2
1
mirror of https://github.com/qpdf/qpdf.git synced 2024-10-31 19:02:30 +00:00
qpdf/manual/qpdf-manual.xml
Jay Berkenbilt 701e6db7b0 2.0.6
git-svn-id: svn+q:///qpdf/trunk@673 71b93d88-0707-0410-a8cf-f5a4172ac649
2009-05-03 16:38:04 +00:00

2036 lines
79 KiB
XML

<?xml version="1.0" encoding="utf-8"?>
<!DOCTYPE book [
<!ENTITY ldquo "&#x201C;">
<!ENTITY rdquo "&#x201D;">
<!ENTITY mdash "&#x2014;">
<!ENTITY ndash "&#x2013;">
<!ENTITY nbsp "&#xA0;">
<!ENTITY swversion "2.0.6">
<!ENTITY lastreleased "May 3, 2009">
]>
<book>
<bookinfo>
<title>QPDF Manual</title>
<subtitle>For QPDF Version &swversion;, &lastreleased;</subtitle>
<author>
<firstname>Jay</firstname><surname>Berkenbilt</surname>
</author>
<copyright>
<year>2005&ndash;2009</year>
<holder>Jay Berkenbilt</holder>
</copyright>
</bookinfo>
<preface id="acknowledgments">
<title>General Information</title>
<para>
QPDF is a program that does structural, content-preserving
transformations on PDF files. QPDF's website is located at <ulink
url="http://qpdf.qbilt.org/">http://qpdf.qbilt.org/</ulink>.
</para>
<para>
QPDF has been released under the terms of <ulink
url="http://www.opensource.org/licenses/artistic-license-2.0.php">Version
2.0 of the Artistic License</ulink>, a copy of which appears in the
file <filename>Artistic-2.0</filename> in the source distribution.
</para>
<para>
QPDF was originally created in 2001 and modified periodically
between 2001 and 2005 during my employment at <ulink
url="http://www.apexcovantage.com">Apex CoVantage</ulink>. Upon my
departure from Apex, the company graciously allowed me to take
ownership of the software and continue maintaining as an open
source project, a decision for which I am very grateful. I have
made considerable enhancements to it since that time. I feel
fortunate to have worked for people who would make such a decision.
This work would not have been possible without their support.
</para>
</preface>
<chapter id="ref.overview">
<title>What is QPDF?</title>
<para>
QPDF is a program that does structural, content-preserving
transformations on PDF files. It could have been called something
like <emphasis>pdf-to-pdf</emphasis>. It also provides many useful
capabilities to developers of PDF-producing software or for people
who just want to look at the innards of a PDF file to learn more
about how they work.
</para>
<para>
QPDF is <emphasis>not</emphasis> a PDF content creation library, a
PDF viewer, or a program capable of converting PDF into other
formats. In particular, QPDF knows nothing about the semantics of
PDF content streams. If you are looking for something that can do
that, you should look elsewhere. However, once you have a valid
PDF file, QPDF can be used to transform that file in ways perhaps
your original PDF creation can't handle. For example, programs
generate simple PDF files but can't password-protect them,
web-optimize them, or perform other transformations of that type.
</para>
</chapter>
<chapter id="ref.installing">
<title>Building and Installing QPDF</title>
<para>
This chapter describes how to build and install qpdf. Please see
also the <filename>README</filename> and
<filename>INSTALL</filename> files in the source distribution.
</para>
<sect1 id="ref.prerequisites">
<title>System Requirements</title>
<para>
The qpdf package has relatively few external dependencies. In
order to build qpdf, the following packages are required:
<itemizedlist>
<listitem>
<para>
zlib: <ulink url="http://www.zlib.net/">http://www.zlib.net/</ulink>
</para>
</listitem>
<listitem>
<para>
pcre: <ulink url="http://www.pcre.org/">http://www.pcre.org/</ulink>
</para>
</listitem>
<listitem>
<para>
gnu make 3.81 or newer: <ulink url="http://www.gnu.org/software/make">http://www.gnu.org/software/make</ulink>
</para>
</listitem>
<listitem>
<para>
perl version 5.8 or newer:
<ulink url="http://www.perl.org/">http://www.perl.org/</ulink>;
required for <command>fix-qdf</command> and the test suite.
</para>
</listitem>
<listitem>
<para>
GNU diffutils (any version): <ulink
url="http://www.gnu.org/software/diffutils/">http://www.gnu.org/software/diffutils/</ulink>
is required to run the test suite. Note that this is the
version of diff present on virtually all GNU/Linux systems.
This is required because the test suite uses <command>diff
-u</command>.
</para>
</listitem>
</itemizedlist>
</para>
<para>
Part of qpdf's test suite does comparisons of the contents PDF
files by converting them images and comparing the images. You can
optionally disable this part of the test suite by running
<command>configure</command> with the
<option>--disable-test-compare-images</option> flag. If you leave
this enabled, the following additional requirements are required
by the test suite. Note that in no case are these items required
to use qpdf.
<itemizedlist>
<listitem>
<para>
libtiff: <ulink url="http://www.remotesensing.org/libtiff/">http://www.remotesensing.org/libtiff/</ulink>
</para>
</listitem>
<listitem>
<para>
GhostScript version 8.60 or newer: <ulink
url="http://pages.cs.wisc.edu/~ghost/">http://pages.cs.wisc.edu/~ghost/</ulink>
</para>
</listitem>
</itemizedlist>
This option is primarily intended for use by packagers of qpdf so
that they can avoid having the qpdf packages depend on tiff and
ghostscript software.
</para>
<para>
If Adobe Reader is installed as <command>acroread</command>, some
additional test cases will be enabled. These test cases simply
verify that Adobe Reader can open the files that qpdf creates.
They require version 8.0 or newer to pass. However, in order to
avoid having qpdf depend on non-free (as in liberty) software, the
test suite will still pass without Adobe reader, and the test
suite still exercises the full functionality of the software.
</para>
<para>
Pre-built documentation is distributed with qpdf, so you should
generally not need to rebuild the documentation. In order to
build the documentation from its docbook sources, you need the
docbook XML style sheets (<ulink
url="http://downloads.sourceforge.net/docbook/">http://downloads.sourceforge.net/docbook/</ulink>).
To build the PDF version of the documentation, you need Apache fop
(<ulink
url="http://xml.apache.org/fop/">http://xml.apache.org/fop/</ulink>)
version 0.94 of higher.
</para>
</sect1>
<sect1 id="ref.building">
<title>Build Instructions</title>
<para>
Building qpdf on UNIX is generally just a matter of running
<programlisting>./configure
make
</programlisting>
You can also run <command>make check</command> to run the test
suite and <command>make install</command> to install. Please run
<command>./configure --help</command> for options on what can be
configured. You can also set the value of
<varname>DESTDIR</varname> during installation to install to a
temporary location, as is common with many open source packages.
Please see also the <filename>README</filename> and
<filename>INSTALL</filename> files in the source distribution.
</para>
<para>
There is currently no support for building qpdf on Windows. The
code is reasonably portable, however, and making it work on
Windows would probably be reasonably straightforward. A
significant amount of the code in QPDF has been known to work on
Windows in the past.
</para>
<para>
There are some other things you can do with the build. Although
qpdf uses <application>autoconf</application>, it does not use
<application>automake</application> but instead uses a
hand-crafted non-recursive Makefile that requires gnu make. If
you're really interested, please read the comments in the
top-level <filename>Makefile</filename>.
</para>
</sect1>
</chapter>
<chapter id="ref.using">
<title>Running QPDF</title>
<para>
This chapter describes how to run the qpdf program from the command
line.
</para>
<sect1 id="ref.invocation">
<title>Basic Invocation</title>
<para>
When running qpdf, the basic invocation is as follows:
<programlisting><command>qpdf</command><option> [ <replaceable>options</replaceable> ] <replaceable>infilename</replaceable> [ <replaceable>outfilename</replaceable> ]</option>
</programlisting>
This converts PDF file <option>infilename</option> to PDF file
<option>outfilename</option>. The output file is functionally
identical to the input file but may have been structurally
reorganized. Also, orphaned objects will be removed from the
file. Many transformations are available as controlled by the
options below.
