rewrote how-to

git-svn-id: https://svn.apache.org/repos/asf/jakarta/poi/trunk@352085 13f79535-47bb-0310-9956-ffa450edef68
This commit is contained in:
Mark Johnson 2002-02-11 04:15:29 +00:00
parent d3a5bd28f8
commit 18ffb78a0b
2 changed files with 369 additions and 190 deletions

View File

@ -0,0 +1,369 @@
<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE document PUBLIC "-//APACHE//DTD Documentation V1.0//EN" "../dtd/document-v10.dtd">
<document>
<header>
<authors>
<person email="mjohnson@apache.org" name="Marc Johnson" id="MJ"/>
</authors>
</header>
<body>
<s1 title="How To Use the POIFS APIs">
<p>This document describes how to use the POIFS APIs to read, write, and modify files that employ a POIFS-compatible data structure to organize their content.</p>
<s2 title="Revision History">
<ul>
<li>02.10.2002 - completely rewritten from original documents on <link href="https://sourceforge.net/cvs/?group_id=32701">sourceforge</link></li>
</ul>
</s2>
<s2 title="Target Audience">
<p>This document is intended for Java developers who need to use the POIFS APIs to read, write, or modify files that employ a POIFS-compatible data structure to organize their content. It is not necessary for developers to understand the POIFS data structures, and an explanation of those data structures is beyond the scope of this document. It is expected that the members of the target audience will understand the rudiments of a hierarchical file system, and familiarity with the event pattern employed by Java APIs such as AWT would be helpful.</p>
</s2>
<s2 title="Glossary">
<p>This document attempts to be consistent in its terminology, which is defined here:</p>
<table>
<tr>
<td><b>Term</b></td>
<td><b>Definition</b></td>
</tr>
<tr>
<td>Directory</td>
<td>A special file that may contain other directories and documents.</td>
</tr>
<tr>
<td>DirectoryEntry</td>
<td>Representation of a directory within another directory.</td>
</tr>
<tr>
<td>Document</td>
<td>A file containing data, such as word processing data or a spreadsheet workbook.</td>
</tr>
<tr>
<td>DocumentEntry</td>
<td>Representation of a document within a directory.</td>
</tr>
<tr>
<td>Entry</td>
<td>Representation of a file in a directory.</td>
</tr>
<tr>
<td>File</td>
<td>A named entity, managed and contained by the file system.</td>
</tr>
<tr>
<td>File System</td>
<td>The POIFS data structures, plus the contained directories and documents, which are maintained in a hierarchical directory structure.</td>
</tr>
<tr>
<td>Root Directory</td>
<td>The directory at the base of a file system. All file systems have a root directory. The POIFS APIs will not allow the root directory to be removed or renamed, but it can be accessed for the purpose of reading its contents or adding files (directories and documents) to it.</td>
</tr>
</table>
</s2>
</s1>
<s1 title="Reading a File System">
<p>This section covers reading a file system. There are two ways to read a file system; these techniques are sketched out in the following table, and then explained in greater depth in the sections following the table.</p>
<table>
<tr>
<td><b>Technique</b></td>
<td><b>Advantages</b></td>
<td><b>Disadvantages</b></td>
</tr>
<tr>
<td>Conventional Reading</td>
<td>
<ul>
<li>Simpler API similar to reading a conventional file system.</li>
<li>Can read documents in any order.</li>
</ul>
</td>
<td>
<ul>
<li>All files are resident in memory, whether your application needs them or not.</li>
</ul>
</td>
</tr>
<tr>
<td>Event-Driven Reading</td>
<td>
<ul>
<li>Reduced footprint -- only the documents you care about are processed.</li>
<li>Improved performance -- no time is wasted reading the documents you're not interested in.</li>
</ul>
</td>
<td>
<ul>
<li>More complicated API.</li>
<li>Need to know in advance which documents you want to read.</li>
<li>No control over the order in which the documents are read.</li>
<li>No way to go back and get additional documents except to re-read the file system, which may not be possible, e.g., if the file system is being read from an input stream that lacks random access support.</li>
</ul>
</td>
</tr>
</table>
<s2 title="Conventional Reading">
<p>In this technique for reading, the entire file system is loaded into memory, and the entire directory tree can be walked by an application, reading specific documents at the application's leisure.</p>
<s3 title="Preparation">
<p>Before an application can read a file from the file system, the file system needs to be loaded into memory. This is done by using the <code>org.apache.poi.poifs.filesystem.POIFSFileSystem</code> class. Once the file system has been loaded into memory, the application may need the root directory. The following code fragment will accomplish this preparation stage:</p>
<source>
// need an open InputStream; for a file-based system, this would be appropriate:
// InputStream stream = new FileInputStream(fileName);
POIFSFileSystem fs;
try
{
fs = new POIFSFileSystem(inputStream);
}
catch (IOException e)
{
// an I/O error occurred, or the InputStream did not provide a compatible
// POIFS data structure
}
DirectoryEntry root = fs.getRoot();</source>
<p>Assuming no exception was thrown, the file system can then be read.</p>
<p>Note: loading the file system can take noticeable time, particularly for large file systems.</p>
</s3>
<s3 title="Reading the Directory Tree">
<p>Once the file system has been loaded into memory and the root directory has been obtained, the root directory can be read. The following code fragment shows how to read the entries in an <code>org.apache.poi.poifs.filesystem.DirectoryEntry</code> instance:</p>
<source>
// dir is an instance of DirectoryEntry ...
