PDFMergerUtility.java
/*
* Licensed to the Apache Software Foundation (ASF) under one or more
* contributor license agreements. See the NOTICE file distributed with
* this work for additional information regarding copyright ownership.
* The ASF licenses this file to You under the Apache License, Version 2.0
* (the "License"); you may not use this file except in compliance with
* the License. You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
package org.apache.pdfbox.multipdf;
import java.io.File;
import java.io.FileNotFoundException;
import java.io.IOException;
import java.io.InputStream;
import java.io.OutputStream;
import java.util.ArrayList;
import java.util.Arrays;
import java.util.Collections;
import java.util.HashMap;
import java.util.HashSet;
import java.util.LinkedHashMap;
import java.util.List;
import java.util.Map;
import java.util.Set;
import org.apache.commons.logging.Log;
import org.apache.commons.logging.LogFactory;
import org.apache.pdfbox.Loader;
import org.apache.pdfbox.cos.COSArray;
import org.apache.pdfbox.cos.COSBase;
import org.apache.pdfbox.cos.COSDictionary;
import org.apache.pdfbox.cos.COSInteger;
import org.apache.pdfbox.cos.COSName;
import org.apache.pdfbox.cos.COSNumber;
import org.apache.pdfbox.cos.COSObject;
import org.apache.pdfbox.cos.COSStream;
import org.apache.pdfbox.io.IOUtils;
import org.apache.pdfbox.io.MemoryUsageSetting;
import org.apache.pdfbox.pdfwriter.compress.CompressParameters;
import org.apache.pdfbox.pdmodel.PDDocument;
import org.apache.pdfbox.pdmodel.PDDocumentCatalog;
import org.apache.pdfbox.pdmodel.PDDocumentInformation;
import org.apache.pdfbox.pdmodel.PDDocumentNameDestinationDictionary;
import org.apache.pdfbox.pdmodel.PDDocumentNameDictionary;
import org.apache.pdfbox.pdmodel.PDPage;
import org.apache.pdfbox.pdmodel.PDResources;
import org.apache.pdfbox.pdmodel.PDStructureElementNameTreeNode;
import org.apache.pdfbox.pdmodel.PageMode;
import org.apache.pdfbox.pdmodel.common.COSObjectable;
import org.apache.pdfbox.pdmodel.common.PDDestinationOrAction;
import org.apache.pdfbox.pdmodel.common.PDMetadata;
import org.apache.pdfbox.pdmodel.common.PDNameTreeNode;
import org.apache.pdfbox.pdmodel.common.PDNumberTreeNode;
import org.apache.pdfbox.pdmodel.common.PDStream;
import org.apache.pdfbox.pdmodel.documentinterchange.logicalstructure.PDMarkInfo;
import org.apache.pdfbox.pdmodel.documentinterchange.logicalstructure.PDParentTreeValue;
import org.apache.pdfbox.pdmodel.documentinterchange.logicalstructure.PDStructureElement;
import org.apache.pdfbox.pdmodel.documentinterchange.logicalstructure.PDStructureTreeRoot;
import org.apache.pdfbox.pdmodel.graphics.color.PDOutputIntent;
import org.apache.pdfbox.pdmodel.interactive.action.PDActionGoTo;
import org.apache.pdfbox.pdmodel.interactive.annotation.PDAnnotation;
import org.apache.pdfbox.pdmodel.interactive.documentnavigation.destination.PDDestination;
import org.apache.pdfbox.pdmodel.interactive.documentnavigation.destination.PDPageDestination;
import org.apache.pdfbox.pdmodel.interactive.documentnavigation.outline.PDDocumentOutline;
import org.apache.pdfbox.pdmodel.interactive.documentnavigation.outline.PDOutlineItem;
import org.apache.pdfbox.pdmodel.interactive.form.PDAcroForm;
import org.apache.pdfbox.pdmodel.interactive.form.PDField;
import org.apache.pdfbox.pdmodel.interactive.viewerpreferences.PDViewerPreferences;
/**
* This class will take a list of pdf documents and merge them, saving the
* result in a new document.
*
* @author Ben Litchfield
*/
public class PDFMergerUtility
{
/**
* Log instance.