</para>
<para>
<option>outfilename</option> does not have to be seekable, even
when generating linearized files. Specifying
&ldquo;<option>-</option>&rdquo; as <option>outfilename</option>
means to write to standard output.
</para>
<para>
Most options require an output file, but some testing or
inspection commands do not. These are specifically noted.
</para>
</sect1>
<sect1 id="ref.basic-options">
<title>Basic Options</title>
<para>
The following options are the most common ones and perform
commonly needed transformations.
<variablelist>
<varlistentry>
<term><option>--password=password</option></term>
<listitem>
<para>
Specifies a password for accessing encrypted files.
</para>
</listitem>
</varlistentry>
<varlistentry>
<term><option>--linearize</option></term>
<listitem>
<para>
Causes generation of a linearized (web optimized) output file.
</para>
</listitem>
</varlistentry>
<varlistentry>
<term><option>--encrypt options --</option></term>
<listitem>
<para>
Causes generation an encrypted output file. Please see <xref
linkend="ref.encryption-options"/> for details on how to
specify encryption parameters.
</para>
</listitem>
</varlistentry>
<varlistentry>
<term><option>--decrypt</option></term>
<listitem>
<para>
Removes any encryption on the file. A password must be
supplied if the file is password protected.
</para>
</listitem>
</varlistentry>
</variablelist>
</para>
<para>
Password-protected files may be opened by specifying a password.
By default, qpdf will preserve any encryption data associated with
a file. If <option>--decrypt</option> is specified, qpdf will
attempt to remove any encryption information. If
<option>--encrypt</option> is specified, qpdf will replace the
document's encryption parameters with whatever is specified.
</para>
<para>
Note that qpdf does not obey encryption restrictions already
imposed on the file. Doing so would be meaningless since qpdf can
be used to remove encryption from the file entirely. This
functionality is not intended to be used for bypassing copyright
restrictions or other restrictions placed on files by their
producers.
</para>
</sect1>
<sect1 id="ref.encryption-options">
<title>Encryption Options</title>
<para>
To change the encryption parameters of a file, use the --encrypt
flag. The syntax is
<programlisting><option>--encrypt <replaceable>user-password</replaceable> <replaceable>owner-password</replaceable> <replaceable>key-length</replaceable> [ <replaceable>restrictions</replaceable> ] --</option>
</programlisting>
Note that &ldquo;<option>--</option>&rdquo; terminates parsing of
encryption flags and must be present even if no restrictions are
present.
</para>
<para>
Either or both of the user password and the owner password may be
empty strings.
</para>
<para>
The value for
<option><replaceable>key-length</replaceable></option> may be 40
or 128. The restriction flags are dependent upon key length.
When no additional restrictions are given, the default is to be
fully permissive.
</para>
<para>
If <option><replaceable>key-length</replaceable></option> is 40,
the following restriction options are available:
<variablelist>
<varlistentry>
<term><option>--print=[yn]</option></term>
<listitem>
<para>
Determines whether or not to allow printing.
</para>
</listitem>
</varlistentry>
<varlistentry>
<term><option>--modify=[yn]</option></term>
<listitem>
<para>
Determines whether or not to allow document modification.
</para>
</listitem>
</varlistentry>
<varlistentry>
<term><option>--extract=[yn]</option></term>
<listitem>
<para>
Determines whether or not to allow text/image extraction.
</para>
</listitem>
</varlistentry>
<varlistentry>
<term><option>--annotate=[yn]</option></term>
<listitem>
<para>
Determines whether or not to allow comments and form fill-in
and signing.
</para>
</listitem>
</varlistentry>
</variablelist>
If <option><replaceable>key-length</replaceable></option> is 128,
the following restriction options are available:
<variablelist>
<varlistentry>
<term><option>--accessibility=[yn]</option></term>
<listitem>
<para>
Determines whether or not to allow accessibility to visually
impaired.
</para>
</listitem>
</varlistentry>
<varlistentry>
<term><option>--extract=[yn]</option></term>
<listitem>
<para>
Determines whether or not to allow text/graphic extraction.
</para>
</listitem>
</varlistentry>
<varlistentry>
<term><option>--print=<replaceable>print-opt</replaceable></option></term>
<listitem>
<para>
Controls printing access.
<option><replaceable>print-opt</replaceable></option> may be
one of the following:
<itemizedlist>
<listitem>
<para>
<option>full</option>: allow full printing
</para>
</listitem>
<listitem>
<para>
<option>low</option>: allow low-resolution printing only
</para>
</listitem>
<listitem>
<para>
<option>none</option>: disallow printing
</para>
</listitem>
</itemizedlist>
</para>
</listitem>
</varlistentry>
<varlistentry>
<term><option>--modify=<replaceable>modify-opt</replaceable></option></term>
<listitem>
<para>
Controls modify access.
<option><replaceable>modify-opt</replaceable></option> may be
one of the following:
<itemizedlist>
<listitem>
<para>
<option>all</option>: allow full document modification
</para>
</listitem>
<listitem>
<para>
<option>annotate</option>: allow comment authoring and form operations
</para>
</listitem>
<listitem>
<para>
<option>form</option>: allow form field fill-in and signing
</para>
</listitem>
<listitem>
<para>
<option>assembly</option>: allow document assembly only
</para>
</listitem>
<listitem>
<para>
<option>none</option>: allow no modifications
</para>
</listitem>
</itemizedlist>
</para>
</listitem>
</varlistentry>
</variablelist>
The default for each permission option is to be fully permissive.
</para>
</sect1>
<sect1 id="ref.advanced-transformation">
<title>Advanced Transformation Options</title>
<para>
These transformation options control fine points of how qpdf
creates the output file. Mostly these are of use only to people
who are very familiar with the PDF file format or who are PDF
developers. The following options are available:
<variablelist>
<varlistentry>
<term><option>--stream-data=<replaceable>option</replaceable></option></term>
<listitem>
<para>
Controls transformation of stream data. The value of
<option><replaceable>option</replaceable></option> may be one
of the following:
<itemizedlist>
<listitem>
<para>
<option>compress</option>: recompress stream data when
possible (default)
</para>
</listitem>
<listitem>
<para>
<option>preserve</option>: leave all stream data as is
</para>
</listitem>
<listitem>
<para>
<option>uncompress</option>: uncompress stream data when
possible
</para>
</listitem>
</itemizedlist>
</para>
</listitem>
</varlistentry>
<varlistentry>
<term><option>--normalize-content=[yn]</option></term>
<listitem>
<para>
Enables or disables normalization of content streams.
</para>
</listitem>
</varlistentry>
<varlistentry>
<term><option>--suppress-recovery</option></term>
<listitem>
<para>
Prevents qpdf from attempting to recover damaged files.
</para>
</listitem>
</varlistentry>
<varlistentry>
<term><option>--object-streams=<replaceable>mode</replaceable></option></term>
<listitem>
<para>
Controls handing of object streams. The value of
<option><replaceable>mode</replaceable></option> may be one of
the following:
<itemizedlist>
<listitem>
<para>
<option>preserve</option>: preserve original object streams
(default)
</para>
</listitem>
<listitem>
<para>
<option>disable</option>: don't write any object streams
</para>
</listitem>
<listitem>
<para>
<option>generate</option>: use object streams wherever
possible
</para>
</listitem>
</itemizedlist>
</para>
</listitem>
</varlistentry>
<varlistentry>
<term><option>--ignore-xref-streams</option></term>
<listitem>
<para>
Tells qpdf to ignore any cross-reference streams.
</para>
</listitem>
</varlistentry>
<varlistentry>
<term><option>--qdf</option></term>
<listitem>
<para>
Turns on QDF mode. For additional information on QDF, please
see <xref linkend="ref.qdf"/>.