for (Iterator iter = dir.getEntries(); iter.hasNext(); )
{
Entry entry = (Entry)iter.next();
System.out.println("found entry: " + entry.getName());
if (entry instanceof DirectoryEntry)
{
// .. recurse into this directory
}
else if (entry instanceof DocumentEntry)
{
// entry is a document, which you can read
}
else
{
// currently, either an Entry is a DirectoryEntry or a DocumentEntry,
// but in the future, there may be other entry subinterfaces. The
// internal data structure certainly allows for a lot more entry types.
}
}</source>
</s3>
<s3 title="Reading a Specific Document">
<p>There are a couple of ways to read a document, depending on whether the document resides in the root directory or in another directory. Either way, you will obtain an <code>org.apache.poi.poifs.filesystem.DocumentInputStream</code> instance.</p>
<s4 title="DocumentInputStream">
<p>The DocumentInputStream class is a simple implementation of InputStream that makes a few guarantees worth noting:</p>
<ul>
<li><code>available()</code> always returns the number of bytes in the document from your current position in the document.</li>
<li><code>markSupported()</code> returns <code>true</code>.</li>
<li><code>mark(int limit)</code> ignores the limit parameter; basically the method marks the current position in the document.</li>
<li><code>reset()</code> takes you back to the position when <code>mark()</code> was last called, or to the beginning of the document if <code>mark()</code> has not been called.</li>
<li><code>skip(long n)</code> will take you to your current position + n (but not past the end of the document).</li>
</ul>
<p>The behavior of <code>available</code> means you can read in a document in a single read call like this:</p>
<source>
byte[] content = new byte[ stream.available() ];
stream.read(content);
stream.close();</source>
<p>The combination of <code>mark</code>, <code>reset</code>, and <code>skip</code> provide the basic mechanisms needed for random access of the document contents.</p>
</s4>
<s4 title="Reading a Document From the Root Directory">
<p>If the document resides in the root directory, you can obtain a <code>DocumentInputStream</code> like this:</p>
<source>
// load file system
try
{
DocumentInputStream stream = filesystem.createDocumentInputStream(documentName);
// process data from stream
}
catch (IOException e)
{
// no such document, or the Entry represented by documentName is not a
// DocumentEntry
}</source>
</s4>
<s4 title="Reading a Document From an Arbitrary Directory">
<p>A more generic technique for reading a document is to obtain an <code>org.apache.poi.poifs.filesystem.DirectoryEntry</code> instance for the directory containing the desired document (recall that you can use <code>getRoot()</code> to obtain the root directory from its file system). From that DirectoryEntry, you can then obtain a <code>DocumentInputStream</code> like this:</p>
<source>
DocumentEntry document = (DocumentEntry)directory.getEntry(documentName);
DocumentInputStream stream = new DocumentInputStream(document);
</source>
</s4>
</s3>
</s2>
<s2 title="Event-Driven Reading">
<p>The event-driven API for reading documents is a little more complicated and requires that your application know, in advance, which files it wants to read. The benefit of using this API is that each document is in memory just long enough for your application to read it, and documents that you never read at all are not in memory at all. When you're finished reading the documents you wanted, the file system has no data structures associated with it at all and can be discarded.</p>
<s3 title="Preparation">
<p>The preparation phase involves creating an instance of <code>org.apache.poi.poifs.eventfilesystem.POIFSReader</code> and to then register one or more <code>org.apache.poi.poifs.eventfilesystem.POIFSReaderListener</code> instances with the <code>POIFSReader</code>.</p>
<source>
POIFSReader reader = new POIFSReader();
// register for everything
reader.registerListener(myOmnivorousListener);
// register for selective files
reader.registerListener(myPickyListener, "foo");
reader.registerListener(myPickyListener, "bar");
// register for selective files