*/
private static final Log LOG = LogFactory.getLog(PDFMergerUtility.class);
private final List<Object> sources;
private String destinationFileName;
private OutputStream destinationStream;
private boolean ignoreAcroFormErrors = false;
private PDDocumentInformation destinationDocumentInformation = null;
private PDMetadata destinationMetadata = null;
private DocumentMergeMode documentMergeMode = DocumentMergeMode.PDFBOX_LEGACY_MODE;
private AcroFormMergeMode acroFormMergeMode = AcroFormMergeMode.PDFBOX_LEGACY_MODE;
/**
* The mode to use when merging documents:
*
* <ul>
* <li>{@link DocumentMergeMode#OPTIMIZE_RESOURCES_MODE} Optimizes resource handling such as
* closing documents early. <strong>Not all document elements are merged</strong> compared to
* the PDFBOX_LEGACY_MODE. Currently supported are:
* <ul>
* <li>Page content and resources
* </ul>
* <li>{@link DocumentMergeMode#PDFBOX_LEGACY_MODE} Keeps all files open until the
* merge has been completed. This is currently necessary to merge documents
* containing a Structure Tree.
* </ul>
*/
public enum DocumentMergeMode
{
OPTIMIZE_RESOURCES_MODE,
PDFBOX_LEGACY_MODE
}
/**
* The mode to use when merging AcroForm between documents:
*
* <ul>
* <li>{@link AcroFormMergeMode#JOIN_FORM_FIELDS_MODE} fields with the same fully qualified name
* will be merged into one with the widget annotations of the merged fields
* becoming part of the same field.
* <li>{@link AcroFormMergeMode#PDFBOX_LEGACY_MODE} fields with the same fully qualified name
* will be renamed and treated as independent. This mode was used in versions
* of PDFBox up to 2.x.
* </ul>
*/
public enum AcroFormMergeMode
{
JOIN_FORM_FIELDS_MODE,
PDFBOX_LEGACY_MODE
}
/**
* Instantiate a new PDFMergerUtility.
*/
public PDFMergerUtility()
{
sources = new ArrayList<>();
}
/**
* Get the merge mode to be used for merging AcroForms between documents
*
* {@link AcroFormMergeMode}
*/
public AcroFormMergeMode getAcroFormMergeMode()
{
return acroFormMergeMode;
}
/**
* Set the merge mode to be used for merging AcroForms between documents
*
* {@link AcroFormMergeMode}
*/
public void setAcroFormMergeMode(AcroFormMergeMode theAcroFormMergeMode)
{
this.acroFormMergeMode = theAcroFormMergeMode;
}
/**
* Get the merge mode to be used for merging documents
*
* {@link DocumentMergeMode}
*/
public DocumentMergeMode getDocumentMergeMode()
{
return documentMergeMode;
}
/**
* Set the merge mode to be used for merging documents
*
* {@link DocumentMergeMode}
*/
public void setDocumentMergeMode(DocumentMergeMode theDocumentMergeMode)
{
this.documentMergeMode = theDocumentMergeMode;
}
/**
* Set the mode to be used for merging the documents
*
* {@link DocumentMergeMode}
*/
public void setAcroFormMergeMode(DocumentMergeMode theDocumentMergeMode)
{
this.documentMergeMode = theDocumentMergeMode;
}
/**
* Get the name of the destination file.
*
* @return Returns the destination.
*/
public String getDestinationFileName()
{
return destinationFileName;
}
/**
* Set the name of the destination file.
*
* @param destination The destination to set.
*/
public void setDestinationFileName(String destination)
{
destinationFileName = destination;
}
/**
* Get the destination OutputStream.
*
* @return Returns the destination OutputStream.
*/
public OutputStream getDestinationStream()
{
return destinationStream;
}
/**
* Set the destination OutputStream.
*
* @param destStream The destination to set.
*/
public void setDestinationStream(OutputStream destStream)
{
destinationStream = destStream;
}
/**
* Get the destination document information that is to be set in {@link #mergeDocuments(org.apache.pdfbox.io.MemoryUsageSetting)
* }. The default is null, which means that it is ignored.
*
* @return The destination document information.
*/
public PDDocumentInformation getDestinationDocumentInformation()
{
return destinationDocumentInformation;
}
/**
* Set the destination document information that is to be set in {@link #mergeDocuments(org.apache.pdfbox.io.MemoryUsageSetting)
* }. The default is null, which means that it is ignored.
*
* @param info The destination document information.
*/
public void setDestinationDocumentInformation(PDDocumentInformation info)
{
destinationDocumentInformation = info;
}
/**
* Set the destination metadata that is to be set in {@link #mergeDocuments(org.apache.pdfbox.io.MemoryUsageSetting)
* }. The default is null, which means that it is ignored.
*
* @return The destination metadata.