</para>
</listitem>
</varlistentry>
</variablelist>
</para>
<para>
By default, when a stream is encoded using non-lossy filters that
qpdf understands and is not already compressed using a good
compression scheme, qpdf will uncompress and recompress streams.
Assuming proper filter implements, this is safe and generally
results in smaller files. This behavior may also be explicitly
requested with <option>--stream-data=compress</option>.
</para>
<para>
When <option>--stream-data=preserve</option> is specified, qpdf
will never attempt to change the filtering of any stream data.
</para>
<para>
When <option>--stream-data=uncompress</option> is specified, qpdf
will attempt to remove any non-lossy filters that it supports.
This includes <literal>/FlateDecode</literal>,
<literal>/LZWDecode</literal>, <literal>/ASCII85Decode</literal>,
and <literal>/ASCIIHexDecode</literal>. This can be very useful
for inspecting the contents of various streams.
</para>
<para>
When <option>--normalize-content=y</option> is specified, qpdf
will attempt to normalize whitespace and newlines in page content
streams. This is generally safe but could, in some cases, cause
damage to the content streams. This option is intended for people
who wish to study PDF content streams or to debug PDF content.
You should not use this for &ldquo;production&rdquo; PDF files.
</para>
<para>
Ordinarily, qpdf will attempt to recover from certain types of
errors in PDF files. These include errors in the cross-reference
table, certain types of object numbering errors, and certain types
of stream length errors. Sometimes, qpdf may think it has
recovered but may not have actually recovered, so care should be
taken when using this option as some data loss is possible. The
<option>--suppress-recovery</option> option will prevent qpdf from
attempting recovery. In this case, it will fail on the first
error that it encounters.
</para>
<para>
Object streams, also known as compressed objects, were introduced
into the PDF specification at version 1.5, corresponding to
Acrobat 6. Some older PDF viewers may not support files with
object streams. qpdf can be used to transform files with object
streams to files without object streams or vice versa. As
mentioned above, there are three object stream modes:
<option>preserve</option>, <option>disable</option>, and
<option>generate</option>.
</para>
<para>
In <option>preserve</option> mode, the relationship to objects and
the streams that contain them is preserved from the original file.
In <option>disable</option> mode, all objects are written as
regular, uncompressed objects. The resulting file should be
readable by older PDF viewers. (Of course, the content of the
files may include features not supported by older viewers, but at
least the structure will be supported.) In
<option>generate</option> mode, qpdf will create its own object
streams. This will usually result in more compact PDF files,
though they may not be readable by older viewers. In this mode,
qpdf will also make sure the PDF version number in the header is
at least 1.5.
</para>
<para>
Ordinarily, qpdf reads cross-reference streams when they are
present in a PDF file. If <option>--ignore-xref-streams</option>
is specified, qpdf will ignore any cross-reference streams for
hybrid PDF files. The purpose of hybrid files is to make some
content available to viewers that are not aware of cross-reference
streams. It is almost never desirable to ignore them. The only
time when you might want to use this feature is if you are testing
creation of hybrid PDF files and wish to see how a PDF consumer
that doesn't understand object and cross-reference streams would
interpret such a file.
</para>
<para>
The <option>--qdf</option> flag turns on QDF mode, which changes
some of the defaults described above. Specifically, in QDF mode,
by default, stream data is uncompressed, content streams are
normalized, and encryption is removed. These defaults can still
be overridden by specifying the appropriate options as described
above. Additionally, in QDF mode, stream lengths are stored as
indirect objects, objects are laid out in a less efficient but
more readable fashion, and the documents are interspersed with
comments that make it easier for the user to find things and also
make it possible for <command>fix-qdf</command> to work properly.
QDF mode is intended for people, mostly developers, who wish to
inspect or modify PDF files in a text editor. For details, please
see <xref linkend="ref.qdf"/>.
</para>
</sect1>
<sect1 id="ref.testing-options">
<title>Testing, Inspection, and Debugging Options</title>
<para>
These options can be useful for digging into PDF files or for use
in automated test suites for software that uses the qpdf library.
When any of the options in this section are specified, no output
file should be given. The following options are available:
<variablelist>
<varlistentry>
<term><option>--static-id</option></term>
<listitem>
<para>
Causes generation of a fixed value for /ID. This is intended
for testing only. Never use it for production files.
</para>
</listitem>
</varlistentry>
<varlistentry>
<term><option>--no-original-object-ids</option></term>
<listitem>
<para>
Suppresses inclusion of original object ID comments in QDF
files. This can be useful when generating QDF files for test
purposes, particularly when comparing them to determine
whether two PDF files have identical content.
</para>
</listitem>
</varlistentry>
<varlistentry>
<term><option>-show-encryption</option></term>
<listitem>
<para>
Shows document encryption parameters. Also shows the
document's user password if the owner password is given.
</para>
</listitem>
</varlistentry>
<varlistentry>
<term><option>-check-linearization</option></term>
<listitem>
<para>
Checks file integrity and linearization status.
</para>
</listitem>
</varlistentry>
<varlistentry>
<term><option>-show-linearization</option></term>
<listitem>
<para>
Checks and displays all data in the linearization hint tables.
</para>
</listitem>
</varlistentry>
<varlistentry>
<term><option>-show-xref</option></term>
<listitem>
<para>
Shows the contents of the cross-reference table in a
human-readable form. This is especially useful for files with
cross-reference streams which are stored in a binary format.
</para>
</listitem>
</varlistentry>
<varlistentry>
<term><option>-show-object=obj[,gen]</option></term>
<listitem>
<para>
Show the contents of the given object. This is especially
useful for inspecting objects that are inside of object
streams (also known as &ldquo;compressed objects&rdquo;).
</para>
</listitem>
</varlistentry>
<varlistentry>
<term><option>-raw-stream-data</option></term>
<listitem>
<para>
When used along with the <option>--show-object</option>
option, if the object is a stream, shows the raw stream data
instead of object's contents.
</para>
</listitem>
</varlistentry>
<varlistentry>
<term><option>-filtered-stream-data</option></term>
<listitem>
<para>
When used along with the <option>--show-object</option>
option, if the object is a stream, shows the filtered stream
data instead of object's contents. If the stream is filtered
using filters that qpdf does not support, an error will be
issued.
</para>
</listitem>
</varlistentry>
<varlistentry>
<term><option>-show-pages</option></term>
<listitem>
<para>
Shows the object and generation number for each page
dictionary object and for each content stream associated with
the page. Having this information makes it more convenient to
inspect objects from a particular page.
</para>
</listitem>
</varlistentry>
<varlistentry>
<term><option>-with-images</option></term>
<listitem>
<para>
When used along with <option>--show-pages</option>, also shows
the object and generation numbers for the image objects on
each page. (At present, information about images in shared
resource dictionaries are not output by this command. This is
discussed in a comment in the source code.)
</para>
</listitem>
</varlistentry>
<varlistentry>
<term><option>-check</option></term>
<listitem>
<para>
Checks file structure and well as encryption, linearization,
and encoding of stream data. A file for which
<option>--check</option> reports no errors may still have
errors in stream data content but should otherwise be
structurally sound. If <option>--check</option> any errors,
qpdf will exit with a status of 2. There are some recoverable
conditions that <option>--check</option> detects. These are
issued as warnings instead of errors. If qpdf finds no errors
but finds warnings, it will exit with a status of 3 (as of
version&nbsp;2.0.4).
</para>
</listitem>
</varlistentry>
</variablelist>
</para>
<para>
The <option>--raw-stream-data</option> and
<option>--filtered-stream-data</option> options are ignored unless
<option>--show-object</option> is given. Either of these options
will cause the stream data to be written to standard output. In
order to avoid commingling of stream data with other output, it is
recommend that these objects not be combined with other
test/inspection options.
</para>
<para>
If <option>--filtered-stream-data</option> is given and
<option>--normalize-content=y</option> is also given, qpdf will
attempt to normalize the stream data as if it is a page content
stream. This attempt will be made even if it is not a page
content stream, in which case it will produce unusuable results.