reader.registerListener(myOtherPickyListener, new POIFSDocumentPath(),
"fubar");
reader.registerListener(myOtherPickyListener, new POIFSDocumentPath(
new String[] { "usr", "bin" ), "fubar");</source>
</s3>
<s3 title="POIFSReaderListener">
<p><code>org.apache.poi.poifs.eventfilesystem.POIFSReaderListener</code> is an interface used to register for documents. When a matching document is read by the <code>org.apache.poi.poifs.eventfilesystem.POIFSReader</code>, the <code>POIFSReaderListener</code> instance receives an <code>org.apache.poi.poifs.eventfilesystem.POIFSReaderEvent</code> instance, which contains an open <code>DocumentInputStream</code> and information about the document.</p>
<p>A <code>POIFSReaderListener</code> instance can register for individual documents, or it can register for all documents; once it has registered for all documents, subsequent (and previous!) registration requests for individual documents are ignored. There is no way to unregister a <code>POIFSReaderListener</code>.</p>
<p>Thus, it is possible to register a single <code>POIFSReaderListener</code> for multiple documents - one, some, or all documents. It is guaranteed that a single <code>POIFSReaderListener</code> will receive exactly one notification per registered document. There is no guarantee as to the order in which it will receive notification of its documents, as future implementations of <code>POIFSReader</code> are free to change the algorithm for walking the file system's directory structure.</p>
<p>It is also permitted to register more than one <code>POIFSReaderListener</code> for the same document. There is no guarantee of ordering for notification of <code>POIFSReaderListener</code> instances that have registered for the same document when <code>POIFSReader</code> processes that document.</p>
<p>It is guaranteed that all notifications occur in the same thread. A future enhancement may be made to provide multi-threaded notifications, but such an enhancement would very probably be made in a new reader class, a <code>ThreadedPOIFSReader</code> perhaps.</p>
<p>The following table describes the three ways to register a <code>POIFSReaderListener</code> for a document or set of documents:</p>
<table>
<tr>
<td><b>Method Signature</b></td>
<td><b>What it does</b></td>
</tr>
<tr>
<td>registerListener(POIFSReaderListener <b>listener</b>)</td>
<td>registers <b>listener</b> for all documents.</td>
</tr>
<tr>
<td>registerListener(POIFSReaderListener <b>listener</b>, String <b>name</b>)</td>
<td>registers <b>listener</b> for a document with the specified <b>name</b> in the root directory.</td>
</tr>
<tr>
<td>registerListener(POIFSReaderListener <b>listener</b>, POIFSDocumentPath <b>path</b>, String <b>name</b>)</td>
<td>registers <b>listener</b> for a document with the specified <b>name</b> in the directory described by <b>path</b></td>
</tr>
</table>
</s3>
<s3 title="POIFSDocumentPath">
<p>The <code>org.apache.poi.poifs.filesystem.POIFSDocumentPath</code> class is used to describe a directory in a POIFS file system. Since there are no reserved characters in the name of a file in a POIFS file system, a more traditional string-based solution for describing a directory, with special characters delimiting the components of the directory name, is not feasible. The constructors for the class are used as follows:</p>
<table>
<tr>
<td><b>Constructor example</b></td>
<td><b>Directory described</b></td>
</tr>
<tr>
<td>new POIFSDocumentPath()</td>
<td>The root directory.</td>
</tr>
<tr>
<td>new POIFSDocumentPath(null)</td>
<td>The root directory.</td>
</tr>
<tr>
<td>new POIFSDocumentPath(new String[ 0 ])</td>
<td>The root directory.</td>
</tr>
<tr>
<td>new POIFSDocumentPath(new String[ ] { "foo", "bar"} )</td>
<td>in Unix terminology, "/foo/bar".</td>
</tr>
<tr>
<td>new POIFSDocumentPath(new POIFSDocumentPath(new String[] { "foo" }), new String[ ] { "fu", "bar"} )</td>
<td>in Unix terminology, "/foo/fu/bar".</td>
</tr>
</table>
</s3>
<s3 title="Processing POIFSReaderEvent Events">
<p>Processing <code>org.apache.poi.poifs.eventfilesystem.POIFSReaderEvent</code> events is relatively easy. After all of the <code>POIFSReaderListener</code> instances have been registered with <code>POIFSReader</code>, the <code>POIFSReader.read(InputStream stream)</code> method is called.</p>