*/
public PDMetadata getDestinationMetadata()
{
return destinationMetadata;
}
/**
* Set the destination metadata that is to be set in {@link #mergeDocuments(org.apache.pdfbox.io.MemoryUsageSetting)
* }. The default is null, which means that it is ignored.
*
* @param meta The destination metadata.
*/
public void setDestinationMetadata(PDMetadata meta)
{
destinationMetadata = meta;
}
/**
* Add a source file to the list of files to merge.
*
* @param source Full path and file name of source document.
*
* @throws FileNotFoundException If the file doesn't exist
*/
public void addSource(String source) throws FileNotFoundException
{
addSource(new File(source));
}
/**
* Add a source file to the list of files to merge.
*
* @param source File representing source document
*
* @throws FileNotFoundException If the file doesn't exist
*/
public void addSource(File source) throws FileNotFoundException
{
sources.add(source);
}
/**
* Add a source to the list of documents to merge.
*
* @param source InputStream representing source document
*/
public void addSource(InputStream source)
{
sources.add(source);
}
/**
* Add a list of sources to the list of documents to merge.
*
* @param sourcesList List of InputStream objects representing source
* documents
*/
public void addSources(List<InputStream> sourcesList)
{
sources.addAll(sourcesList);
}
/**
* Merge the list of source documents, saving the result in the destination
* file.
*
* @param memUsageSetting defines how memory is used for buffering PDF streams;
* in case of <code>null</code> unrestricted main memory is used
*
* @throws IOException If there is an error saving the document.
*/
public void mergeDocuments(MemoryUsageSetting memUsageSetting) throws IOException
{
mergeDocuments(memUsageSetting, CompressParameters.DEFAULT_COMPRESSION);
}
/**
* Merge the list of source documents, saving the result in the destination file.
*
* @param memUsageSetting defines how memory is used for buffering PDF streams; in case of <code>null</code>
* unrestricted main memory is used
* @param compressParameters defines if compressed object streams are enabled
*
* @throws IOException If there is an error saving the document.
*/
public void mergeDocuments(MemoryUsageSetting memUsageSetting, CompressParameters compressParameters) throws IOException
{
if (documentMergeMode == DocumentMergeMode.PDFBOX_LEGACY_MODE)
{
legacyMergeDocuments(memUsageSetting, compressParameters);
}
else if (documentMergeMode == DocumentMergeMode.OPTIMIZE_RESOURCES_MODE)
{
optimizedMergeDocuments(memUsageSetting, compressParameters);
}
}
private void optimizedMergeDocuments(MemoryUsageSetting memUsageSetting,
CompressParameters compressParameters) throws IOException
{
try (PDDocument destination = new PDDocument(memUsageSetting))
{
PDFCloneUtility cloner = new PDFCloneUtility(destination);
for (Object sourceObject : sources)
{
PDDocument sourceDoc = null;
try
{
if (sourceObject instanceof File)
{
sourceDoc = Loader.loadPDF((File) sourceObject, memUsageSetting);
}
else
{
sourceDoc = Loader.loadPDF((InputStream) sourceObject, memUsageSetting);
}
for (PDPage page : sourceDoc.getPages())
{
PDPage newPage = new PDPage(
(COSDictionary) cloner.cloneForNewDocument(page.getCOSObject()));
newPage.setCropBox(page.getCropBox());
newPage.setMediaBox(page.getMediaBox());
newPage.setRotation(page.getRotation());
PDResources resources = page.getResources();
if (resources != null)
{
// this is smart enough to just create references for resources that are used on multiple
// pages
newPage.setResources(new PDResources(
(COSDictionary) cloner.cloneForNewDocument(resources)));
}
else
{
newPage.setResources(new PDResources());
}
destination.addPage(newPage);
}
}
finally
{
IOUtils.closeQuietly(sourceDoc);
}
}
if (destinationStream == null)
{
destination.save(destinationFileName, compressParameters);
}
else
{
destination.save(destinationStream, compressParameters);
}
}
}
/**
* Merge the list of source documents, saving the result in the destination
* file.
*
* @param memUsageSetting defines how memory is used for buffering PDF streams;
* in case of <code>null</code> unrestricted main memory is used
*
* @throws IOException If there is an error saving the document.