</para>
</sect1>
</chapter>
<chapter id="ref.qdf">
<title>QDF Mode</title>
<para>
In QDF mode, qpdf creates PDF files in what we call <firstterm>QDF
form</firstterm>. A PDF file in QDF form, sometimes called a QDF
file, is a completely valid PDF file that has
<literal>%QDF-1.0</literal> as its third line (after the pdf header
and binary characters) and has certain other characteristics. The
purpose of QDF form is to make it possible to edit PDF files, with
some restrictions, in an ordinary text editor. This can be very
useful for experimenting with different PDF constructs or for
making one-off edits to PDF files (though there are other reasons
why this may not always work).
</para>
<para>
It is ordinarily very difficult to edit PDF files in a text editor
for two reasons: most meaningful data in PDF files is compressed,
and PDF files are full of offset and length information that makes
it hard to add or remove data. A QDF file is organized in a manner
such that, if edits are kept within certain constraints, the
<command>fix-qdf</command> program, distributed with qpdf, is able
to restore edited files to a correct state. The
<command>fix-qdf</command> program takes no command-line
arguments. It reads a possibly edited QDF file from standard input
and writes a repaired file to standard output.
</para>
<para>
The following attributes characterize a QDF file:
<itemizedlist>
<listitem>
<para>
All objects appear in numerical order in the PDF file, including
when objects appear in object streams.
</para>
</listitem>
<listitem>
<para>
Objects are printed in an easy-to-read format, and all line
endings are normalized to UNIX line endings.
</para>
</listitem>
<listitem>
<para>
Unless specifically overridden, streams appear uncompressed
(when qpdf supports the filters and they are compressed with a
non-lossy compression scheme), and most content streams are
normalized (line endings are converted to just a UNIX-style
linefeeds).
</para>
</listitem>
<listitem>
<para>
All streams lengths are represented as indirect objects, and the
stream length object is always the next object after the stream.
If the stream data does not end with a newline, an extra newline
is inserted, and a special comment appears after the stream
indicating that this has been done.
</para>
</listitem>
<listitem>
<para>
If the PDF file contains object streams, if object stream
<emphasis>n</emphasis> contains <emphasis>k</emphasis> objects,
those objects are numbered from <emphasis>n+1</emphasis> through
<emphasis>n+k</emphasis>, and the object number/offset pairs
appear on a separate line for each object. Additionally, each
object in the object stream is preceded by a comment indicating
its object number and index. This makes it very easy to find
objects in object streams.
</para>
</listitem>
<listitem>
<para>
All beginnings of objects, <literal>stream</literal> tokens,
<literal>endstream</literal> tokens, and
<literal>endobj</literal> tokens appear on lines by themselves.
A blank line follows every <literal>endobj</literal> token.
</para>
</listitem>
<listitem>
<para>
If there is a cross-reference stream, it is unfiltered.
</para>
</listitem>
<listitem>
<para>
Page dictionaries and page content streams are marked with
special comments that make them easy to find.
</para>
</listitem>
<listitem>
<para>
Comments precede each object indicating the object number of the
corresponding object in the original file.
</para>
</listitem>
</itemizedlist>
</para>
<para>
When editing a QDF file, any edits can be made as long as the above
constraints are maintained. This means that you can freely edit a
page's content without worrying about messing up the QDF file. It
is also possible to add new objects so long as those objects are
added after the last object in the file or subsequent objects are
renumbered. If a QDF file has object streams in it, you can always
add the new objects before the xref stream and then change the
number of the xref stream, since nothing generally ever references
it by number.
</para>
<para>
It is not generally practical to remove objects from QDF files
without messing up object numbering, but if you remove all
references to an object, you can run qpdf on the file (after
running <command>fix-qdf</command>), and qpdf will omit the
now-orphaned object.
</para>
<para>
When <command>fix-qdf</command> is run, it goes through the file
and recomputes the following parts of the file:
<itemizedlist>
<listitem>
<para>
the <literal>/N</literal>, <literal>/W</literal>, and
<literal>/First</literal> keys of all object stream dictionaries
</para>
</listitem>
<listitem>
<para>
the pairs of numbers representing object numbers and offsets of
objects in object streams
</para>
</listitem>
<listitem>
<para>
all stream lengths
</para>
</listitem>
<listitem>
<para>
the cross-reference table or cross-reference stream
</para>
</listitem>
<listitem>
<para>
the offset to the cross-reference table or cross-reference
stream following the <literal>startxref</literal> token
</para>
</listitem>
</itemizedlist>
</para>
</chapter>
<chapter id="ref.using-library">
<title>Using the QPDF Library</title>
<para>
The source tree for the qpdf package has an
<filename>examples</filename> directory that contains a few
example programs. The <filename>qpdf/qpdf.cc</filename> source
file also serves as a useful example since it exercises almost all
of the qpdf library's public interface. The best source of
documentation on the library itself is reading comments in
<filename>include/qpdf/QPDF.hh</filename>,
<filename>include/qpdf/QDFWriter.hh</filename>, and
<filename>include/qpdf/QPDFObjectHandle.hh</filename>.
</para>
<para>
All header files are installed in the <filename>include/qpdf</filename> directory. It
is recommend that you use <literal>#include
&lt;qpdf/QPDF.hh&gt;</literal> rather than adding
<filename>include/qpdf</filename> to your include path.
</para>
<para>
When linking against the qpdf library, you may also need to
specify <literal>-lpcre -lz</literal> on your link command. If
your system understands how to read libtool
<filename>.la</filename> files, this may not be necessary.
</para>
</chapter>
<chapter id="ref.design">
<title>Design and Library Notes</title>
<sect1 id="ref.design.intro">
<title>Introduction</title>
<para>
This section was written prior to the implementation of the qpdf
package and was subsequently modified to reflect the
implementation. In some cases, for purposes of explanation, it
may differ slightly from the actual implementation. As always,
the source code and test suite are authoritative. Even if there
are some errors, this document should serve as a road map to
understanding how this code works.
</para>
<para>
In general, one should adhere strictly to a specification when
writing but be liberal in reading. This way, the product of our
software will be accepted by the widest range of other programs,
and we will accept the widest range of input files. This library
attempts to conform to that philosophy whenever possible but also
aims to provide strict checking for people who want to validate
PDF files. If you don't want to see warnings and are trying to
write something that is tolerant, you can call
<literal>setSuppressWarnings(true)</literal>. If you want to fail
on the first error, you can call
<literal>setAttemptRecovery(false)</literal>. The default
behavior is to generating warnings for recoverable problems. Note
that recovery will not always produce the desired results even if
it is able to get through the file. Unlike most other PDF files
that produce generic warnings such as &ldquo;This file is
damaged,&rdquo;, qpdf generally issues a detailed error message
that would be most useful to a PDF developer. This is by design
as there seems to be a shortage of PDF validation tools out
there. (This was, in fact, one of the major motivations behind
the initial creation of qpdf.)
</para>
</sect1>
<sect1 id="ref.design-goals">
<title>Design Goals</title>
<para>
The QPDF package includes support for reading and rewriting PDF
files. It aims to hide from the user details involving object
locations, modified (appended) PDF files, the
directness/indirectness of objects, and stream filters including
encryption. It does not aim to hide knowledge of the object
hierarchy or content stream contents. Put another way, a user of
the qpdf library is expected to have knowledge about how PDF files
work, but is not expected to have to keep track of bookkeeping
details such as file positions.
</para>
<para>
A user of the library never has to care whether an object is
direct or indirect. All access to objects deals with this
transparently. All memory management details are also handled by
the library.
</para>
<para>
The <classname>PointerHolder</classname> object is used internally
by the library to deal with memory management. This is basically
a smart pointer object very similar in spirit to the Boost
library's <classname>shared_ptr</classname> object, but predating
it by several years. This library also makes use of a technique
for giving fine-grained access to methods in one class to other
classes by using public subclasses with friends and only private
members that in turn call private methods of the containing class.