<p>Assuming that there are no problems with the data, as the <code>POIFSReader</code> processes the documents in the specified <code>InputStream</code>'s data, it calls registered <code>POIFSReaderListener</code> instances' <code>processPOIFSReaderEvent</code> method with a <code>POIFSReaderEvent</code> instance.</p>
<p>The <code>POIFSReaderEvent</code> instance contains information to identify the document (a <code>POIFSDocumentPath</code> object to identify the directory that the document is in, and the document name), and an open <code>DocumentInputStream</code> instance from which to read the document.</p>
</s3>
</s2>
</s1>
<s1 title="Writing a File System">
<p>Writing a file system is very much like reading a file system in that there are multiple ways to do so. You can load an existing file system into memory and modify it (removing files, renaming files) and/or add new files to it, and write it, or you can start with a new, empty file system:</p>
<source>
POIFSFileSystem fs = new POIFSFileSystem();</source>
<s2 title="The Naming of Names">
<p>There are two restrictions on the names of files in a file system that must be considered when creating files:</p>
<ol>
<li>The name of the file must not exceed 31 characters. If it does, the POIFS API will silently truncate the name to fit.</li>
<li>The name of the file must be unique within its containing directory. This seems pretty obvious, but if it isn't spelled out, there'll be hell to pay, to be sure. Uniqueness, of course, is determined <b>after</b> the name has been truncated, if the original name was too long to begin with.</li>
</ol>
</s2>
<s2 title="Creating a Document">
<p>A document can be created by acquiring a <code>DirectoryEntry</code> and calling one of the two <code>createDocument</code> methods:</p>
<table>
<tr>
<td><b>Method Signature</b></td>
<td><b>Advantages</b></td>
<td><b>Disadvantages</b></td>
</tr>
<tr>
<td>CreateDocument(String name, InputStream stream)</td>
<td>
<ul>
<li>Simple API.</li>
</ul>
</td>
<td>
<ul>
<li>Increased memory footprint (document is in memory until file system is written).</li>
</ul>
</td>
</tr>
<tr>
<td>CreateDocument(String name, int size, POIFSWriterListener writer)</td>
<td>
<ul>
<li>Decreased memory footprint (only very small documents are held in memory, and then only for a short time).</li>
</ul>
</td>
<td>
<ul>
<li>More complex API.</li>
<li>Determining document size in advance may be difficult.</li>
<li>Lose control over when document is to be written.</li>
</ul>
</td>
</tr>
</table>
<p>Unlike reading, you don't have to choose between the in-memory and event-driven writing models; both can co-exist in the same file system.</p>
<p>Writing is initiated when the <code>POIFSFileSystem</code> instance's <code>writeFilesystem()</code> method is called with an <code>OutputStream</code> to write to.</p>
<p>The event-driven model is quite similar to the event-driven model for reading, in that the file system calls your <code>org.apache.poi.poifs.filesystem.POIFSWriterListener</code> when it's time to write your document, just as the <code>POIFSReader</code> calls your <code>POIFSReaderListener</code> when it's time to read your document. Internally, when <code>writeFilesystem()</code> is called, the final POIFS data structures are created and are written to the specified <code>OutputStream</code>. When the file system needs to write a document out that was created with the event-driven model, it calls the <code>POIFSWriterListener</code> back, calling its <code>processPOIFSWriterEvent()</code> method, passing an <code>org.apache.poi.poifs.filesystem.POIFSWriterEvent</code> instance. This object contains the <code>POIFSDocumentPath</code> and name of the document, its size, and an open <code>org.apache.poi.poifs.filesystem.DocumentOutputStream</code> to which to write. A <code>DocumentOutputStream</code> is a wrapper over the <code>OutputStream</code> that was provided to the <code>POIFSFileSystem</code> to write to, and has the responsibility of making sure that the document your application writes fits within the size you specified for it.</p>