*/
private void legacyMergeDocuments(MemoryUsageSetting memUsageSetting,
CompressParameters compressParameters) throws IOException
{
if (!sources.isEmpty())
{
// Make sure that:
// - first Exception is kept
// - all PDDocuments are closed
// - all FileInputStreams are closed
// - there's a way to see which errors occurred
List<PDDocument> tobeclosed = new ArrayList<>(sources.size());
MemoryUsageSetting partitionedMemSetting = memUsageSetting != null ?
memUsageSetting.getPartitionedCopy(sources.size()+1) :
MemoryUsageSetting.setupMainMemoryOnly();
try (PDDocument destination = new PDDocument(partitionedMemSetting))
{
for (Object sourceObject : sources)
{
PDDocument sourceDoc = null;
if (sourceObject instanceof File)
{
sourceDoc = Loader.loadPDF((File) sourceObject, partitionedMemSetting);
}
else
{
sourceDoc = Loader.loadPDF((InputStream) sourceObject,
partitionedMemSetting);
}
tobeclosed.add(sourceDoc);
appendDocument(destination, sourceDoc);
}
// optionally set meta data
if (destinationDocumentInformation != null)
{
destination.setDocumentInformation(destinationDocumentInformation);
}
if (destinationMetadata != null)
{
destination.getDocumentCatalog().setMetadata(destinationMetadata);
}
if (destinationStream == null)
{
destination.save(destinationFileName, compressParameters);
}
else
{
destination.save(destinationStream, compressParameters);
}
}
finally
{
for (PDDocument doc : tobeclosed)
{
IOUtils.closeAndLogException(doc, LOG, "PDDocument", null);
}
}
}
}
/**
* append all pages from source to destination.
*
* @param destination the document to receive the pages
* @param source the document originating the new pages
*
* @throws IOException If there is an error accessing data from either
* document.
*/
public void appendDocument(PDDocument destination, PDDocument source) throws IOException
{
if (source.getDocument().isClosed())
{
throw new IOException("Error: source PDF is closed.");
}
if (destination.getDocument().isClosed())
{
throw new IOException("Error: destination PDF is closed.");
}
PDDocumentCatalog srcCatalog = source.getDocumentCatalog();
if (isDynamicXfa(srcCatalog.getAcroForm()))
{
throw new IOException("Error: can't merge source document containing dynamic XFA form content.");
}
PDDocumentInformation destInfo = destination.getDocumentInformation();
PDDocumentInformation srcInfo = source.getDocumentInformation();
mergeInto(srcInfo.getCOSObject(), destInfo.getCOSObject(), Collections.<COSName>emptySet());
// use the highest version number for the resulting pdf
float destVersion = destination.getVersion();
float srcVersion = source.getVersion();
if (destVersion < srcVersion)
{
destination.setVersion(srcVersion);
}
int pageIndexOpenActionDest = -1;
PDDocumentCatalog destCatalog = destination.getDocumentCatalog();
if (destCatalog.getOpenAction() == null)
{
// PDFBOX-3972: get local dest page index, it must be reassigned after the page cloning
PDDestinationOrAction openAction = null;
try
{
openAction = srcCatalog.getOpenAction();
}
catch (IOException ex)
{
// PDFBOX-4223
LOG.error("Invalid OpenAction ignored", ex);
}
PDDestination openActionDestination = null;
if (openAction instanceof PDActionGoTo)
{
openActionDestination = ((PDActionGoTo) openAction).getDestination();
}
else if (openAction instanceof PDDestination)
{
openActionDestination = (PDDestination) openAction;
}
// note that it can also be something else, e.g. PDActionJavaScript, then do nothing
if (openActionDestination instanceof PDPageDestination)
{
PDPage page = ((PDPageDestination) openActionDestination).getPage();
if (page != null)
{
pageIndexOpenActionDest = srcCatalog.getPages().indexOf(page);
}
}
destCatalog.setOpenAction(openAction);
}
PDFCloneUtility cloner = new PDFCloneUtility(destination);
mergeAcroForm(cloner, destCatalog, srcCatalog);
COSArray destThreads = destCatalog.getCOSObject().getCOSArray(COSName.THREADS);
COSArray srcThreads = (COSArray) cloner.cloneForNewDocument(destCatalog.getCOSObject().getDictionaryObject(
COSName.THREADS));
if (destThreads == null)
{
destCatalog.getCOSObject().setItem(COSName.THREADS, srcThreads);