See <classname>QPDFObjectHandle::Factory</classname> as an
example.
</para>
<para>
The top-level qpdf class is <classname>QPDF</classname>. A
<classname>QPDF</classname> object represents a PDF file. The
library provides methods for both accessing and mutating PDF
files.
</para>
<para>
<classname>QPDFObject</classname> is the basic PDF Object class.
It is an abstract base class from which are derived classes for
each type of PDF object. Clients do not interact with Objects
directly but instead interact with
<classname>QPDFObjectHandle</classname>.
</para>
<para>
<classname>QPDFObjectHandle</classname> contains
<classname>PointerHolder&lt;QPDFObject&gt;</classname> and
includes accessor methods that are type-safe proxies to the
methods of the derived object classes as well as methods for
querying object types. They can be passed around by value,
copied, stored in containers, etc. with very low overhead.
Instances of <classname>QPDFObjectHandle</classname> always
contain a reference back to the <classname>QPDF</classname> object
from which they were created. A
<classname>QPDFObjectHandle</classname> may be direct or indirect.
If indirect, the <classname>QPDFObject</classname> the
<classname>PointerHolder</classname> initially points to is a null
pointer. In this case, the first attempt to access the underlying
<classname>QPDFObject</classname> will result in the
<classname>QPDFObject</classname> being resolved via a call to the
referenced <classname>QPDF</classname> instance. This makes it
essentially impossible to make coding errors in which certain
things will work for some PDF files and not for others based on
which objects are direct and which objects are indirect.
</para>
<para>
There is no public interface for creating instances of
QPDFObjectHandle. They can be created only inside the QPDF
library. This is generally done through a call to the private
method <function>QPDF::readObject</function> which uses
<classname>QPDFTokenizer</classname> to read an indirect object at
a given file position and return a
<classname>QPDFObjectHandle</classname> that encapsulates it.
There are also internal methods to create fabricated indirect
objects from existing direct objects or to change an indirect
object into a direct object, though these steps are not performed
except to support rewriting.
</para>
<para>
When the <classname>QPDF</classname> class creates a new object,
it dynamically allocates the appropriate type of
<classname>QPDFObject</classname> and immediately hands the
pointer to an instance of <classname>QPDFObjectHandle</classname>.
The parser reads a token from the current file position. If the
token is a not either a dictionary or array opener, an object is
immediately constructed from the single token and the parser
returns. Otherwise, the parser is invoked recursively in a
special mode in which it accumulates objects until it finds a
balancing closer. During this process, the
&ldquo;<literal>R</literal>&rdquo; keyword is recognized and an
indirect <classname>QPDFObjectHandle</classname> may be
constructed.
</para>
<para>
The <function>QPDF::resolve()</function> method, which is used to
resolve an indirect object, may be invoked from the
<classname>QPDFObjectHandle</classname> class. It first checks a
cache to see whether this object has already been read. If not,
it reads the object from the PDF file and caches it. It the
returns the resulting <classname>QPDFObjectHandle</classname>.
The calling object handle then replaces its
<classname>PointerHolder&lt;QDFObject&gt;</classname> with the one
from the newly returned <classname>QPDFObjectHandle</classname>.
In this way, only a single copy of any direct object need exist
and clients can access objects transparently without knowing
caring whether they are direct or indirect objects. Additionally,
no object is ever read from the file more than once. That means
that only the portions of the PDF file that are actually needed
are ever read from the input file, thus allowing the qpdf package
to take advantage of this important design goal of PDF files.
</para>
<para>
If the requested object is inside of an object stream, the object
stream itself is first read into memory. Then the tokenizer reads
objects from the memory stream based on the offset information
stored in the stream. Those individual objects are cached, after
which the temporary buffer holding the object stream contents are
discarded. In this way, the first time an object in an object
stream is requested, all objects in the stream are cached.
</para>
<para>
An instance of <classname>QPDF</classname> is constructed by using
the class's default constructor. If desired, the
<classname>QPDF</classname> object may be configured with various
methods that change its default behavior. Then the
<function>QPDF::processFile()</function> method is passed the name
of a PDF file, which permanently associates the file with that
QPDF object. A password may also be given for access to
password-protected files. QPDF does not enforce encryption
parameters and will treat user and owner passwords equivalently.
Either password may be used to access an encrypted file.
<footnote>
<para>
As pointed out earlier, the intention is not for qpdf to be used
to bypass security on files. but as any open source PDF consumer
may be easily modified to bypass basic PDF document security,
and qpdf offers may transformations that can do this as well,
there seems to be little point in the added complexity of
conditionally enforcing document security.
</para>
</footnote>
<classname>QPDF</classname> will allow recovery of a user password
given an owner password. The input PDF file must be seekable.
(Output files written by <classname>QPDFWriter</classname> need
not be seekable, even when creating linearized files.) During
construction, <classname>QPDF</classname> validates the PDF file's
header, and then reads the cross reference tables and trailer
dictionaries. The <classname>QPDF</classname> class keeps only
the first trailer dictionary though it does read all of them so it
can check the <literal>/Prev</literal> key.
<classname>QPDF</classname> class users may request the root
object and the trailer dictionary specifically. The cross
reference table is kept private. Objects may then be requested by
number of by walking the object tree.
</para>
<para>
When a PDF file has a cross-reference stream instead of a
cross-reference table and trailer, requesting the document's
trailer dictionary returns the stream dictionary from the
cross-reference stream instead.
</para>
<para>
There are some convenience routines for very common operations
such as walking the page tree and returning a vector of all page
objects. For full details, please see the header file
<filename>QPDF.hh</filename>.
</para>
<para>
The following example should clarify how
<classname>QPDF</classname> processes a simple file.
<itemizedlist>
<listitem>
<para>
Client constructs <classname>QPDF</classname>
<varname>pdf</varname> and calls
<function>pdf.processFile("a.pdf");</function>.
</para>
</listitem>
<listitem>
<para>
The <classname>QPDF</classname> class checks the beginning of
<filename>a.pdf</filename> for
<literal>%!PDF-1.[0-9]+</literal>. It then reads the cross
reference table mentioned at the end of the file, ensuring that
it is looking before the last <literal>%%EOF</literal>. After
getting to <literal>trailer</literal> keyword, it invokes the
parser.
</para>
</listitem>
<listitem>
<para>
The parser sees &ldquo;<literal>&lt;&lt;</literal>&rdquo;, so
it calls itself recursively in dictionary creation mode.
</para>
</listitem>
<listitem>
<para>
In dictionary creation mode, the parser keeps accumulating
objects until it encounters
&ldquo;<literal>&gt;&gt;</literal>&rdquo;. Each object that is
read is pushed onto a stack. If
&ldquo;<literal>R</literal>&rdquo; is read, the last two
objects on the stack are inspected. If they are integers, they
are popped off the stack and their values are used to construct
an indirect object handle which is then pushed onto the stack.
When &ldquo;<literal>&gt;&gt;</literal>&rdquo; is finally read,
the stack is converted into a
<classname>QPDF_Dictionary</classname> which is placed in a
<classname>QPDFObjectHandle</classname> and returned.
</para>
</listitem>
<listitem>
<para>
The resulting dictionary is saved as the trailer dictionary.
</para>
</listitem>
<listitem>
<para>
The <literal>/Prev</literal> key is searched. If present,
<classname>QPDF</classname> seeks to that point and repeats
except that the new trailer dictionary is not saved. If
<literal>/Prev</literal> is not present, the initial parsing
process is complete.
</para>
<para>
If there is an encryption dictionary, the document's encryption
parameters are initialized.
</para>
</listitem>
<listitem>
<para>
The client requests root object. The
<classname>QPDF</classname> class gets the value of root key
from trailer dictionary and returns it. It is an unresolved
indirect <classname>QPDFObjectHandle</classname>.
</para>
</listitem>
<listitem>
<para>
The client requests the <literal>/Pages</literal> key from root
<classname>QPDFObjectHandle</classname>. The
<classname>QPDFObjectHandle</classname> notices that it is
indirect so it asks <classname>QPDF</classname> to resolve it.