</s2>
<s2 title="Creating a Directory">
<p>Creating a directory is similar to creating a document, except that there's only one way to do so:</p>
<source>
DirectoryEntry createdDir = existingDir.createDirectory(name);</source>
</s2>
<s2 title="Using POIFSFileSystem Directly To Create a Document Or Directory">
<p>As with reading documents, it is possible to create a new document or directory in the root directory by using convenience methods of POIFSFileSystem.</p>
<table>
<tr>
<td>DirectoryEntry Method Signature</td>
<td>POIFSFileSystem Method Signature</td>
</tr>
<tr>
<td>createDocument(String name, InputStream stream)</td>
<td>createDocument(InputStream stream, String name)</td>
</tr>
<tr>
<td>createDocument(String name, int size, POIFSWriterListener writer)</td>
<td>createDocument(String name, int size, POIFSWriterListener writer)</td>
</tr>
<tr>
<td>createDirectory(String name)</td>
<td>createDirectory(String name)</td>
</tr>
</table>
</s2>
</s1>
<s1 title="Modifying a File System">
<p>It is possible to modify an existing POIFS file system, whether it's one your application has loaded into memory, or one which you are creating on the fly.</p>
<s2 title="Removing a Document">
<p>Removing a document is simple: you get the <code>Entry</code> corresponding to the document and call its <code>delete()</code> method. This is a boolean method, but should always return <code>true</code>, indicating that the operation succeeded.</p>
</s2>
<s2 title="Removing a Directory">
<p>Removing a directory is also simple: you get the <code>Entry</code> corresponding to the directory and call its <code>delete()</code> method. This is a boolean method, but, unlike deleting a document, may not always return <code>true</code>, indicating that the operation succeeded. Here are the reasons why the operation may fail:</p>
<ul>
<li>The directory still has files in it (to check, call <code>isEmpty()</code> on its DirectoryEntry; is the return value <code>false</code>?)</li>
<li>The directory is the root directory. You cannot remove the root directory.</li>
</ul>
</s2>
<s2 title="Renaming a File">
<p>Regardless of whether the file is a directory or a document, it can be renamed, with one exception - the root directory has a special name that is expected by the components of a major software vendor's office suite, and the POIFS API will not let that name be changed. Renaming is done by acquiring the file's corresponding <code>Entry</code> instance and calling its <code>renameTo</code> method, passing in the new name.</p>
<p>Like <code>delete</code>, <code>renameTo</code> returns <code>true</code> if the operation succeeded, otherwise <code>false</code>. Reasons for failure include these:</p>
<ul>
<li>The new name is the same as another file in the same directory. And don't forget - if the new name is longer than 31 characters, it <b>will</b> be silently truncated. In its original length, the new name may have been unique, but truncated to 31 characters, it may not be unique any longer.</li>
<li>You tried to rename the root directory.</li>
</ul>
</s2>
</s1>
</body>
</document>

View File

@ -1,190 +0,0 @@
<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2//EN">
<HTML>
<HEAD>
<META HTTP-EQUIV="CONTENT-TYPE" CONTENT="text/html; charset=iso-8859-1">
<TITLE></TITLE>
<META NAME="GENERATOR" CONTENT="StarOffice/5.2 (Linux)">
<META NAME="CREATED" CONTENT="20011031;14571450">
<META NAME="CHANGEDBY" CONTENT=" ">
<META NAME="CHANGED" CONTENT="20011230;13132100">
</HEAD>
<BODY>
<H1>POIFS HOW TO</H1>
<H2>How to use POIFS directly</H2>
<H3>Andrew C. Oliver - December 14, 2001</H3>
<DL>
<DD STYLE="margin-bottom: 0.2in">10.31.2001- initial revision for
build POI 0.12.3
</DD><DD STYLE="margin-bottom: 0.2in">
12.15.2001 - minor revisions - thread safety, entry modification,
name restrictions, and so on.</DD><DD STYLE="margin-bottom: 0.2in">
12.30.2001 - revised for POI 1.0-final - minor revisions
</DD></DL>
<H2>
Capabilities</H2>
<DL>
<DD STYLE="margin-bottom: 0.2in">This release of POIFS contains the
full functionality to read, write and modify (by recreation) files
in the format most commonly referred to as OLE 2 Compound Document
Format (proabably tm - Microsoft).
