p>
<center>
<table border cellpadding=4 width="80%">
<caption align=bottom>
<b>HDF5 Object Header Group Message Layout</b>
</caption>
<tr align=center>
<th width="25%">byte</th>
<th width="25%">byte</th>
<th width="25%">byte</th>
<th width="25%">byte</th>
<tr align=center>
<td colspan=4>B-tree Address</td>
<tr align=center>
<td colspan=4>Heap Address</td>
</table>
</center>
<P>
<dl>
<dt>The elements of the Group Message are described below:
<dd>
<dl>
<dt>B-tree Address (<offset> bytes)
<dd>This value is the offset in bytes from the beginning of the file
where the B-tree is located.
<dt>Heap Address (<offset> bytes)
<dd>This value is the offset in bytes from the beginning of the file
where the group name heap is located.
</dl>
</dl>
<h3><a name="SharedObjectHeader">Disk Format: Level 2b - Shared Data Object Headers</a></h3>
<P>In order to share header messages between several dataset objects, object
header messages may be placed into the global heap. Since these
messages require additional information beyond the basic object header message
information, the format of the shared message is detailed below.
<BR> <BR>
<center>
<table border cellpadding=4 width=60%>
<caption align=bottom>
<B>HDF5 Shared Object Header Message</B>
</caption>
<tr align=center>
<th width=25%>byte</th>
<th width=25%>byte</th>
<th width=25%>byte</th>
<th width=25%>byte</th>
<tr align=center>
<td colspan=4>Reference Count of Shared Header Message</td>
<tr align=center>
<td colspan=4><br> Shared Object Header Message<br> <br></td>
</table>
</center>
<p>
<dl>
<dt> The elements of the shared object header message are described below:
<dd>
<dl>
<dt>Reference Count of Shared Header Message: (32-bit unsigned integer)
<dd>This value is used to keep a count of the number of dataset objects which
refer to this message from their dataset headers. When this count reaches zero,
the shared message header may be removed from the global heap.
<dt>Shared Object Header Message: (various lengths)
<dd>The data stored for the shared object header message is formatted in the
same way as the private object header messages described in the object header
description earlier in this document and begins with the header message Type.
</dl>
</dl>
<h3><a name="DataStorage">Disk Format: Level 2c - Data Object Data Storage</a></h3>
<P>The data for an object is stored separately from the header
information in the file and may not actually be located in the HDF5 file
itself if the header indicates that the data is stored externally. The
information for each record in the object is stored according to the
dimensionality of the object (indicated in the dimensionality header message).
Multi-dimensional data is stored in C order [same as current scheme], i.e. the
"last" dimension changes fastest.
<P>Data whose elements are composed of simple number-types are stored in
native-endian IEEE format, unless they are specifically defined as being stored
in a different machine format with the architecture-type information from the
number-type header message. This means that each architecture will need to
[potentially] byte-swap data values into the internal representation for that
particular machine.
<P> Data with a "variable" sized number-type is stored in a data heap
internal to the HDF5 file. Global heap identifiers are stored in the
data object storage.
<P>Data whose elements are composed of pointer number-types are stored in several
different ways depending on the particular pointer type involved. Simple
pointers are just stored as the dataset offset of the object being pointed to with the
size of the pointer being the same number of bytes as offsets in the file.
Partial-object pointers are stored as a heap-ID which points to the following
information within the file-heap: an offset of the object pointed to, number-type
information (same format as header message), dimensionality information (same
format as header message), sub-set start and end information (i.e. a coordinate
location for each), and field start and end names (i.e. a [pointer to the]
string indicating the first field included and a [pointer to the] string name
for the last field).
<P>Data of a compound datatype is stored as a contiguous stream of the items
in the structure, with each item formatted according to its datatype.
<hr>
<center>
<table border=0 width=98%>
<tr><td valign=top align=left>
<a href="index.html">Other HDF5 documents and links</a> <br>
<a href="H5.intro.html">Introduction to HDF5</a> <br>
</td>
<td> </td>
<td valign=top align=right>
<a href="H5.user.html">HDF5 User Guide</a> <br>
<a href="RM_H5Front.html">HDF5 Reference Manual</a> <br>
</td></tr>
</table>
</center>
<hr>
<!--
<address><a href="mailto:koziol@ncsa.uiuc.edu">Quincey Koziol</a></address>
<address><a href="mailto:matzke@llnl.gov">Robb Matzke</a></address>
-->
<address><a href="mailto:hdfhelp@ncsa.uiuc.edu">HDF Help Desk</a></address>
<!-- hhmts start -->
Last modified: 8 March 2000
<!-- hhmts end -->
<br>
Describes HDF5 Release 1.4 Beta, December 2000
</body>
</html>