<classname>QPDF</classname> looks in the object cache for an
object with the root dictionary's object ID and generation
number. Upon not seeing it, it checks the cross reference
table, gets the offset, and reads the object present at that
offset. It stores the result in the object cache and returns
the cached result. The calling
<classname>QPDFObjectHandle</classname> replaces its object
pointer with the one from the resolved
<classname>QPDFObjectHandle</classname>, verifies that it a
valid dictionary object, and returns the (unresolved indirect)
<classname>QPDFObject</classname> handle to the top of the
Pages hierarchy.
</para>
<para>
As the client continues to request objects, the same process is
followed for each new requested object.
</para>
</listitem>
</itemizedlist>
</para>
</sect1>
<sect1 id="ref.encryption">
<title>Encryption</title>
<para>
Encryption is supported transparently by qpdf. When opening a PDF
file, if an encryption dictionary exists, the
<classname>QPDF</classname> object processes this dictionary using
the password (if any) provided. The primary decryption key is
computed and cached. No further access is made to the encryption
dictionary after that time. When an object is read from a file,
the object ID and generation of the object in which it is
contained is always known. Using this information along with the
stored encryption key, all stream and string objects are
transparently decrypted. Raw encrypted objects are never stored
in memory. This way, nothing in the library ever has to know or
care whether it is reading an encrypted file.
</para>
<para>
An interface is also provided for writing encrypted streams and
strings given an encryption key. This is used by
<classname>QPDFWriter</classname> when it rewrites encrypted
files.
</para>
</sect1>
<sect1 id="ref.rewriting">
<title>Writing PDF Files</title>
<para>
The qpdf library supports file writing of
<classname>QPDF</classname> objects to PDF files through the
<classname>QPDFWriter</classname> class. The
<classname>QPDFWriter</classname> class has two writing modes: one
for non-linearized files, and one for linearized files. See <xref
linkend="ref.linearization"/> for a description of linearization
is implemented. This section describes how we write
non-linearized files including the creation of QDF files (see
<xref linkend="ref.qdf"/>.
</para>
<para>
This outline was written prior to implementation and is not
exactly accurate, but it provides a correct &ldquo;notional&rdquo;
idea of how writing works. Look at the code in
<classname>QPDFWriter</classname> for exact details.
<itemizedlist>
<listitem>
<para>
Initialize state:
<itemizedlist>
<listitem>
<para>
next object number = 1
</para>
</listitem>
<listitem>
<para>
object queue = empty
</para>
</listitem>
<listitem>
<para>
renumber table: old object id/generation to new id/0 = empty
</para>
</listitem>
<listitem>
<para>
xref table: new id -> offset = empty
</para>
</listitem>
</itemizedlist>
</para>
</listitem>
<listitem>
<para>
Create a QPDF object from a file.
</para>
</listitem>
<listitem>
<para>
Write header for new PDF file.
</para>
</listitem>
<listitem>
<para>
Request the trailer dictionary.
</para>
</listitem>
<listitem>
<para>
For each value that is an indirect object, grab the next object
number (via an operation that returns and increments the
number). Map object to new number in renumber table. Push
object onto queue.
</para>
</listitem>
<listitem>
<para>
While there are more objects on the queue:
<itemizedlist>
<listitem>
<para>
Pop queue.
</para>
</listitem>
<listitem>
<para>
Look up object's new number <emphasis>n</emphasis> in the
renumbering table.
</para>
</listitem>
<listitem>
<para>
Store current offset into xref table.
</para>
</listitem>
<listitem>
<para>
Write <literal><replaceable>n</replaceable> 0 obj</literal>.
</para>
</listitem>
<listitem>
<para>
If object is null, whether direct or indirect, write out
null, thus eliminating unresolvable indirect object
references.
</para>
</listitem>
<listitem>
<para>
If the object is a stream stream, write stream contents,
piped through any filters as required, to a memory buffer.
Use this buffer to determine the stream length.
</para>
</listitem>
<listitem>
<para>
If object is not a stream, array, or dictionary, write out
its contents.
</para>
</listitem>
<listitem>
<para>
If object is an array or dictionary (including stream),
traverse its elements (for array) or values (for
dictionaries), handling recursive dictionaries and arrays,
looking for indirect objects. When an indirect object is
found, if it is not resolvable, ignore. (This case is
handled when writing it out.) Otherwise, look it up in the
renumbering table. If not found, grab the next available
object number, assign to the referenced object in the
renumbering table, and push the referenced object onto the
queue. As a special case, when writing out a stream
dictionary, replace length, filters, and decode parameters
as required.
</para>
<para>
Write out dictionary or array, replacing any unresolvable
indirect object references with null (pdf spec says
reference to non-existent object is legal and resolves to
null) and any resolvable ones with references to the
renumbered objects.
</para>
</listitem>
<listitem>
<para>
If the object is a stream, write
<literal>stream\n</literal>, the stream contents (from the
memory buffer), and <literal>\nendstream\n</literal>.
</para>
</listitem>
<listitem>
<para>
When done, write <literal>endobj</literal>.
</para>
</listitem>
</itemizedlist>
</para>
</listitem>
</itemizedlist>
</para>
<para>
Once we have finished the queue, all referenced objects will have
been written out and all deleted objects or unreferenced objects
will have been skipped. The new cross-reference table will
contain an offset for every new object number from 1 up to the
number of objects written. This can be used to write out a new
xref table. Finally we can write out the trailer dictionary with
appropriately computed /ID (see spec, 8.3, File Identifiers), the
cross reference table offset, and <literal>%%EOF</literal>.
</para>
</sect1>
<sect1 id="ref.filtered-streams">
<title>Filtered Streams</title>
<para>
Support for streams is implemented through the
<classname>Pipeline</classname> interface which was designed for
this package.
</para>
<para>
When reading streams, create a series of
<classname>Pipeline</classname> objects. The
<classname>Pipeline</classname> abstract base requires
implementation <function>write()</function> and
<function>finish()</function> and provides an implementation of
<function>getNext()</function>. Each pipeline object, upon
receiving data, does whatever it is going to do and then writes
the data (possibly modified) to its successor. Alternatively, a
pipeline may be an end-of-the-line pipeline that does something
like store its output to a file or a memory buffer ignoring a
successor. For additional details, look at
<filename>Pipeline.hh</filename>.
</para>
<para>
<classname>QPDF</classname> can read raw or filtered streams.
When reading a filtered stream, the <classname>QPDF</classname>
class creates a <classname>Pipeline</classname> object for one of
each appropriate filter object and chains them together. The last
filter should write to whatever type of output is required. The
<classname>QPDF</classname> class has an interface to write raw or
filtered stream contents to a given pipeline.
</para>
</sect1>
</chapter>
<chapter id="ref.linearization">
<title>Linearization</title>
<para>
This chapter describes how <classname>QPDF</classname> and
<classname>QPDFWriter</classname> implement creation and processing
of linearized PDFS.
</para>
<sect1 id="ref.linearization-strategy">
<title>Basic Strategy for Linearization</title>
<para>
To avoid the incestuous problem of having the qpdf library
validate its own linearized files, we have a special linearized
file checking mode which can be invoked via <command>qpdf
--check-linearization</command> (or <command>qpdf
--check</command>). This mode reads the linearization parameter
dictionary and the hint streams and validates that object
ordering, parameters, and hint stream contents are correct. The
validation code was first tested against linearized files created
by external tools (Acrobat and pdlin) and then used to validate
files created by <classname>QPDFWriter</classname> itself.