</DD></DL>
<H2>
Target Audience</H2>
<P>This release candidate is intended for general use. It is
considered to be production-ready. It has not yet been extensively
tested (especially in a high load multi-threaded server situation),
though it's been unit tested quite a bit. This release is considered
to be &quot;golden&quot; as it has been used by HSSF and other users
without problems for some time, and has not changed recently.
</P>
<H2>General Use</H2>
<H3>User API</H3>
<H4>High level description and overview</H4>
<P>Files written with the POIFS library are referred to as POIFS file
systems (or sometimes archives). The OLE 2 Compound Document format
is designed to mimic many of the characteristics of a pre-modern file
system (most similar to FAT). We make the distinction between POIFS
written files and &quot;native&quot; written OLE 2 Compound Document
Format files because while we believe POIFS to be a full, correct and
complete implementation, most of this was accomplished through
researching other open source implementations and flat out guesses.</P>
<P>This overview is in no way intended to be complete (for a more
intense discussion please see POIFSFormat.html in this same
directory), it should give you a good idea into the principals of a
POIFS file system. Please note that specific file formats such as XLS
(HSSF) or DOC utilize POIFS file systems to contain their data, POIFS
itself does not know how to interpret the archived data.</P>
<P>Every POIFS file system contains a hierarchy of directories
starting with the root (there is always one, and only one, root).
Each directory, including the root, may contain one or more
directories and/or documents. Every directory and document has a
name. The root directory has a name, but unlike other directories,
its name is fixed and cannot be renamed.</P>
<P><STRONG>The POIFS API was not designed to be, and is not,
thread-safe.</STRONG> Only one thread of control should ever
manipulate a specific POIFS file system over that file system's
lifetime. You can, of course, have multiple threads, each
manipulating a distinct POIFS file system instance.</P>
<H4>Writing a new one</H4>
<P>To create a new (from scratch) POIFS file system for writing to,
you simply create an instance of
<CODE>net.sourceforge.poi.poifs.filesystem.Filesystem</CODE> using
the default constructor (no arguments). Initially this POIFS file
system will be empty except for containing the essential root
directory.</P>
<P>From there you can create a directory entry by calling&nbsp;<CODE>
Filesystem.createDirectory(name)</CODE>, and passing in the name of
the directory. This will return an instance
of&nbsp;<CODE>net.sourceforge.poi.poifs.filesystem.DirectoryEntry</CODE>
. You can also create a document within the root directory by
calling&nbsp;<CODE> Filesystem.createDocument(name, inputstream)</CODE>,
and passing the name of the document and an instance
of&nbsp;<CODE>java.io.InputStream</CODE> from which the document's
data can be obtained. It is noted that, the most commonly used file
formats of the Microsoft Corporation such as DOC, XLS, etc. are all
POIFS-compatible file systems with documents stored in the root
directory.</P>
<P>Supposing the document is to be stored in a directory other than
the root, you take the instance of&nbsp;<CODE>DirectoryEntry</CODE>
that you created and call&nbsp;<CODE>createDocument(name,
inputstream)</CODE> on it instead. You can also create a child
directory by calling&nbsp;<CODE> createDirectory(name)</CODE>.
Alternatively you can call&nbsp;<CODE>Filesystem.getRoot()</CODE> and
use it just like any other directory entry.</P>
<P>When you've finished creating entries in the filesystem, simply
call&nbsp;<CODE> Filesystem.writeFilesystem(stream)</CODE> passing in
an instance of&nbsp;<CODE> java.io.OutputStream</CODE>. Be sure you
close the stream when you're done.</P>
<H5><A NAME="Names"></A>Names</H5>
<P>The POIFS file system imposes two limitations on document and
directory names:</P>
<OL>
<LI><P STYLE="margin-bottom: 0in">The names of documents and
directories must be unique within their containing directory. Pretty
obvious.
</P>
<LI><P>Names are restricted to 31 characters. If you create a
directory or document with a name longer than that, it will be
silently truncated. When truncated, it may conflict with the name of
another directory or document, and the create operation will fail.