</para>
</sect1>
<sect1 id="ref.linearized.preparation">
<title>Preparing For Linearization</title>
<para>
Before creating a linearized PDF file from any other PDF file, the
PDF file must be altered such that all page attributes are
propagated down to the page level (and not inherited from parents
in the <literal>/Pages</literal> tree). We also have to know
which objects refer to which other objects, being concerned with
page boundaries and a few other cases. We refer to this part of
preparing the PDF file as <firstterm>optimization</firstterm>,
discussed in <xref linkend="ref.optimization"/>. Note the, in
this context, the term <firstterm>optimization</firstterm> is a
qpdf term, and the term <firstterm>linearization</firstterm> is a
term from the PDF specification. Do not be confused by the fact
that many applications refer to linearization as optimization or
web optimization.
</para>
<para>
When creating linearized PDF files from optimized PDF files, there
are really only a few issues that need to be dealt with:
<itemizedlist>
<listitem>
<para>
Creation of hints tables
</para>
</listitem>
<listitem>
<para>
Placing objects in the correct order
</para>
</listitem>
<listitem>
<para>
Filling in offsets and byte sizes
</para>
</listitem>
</itemizedlist>
</para>
</sect1>
<sect1 id="ref.optimization">
<title>Optimization</title>
<para>
In order to perform various operations such as linearization and
splitting files into pages, it is necessary to know which objects
are referenced by which pages, page thumbnails, and root and
trailer dictionary keys. It is also necessary to ensure that all
page-level attributes appear directly at the page level and are
not inherited from parents in the pages tree.
</para>
<para>
We refer to the process of enforcing these constraints as
<firstterm>optimization</firstterm>. As mentioned above, note
that some applications refer to linearization as optimization.
Although this optimization was initially motivated by the need to
create linearized files, we are using these terms separately.
</para>
<para>
PDF file optimization is implemented in the
<filename>QPDF_optimization.cc</filename> source file. That file
is richly commented and serves as the primary reference for the
optimization process.
</para>
<para>
After optimization has been completed, the private member
variables <varname>obj_user_to_objects</varname> and
<varname>object_to_obj_users</varname> in
<classname>QPDF</classname> have been populated. Any object that
has more than one value in the
<varname>object_to_obj_users</varname> table is shared. Any
object that has exactly one value in the
<varname>object_to_obj_users</varname> table is private. To find
all the private objects in a page or a trailer or root dictionary
key, one merely has make this determination for each element in
the <varname>obj_user_to_objects</varname> table for the given
page or key.
</para>
<para>
Note that pages and thumbnails have different object user types,
so the above test on a page will not include objects referenced by
the page's thumbnail dictionary and nothing else.
</para>
</sect1>
<sect1 id="ref.linearization.writing">
<title>Writing Linearized Files</title>
<para>
We will create files with only primary hint streams. We will
never write overflow hint streams. (As of PDF version 1.4,
Acrobat doesn't either, and they are never necessary.) The hint
streams contain offset information to objects that point to where
they would be if the hint stream were not present. This means
that we have to calculate all object positions before we can
generate and write the hint table. This means that we have to
generate the file in two passes. To make this reliable,
<classname>QPDFWriter</classname> in linearization mode invokes
exactly the same code twice to write the file to a pipeline.
</para>
<para>
In the first pass, the target pipeline is a count pipeline chained
to a discard pipeline. The count pipeline simply passes its data
through to the next pipeline in the chain but can return the
number of bytes passed through it at any intermediate point. The
discard pipeline is an end of line pipeline that just throws its
data away. The hint stream is not written and dummy values with
adequate padding are stored in the first cross reference table,
linearization parameter dictionary, and /Prev key of the first
trailer dictionary. All the offset, length, object renumbering
information, and anything else we need for the second pass is
stored.
</para>
<para>
At the end of the first pass, this information is passed to the
<classname>QPDF</classname> class which constructs a compressed
hint stream in a memory buffer and returns it.
<classname>QPDFWriter</classname> uses this information to write a
complete hint stream object into a memory buffer. At this point,
the length of the hint stream is known.
</para>
<para>
In the second pass, the end of the pipeline chain is a regular
file instead of a discard pipeline, and we have known values for
all the offsets and lengths that we didn't have in the first pass.
We have to adjust offsets that appear after the start of the hint
stream by the length of the hint stream, which is known. Anything
that is of variable length is padded, with the padding code
surrounding any writing code that differs in the two passes. This
ensures that changes to the way things are represented never
results in offsets that were gathered during the first pass
becoming incorrect for the second pass.
</para>
<para>
Using this strategy, we can write linearized files to a
non-seekable output stream with only a single pass to disk or
wherever the output is going.
</para>
</sect1>
<sect1 id="ref.linearization-data">
<title>Calculating Linearization Data</title>
<para>
Once a file is optimized, we have information about which objects
access which other objects. We can then process these tables to
decide which part (as described in &ldquo;Linearized PDF Document
Structure&rdquo; in the PDF specification) each object is
contained within. This tells us the exact order in which objects
are written. The <classname>QPDFWriter</classname> class asks for
this information and enqueues objects for writing in the proper
order. It also turns on a check that causes an exception to be
thrown if an object is encountered that has not already been
queued. (This could happen only if there were a bug in the
traversal code used to calculate the linearization data.)
</para>
</sect1>
<sect1 id="ref.linearization-issues">
<title>Known Issues with Linearization</title>
<para>
There are a handful of known issues with this linearization code.
These issues do not appear to impact the behavior of linearized
files which still work as intended: it is possible for a web
browser to begin to display them before they are fully
downloaded. In fact, it seems that various other programs that
create linearized files have many of these same issues. These
items make reference to terminology used in the linearization
appendix of the PDF specification.
<itemizedlist>
<listitem>
<para>
Thread Dictionary information keys appear in part 4 with the
rest of Threads instead of in part 9. Objects in part 9 are
not grouped together functionally.
</para>
</listitem>
<listitem>
<para>
We are not calculating numerators for shared object positions
within content streams or interleaving them within content
streams.
</para>
</listitem>
<listitem>
<para>
We generate only page offset, shared object, and outline hint
tables. It would be relatively easy to add some additional
tables. We gather most of the information needed to create
thumbnail hint tables. There are comments in the code about
this.
</para>
</listitem>
</itemizedlist>
</para>
</sect1>
<sect1 id="ref.linearization-debugging">
<title>Debugging Note</title>
<para>
The <command>qpdf --show-linearization</command> command can show
the complete contents of linearization hint streams. To look at
the raw data, you can extract the filtered contents of the
linearization hint tables using <command>qpdf --show-object=n
--filtered-stream-data</command>. Then, to convert this into a
bit stream (since linearization tables are bit streams written
without regard to byte boundaries), you can pipe the resulting
data through the following perl code:
<programlisting>use bytes;
binmode STDIN;
undef $/;
my $a = &lt;STDIN&gt;;
my @ch = split(//, $a);
map { printf("%08b", ord($_)) } @ch;
print "\n";
</programlisting>
</para>
</sect1>
</chapter>
<chapter id="ref.object-and-xref-streams">
<title>Object and Cross-Reference Streams</title>
<para>
This chapter provides information about the implementation of
object stream and cross-reference stream support in qpdf.
</para>
<sect1 id="ref.object-streams">
<title>Object Streams</title>
<para>
Object streams can contain any regular object except the
following:
<itemizedlist>
<listitem>
<para>
stream objects
</para>
</listitem>
<listitem>
<para>
objects with generation &gt; 0
</para>
</listitem>
<listitem>
<para>
the encryption dictionary
</para>
</listitem>
<listitem>
<para>
objects containing the /Length of another stream
</para>
</listitem>
</itemizedlist>
In addition, Adobe reader (at least as of version 8.0.0) appears
to not be able to handle having the document catalog appear in an
object stream if the file is encrypted, though this is not
specifically disallowed by the specification.
</para>
<para>
There are additional restrictions for linearized files. See <xref
linkend="ref.object-streams-linearization"/>for details.
</para>
<para>
The PDF specification refers to objects in object streams as
&ldquo;compressed objects&rdquo; regardless of whether the object
stream is compressed.
</para>
<para>
The generation number of every object in an object stream must be
zero. It is possible to delete and replace an object in an object
stream with a regular object.
</para>
<para>
The object stream dictionary has the following keys:
<itemizedlist>
<listitem>
<para>
<literal>/N</literal>: number of objects
</para>
</listitem>
<listitem>
<para>
<literal>/First</literal>: byte offset of first object
</para>
</listitem>
<listitem>
<para>
<literal>/Extends</literal>: indirect reference to stream that
this extends
</para>
</listitem>
</itemizedlist>
</para>
<para>
Stream collections are formed with <literal>/Extends</literal>.
They must form a directed acyclic graph. These can be used for
semantic information and are not meaningful to the PDF document's
syntactic structure. Although qpdf preserves stream collections,
it never generates them and doesn't make use of this information
in any way.
</para>
<para>
The specification recommends limiting the number of objects in
object stream for efficiency in reading and decoding. Acrobat 6
uses no more than objects per object stream for linearized files
and no more 200 objects per stream for non-linearized files.
<classname>QPDFWriter</classname>, in object stream generation
mode, never puts more than 100 objects in an object stream.
</para>
<para>
Object stream contents consists of <emphasis>N</emphasis> pairs of
integers, each of which is the object number and the byte offset
of the object relative to the first object in the stream, followed
by the objects themselves, concatenated.
</para>
</sect1>
<sect1 id="ref.xref-streams">
<title>Cross-Reference Streams</title>
<para>
For non-hybrid files, the value following
<literal>startxref</literal> is the byte offset to the xref stream
rather than the word <literal>xref</literal>.
</para>
<para>
For hybrid files (files containing both xref tables and
cross-reference streams), the xref table's trailer dictionary
contains the key <literal>/XRefStm</literal> whose value is the
byte offset to a cross-reference stream that supplements the xref
table. A PDF 1.5-compliant application should read the xref table
first. Then it should replace any object that it has already seen
with any defined in the xref stream. Then it should follow any
<literal>/Prev</literal> pointer in the original xref table's
trailer dictionary. The specification is not clear about what
should be done, if anything, with a <literal>/Prev</literal>
pointer in the xref stream referenced by an xref table. The
<classname>QPDF</classname> class ignores it, which is probably
reasonable since, if this case were to appear for any sensible PDF
file, the previous xref table would probably have a corresponding
<literal>/XRefStm</literal> pointer of its own. For example, if a
hybrid file were appended, the appended section would have its own
xref table and <literal>/XRefStm</literal>. The appended xref
table would point to the previous xref table which would point the
<literal>/XRefStm</literal>, meaning that the new
<literal>/XRefStm</literal> doesn't have to point to it.
</para>
<para>
Since xref streams must be read very early, they may not be
encrypted, and the may not contain indirect objects for keys
required to read them, which are these:
<itemizedlist>
<listitem>
<para>
<literal>/Type</literal>: value <literal>/XRef</literal>
</para>
</listitem>
<listitem>
<para>
<literal>/Size</literal>: value <emphasis>n+1</emphasis>: where
<emphasis>n</emphasis> is highest object number (same as
<literal>/Size</literal> in the trailer dictionary)
</para>
</listitem>
<listitem>
<para>
<literal>/Index</literal> (optional): value
<literal>[<replaceable>n count</replaceable> ...]</literal>
used to determine which objects' information is stored in this
stream. The default is <literal>[0 /Size]</literal>.
</para>
</listitem>
<listitem>
<para>
<literal>/Prev</literal>: value
<replaceable>offset</replaceable>: byte offset of previous xref
stream (same as <literal>/Prev</literal> in the trailer
dictionary)
</para>
</listitem>
<listitem>
<para>
<literal>/W [...]</literal>: sizes of each field in the xref
table
</para>
</listitem>
</itemizedlist>
</para>
<para>
The other fields in the xref stream, which may be indirect if
desired, are the union of those from the xref table's trailer
dictionary.
</para>
<sect2 id="ref.xref-stream-data">
<title>Cross-Reference Stream Data</title>
<para>
The stream data is binary and encoded in big-endian byte order.
Entries are concatenated, and each entry has a length equal to
the total of the entries in <literal>/W</literal> above. Each
entry consists of one or more fields, the first of which is the
type of the field. The number of bytes for each field is given
by <literal>/W</literal> above. A 0 in <literal>/W</literal>
indicates that the field is omitted and has the default value.
The default value for the field type is
&ldquo;<literal>1</literal>&rdquo;. All other default values are
&ldquo;<literal>0</literal>&rdquo;.
</para>
<para>
PDF 1.5 has three field types:
<itemizedlist>
<listitem>
<para>
0: for free objects. Format: <literal>0 obj
next-generation</literal>, same as the free table in a
traditional cross-reference table
</para>
</listitem>
<listitem>
<para>
1: regular non-compressed object. Format: <literal>1 offset
generation</literal>
</para>
</listitem>
<listitem>
<para>
2: for objects in object streams. Format: <literal>2
object-stream-number index</literal>, the number of object
stream containing the object and the index within the object
stream of the object.
</para>
</listitem>
</itemizedlist>
</para>
<para>
It seems standard to have the first entry in the table be
<literal>0 0 0</literal> instead of <literal>0 0 ffff</literal>
if there are no deleted objects.
</para>
</sect2>
</sect1>
<sect1 id="ref.object-streams-linearization">
<title>Implications for Linearized Files</title>
<para>
For linearized files, the linearization dictionary, document
catalog, and page objects may not be contained in object streams.
</para>
<para>
Objects stored within object streams are given the highest range
of object numbers within the main and first-page cross-reference
sections.
</para>
<para>
It is okay to use cross-reference streams in place of regular xref
tables. There are on special considerations.
</para>
<para>
Hint data refers to object streams themselves, not the objects in
the streams. Shared object references should also be made to the
object streams. There are no reference in any hint tables to the
object numbers of compressed objects (objects within object
streams).
</para>
<para>
When numbering objects, all shared objects within both the first
and second halves of the linearized files must be numbered
consecutively after all normal uncompressed objects in that half.
</para>
</sect1>
<sect1 id="ref.object-stream-implementation">
<title>Implementation Notes</title>
<para>
There are three modes for writing object streams:
<option>disable</option>, <option>preserve</option>, and
<option>generate</option>. In disable mode, we do not generate
any object streams, and we also generate an xref table rather than
xref streams. This can be used to generate PDF files that are
viewable with older readers. In preserve mode, we write object
streams such that written object streams contain the same objects
and <literal>/Extends</literal> relationships as in the original
file. This is equal to disable if the file has no object streams.
In generate, we create object streams ourselves by grouping
objects that are allowed in object streams together in sets of no
more than 100 objects. We also ensure that the PDF version is at
least 1.5 in generate mode, but we preserve the version header in
the other modes. The default is <option>preserve</option>.
</para>
<para>
We do not support creation of hybrid files. When we write files,
even in preserve mode, we will lose any xref tables and merge any
appended sections.
</para>
</sect1>
</chapter>
<appendix id="ref.release-notes">
<title>Release Notes</title>
<variablelist>
<varlistentry>
<term>2.0.2: June 30, 2008</term>
<listitem>
<itemizedlist>
<listitem>
<para>
Update test suite to work properly with a
non-<command>bash</command> <filename>/bin/sh</filename> and
with Perl 5.10. No changes were made to the actual qpdf
source code itself for this release.
</para>
</listitem>
</itemizedlist>
</listitem>
</varlistentry>
<varlistentry>
<term>2.0.1: May 6, 2008</term>
<listitem>
<itemizedlist>
<listitem>
<para>
No changes in functionality or interface. This release
includes fixes to the source code so that qpdf compiles
properly and passes its test suite on a broader range of
platforms. See <filename>ChangeLog</filename> in the source
distribution for details.
</para>
</listitem>
</itemizedlist>
</listitem>
</varlistentry>
<varlistentry>
<term>2.0: April 29, 2008</term>
<listitem>
<itemizedlist>
<listitem>
<para>
First public release.
</para>
</listitem>
</itemizedlist>
</listitem>
</varlistentry>
</variablelist>
</appendix>
</book>