</P>
</OL>
<H5>Why not Readers and Writers?</H5>
<P>The POIFS file system uses Streams because HSSF, and virtually all
other applications that would use POIFS, deals with binary files,
which Streams handle correctly. Readers and Writers deal with text
and know how to handle 16-bit characters. If there is a demand for
providing support for Readers and Writers, let us know.</P>
<P>Here is some example code (excerpted and adapted from
net.sourceforge.poi.hssf.usermodel.Workbook class):</P>
<PRE> byte[] bytes = getBytes(); // get the bytes for the document (elsewhere in the class)
FileOutputStream stream = new FileOutputStream(&quot;/home/reportsys/test/text.xls&quot;); // create a new FileOuputStream
Filesystem fs = new Filesystem(); // create a new POIFS Filesystem object
fs.createDocument(new ByteArrayInputStream(bytes), &quot;Workbook&quot;); // create a new document in the root directory of the POIFS filesystem
// close on ByteArrayInputStream is a no-op so we don't bother, no real file handle is used
fs.writeFilesystem(stream); // write the filesystem to the output stream.
Stream.close(); // close our stream (don't leak file handles its bad news)</PRE><H4>
Reading or modifying an existing file</H4>
<P>Reading in an exising POIFS file system is equally simple. Create
a new instance of <CODE>net.sourceforge.poi.poifs.filesystem.Filesystem</CODE>
by calling the <CODE>Filesystem(java.io.InputStream)</CODE>
constructor and passing in your file system's data (this would
probably be a <CODE>FileInputStream</CODE> , but it doesn't matter).
From there you can get documents from the root directory by calling
<CODE>Filesystem.createDocumentInputStream(name)</CODE> and passing a
string representing that document's name.</P>
<P>If you wish to walk the filesystem, the easiest thing to do is
<CODE>DirectoryEntry.getEntries()</CODE>. This will give you a
<CODE>java.util.Iterator</CODE> of <CODE>Entry</CODE> instances
(<CODE>DirectoryEntry </CODE>and <CODE>DocumentEntry</CODE> are
extensions of <CODE>Entry</CODE>) contained by the <CODE>DirectoryEntry</CODE>
. For instance you could call <CODE>Filesystem.getRoot()</CODE> to
retrieve a <CODE>DirectoryEntry</CODE> instance. From there you could
call <CODE>DirectoryEntry.getEntries()</CODE> and retrieve an
<CODE>Iterator</CODE> of those entries. Iterating through these
entries, you'd call <CODE>getName()</CODE> to check the name of the
entry and <CODE>isDocumentEntry()</CODE> or <CODE>isDirectoryEntry()</CODE>
to determine its type. Going the other way, given an <CODE>Entry</CODE>,
you can walk back up the directory chain by calling <CODE>getParent()</CODE>,
which returns the <CODE>Entry</CODE>'s containing <CODE>DirectoryEntry</CODE>
(calling <CODE>getParent()</CODE> on the root directory returns a
<SAMP>null</SAMP> reference).</P>
<P>With a <CODE>DocumentEntry</CODE>, you can create an instance of
<CODE>net.sourceforge.poi.poifs.filesystem.DocumentInputStream</CODE>
, by passing the <CODE>DocumentEntry</CODE> as the only argument to
the constructor of <CODE>DocumentInputStream.</CODE>The
<CODE>DocumentInputStream</CODE> class is a simple extension of
<CODE>java.io.InputStream</CODE> that fully supports the <CODE>InputStream</CODE>
API, including the <CODE>mark</CODE> , <CODE>reset</CODE>, and <CODE>skip</CODE>
methods, providing a form of random access I/O.</P>
<P>To modify the file you would simply walk through the entries and
follow the same instructions for writing a POIFS file system from
scratch. There are also methods to delete an <CODE>Entry</CODE>
(note: you cannot delete the root directory, nor can you delete a
<CODE>DirectoryEntry</CODE> unless it's empty) and to rename an <CODE>Entry</CODE>
(but see the <A HREF="#Names">notes</A> above).
</P>
<H3>POIFS Logging facility</H3>
<P>POIFS does not yet use log4j style logging.</P>
<P>Here is an example
</P>
<PRE STYLE="margin-bottom: 0.2in">Paste log config example</PRE><H3>
POIFS Developer's Tools</H3>
<P>POIFS does not yet have developer's tools.
</P>
<H3>What's Next?</H3>
<OL>
<LI><P STYLE="margin-bottom: 0in">Refactoring of the API to more
cleanly separate write from read.
</P>
<LI><P STYLE="margin-bottom: 0in">Add logging/tracing code
</P>
<LI><P STYLE="margin-bottom: 0in">Add tree viewer (probably Andy)
</P>
<LI><P>Read/write support for creation and modification time stamps
</P>
</OL>
<P><BR><BR>
</P>
</BODY>
</HTML>