}
else
{
destThreads.addAll(srcThreads);
}
PDDocumentNameDictionary destNames = destCatalog.getNames();
PDDocumentNameDictionary srcNames = srcCatalog.getNames();
if (srcNames != null)
{
if (destNames == null)
{
destCatalog.getCOSObject().setItem(COSName.NAMES, cloner.cloneForNewDocument(srcNames));
}
else
{
cloner.cloneMerge(srcNames, destNames);
}
}
if (destNames != null && destNames.getCOSObject().containsKey(COSName.ID_TREE))
{
// found in 001031.pdf from PDFBOX-4417 and doesn't belong there
destNames.getCOSObject().removeItem(COSName.ID_TREE);
LOG.warn("Removed /IDTree from /Names dictionary, doesn't belong there");
}
PDDocumentNameDestinationDictionary srcDests = srcCatalog.getDests();
if (srcDests != null)
{
PDDocumentNameDestinationDictionary destDests = destCatalog.getDests();
if (destDests == null)
{
destCatalog.getCOSObject().setItem(COSName.DESTS, cloner.cloneForNewDocument(srcDests));
}
else
{
cloner.cloneMerge(srcDests, destDests);
}
}
PDDocumentOutline srcOutline = srcCatalog.getDocumentOutline();
if (srcOutline != null)
{
PDDocumentOutline destOutline = destCatalog.getDocumentOutline();
if (destOutline == null || destOutline.getFirstChild() == null)
{
PDDocumentOutline cloned = new PDDocumentOutline((COSDictionary) cloner.cloneForNewDocument(srcOutline));
destCatalog.setDocumentOutline(cloned);
}
else
{
// search last sibling for dest, because /Last entry is sometimes wrong
PDOutlineItem destLastOutlineItem = destOutline.getFirstChild();
while (true)
{
PDOutlineItem outlineItem = destLastOutlineItem.getNextSibling();
if (outlineItem == null)
{
break;
}
destLastOutlineItem = outlineItem;
}
for (PDOutlineItem item : srcOutline.children())
{
// get each child, clone its dictionary, remove siblings info,
// append outline item created from there
COSDictionary clonedDict = (COSDictionary) cloner.cloneForNewDocument(item);
clonedDict.removeItem(COSName.PREV);
clonedDict.removeItem(COSName.NEXT);
PDOutlineItem clonedItem = new PDOutlineItem(clonedDict);
destLastOutlineItem.insertSiblingAfter(clonedItem);
destLastOutlineItem = destLastOutlineItem.getNextSibling();
}
}
}
PageMode destPageMode = destCatalog.getPageMode();
if (destPageMode == null)
{
PageMode srcPageMode = srcCatalog.getPageMode();
destCatalog.setPageMode(srcPageMode);
}
COSDictionary srcLabels = srcCatalog.getCOSObject().getCOSDictionary(COSName.PAGE_LABELS);
if (srcLabels != null)
{
int destPageCount = destination.getNumberOfPages();
COSArray destNums;
COSDictionary destLabels = destCatalog.getCOSObject().getCOSDictionary(COSName.PAGE_LABELS);
if (destLabels == null)
{
destLabels = new COSDictionary();
destNums = new COSArray();
destLabels.setItem(COSName.NUMS, destNums);
destCatalog.getCOSObject().setItem(COSName.PAGE_LABELS, destLabels);
}
else
{
destNums = (COSArray) destLabels.getDictionaryObject(COSName.NUMS);
}
COSArray srcNums = (COSArray) srcLabels.getDictionaryObject(COSName.NUMS);
if (srcNums != null)
{
int startSize = destNums.size();
for (int i = 0; i < srcNums.size(); i += 2)
{
COSBase base = srcNums.getObject(i);
if (!(base instanceof COSNumber))
{
LOG.error("page labels ignored, index " + i + " should be a number, but is " + base);
// remove what we added
while (destNums.size() > startSize)
{
destNums.remove(startSize);
}
break;
}
COSNumber labelIndex = (COSNumber) base;
long labelIndexValue = labelIndex.intValue();
destNums.add(COSInteger.get(labelIndexValue + destPageCount));
destNums.add(cloner.cloneForNewDocument(srcNums.getObject(i + 1)));
}
}
}
COSStream destMetadata = destCatalog.getCOSObject().getCOSStream(COSName.METADATA);
COSStream srcMetadata = srcCatalog.getCOSObject().getCOSStream(COSName.METADATA);
if (destMetadata == null && srcMetadata != null)
{
try
{
PDStream newStream = new PDStream(destination, srcMetadata.createInputStream(), (COSName) null);
mergeInto(srcMetadata, newStream.getCOSObject(),
new HashSet<>(Arrays.asList(COSName.FILTER, COSName.LENGTH)));
destCatalog.getCOSObject().setItem(COSName.METADATA, newStream);
}
catch (IOException ex)
{
// PDFBOX-4227 cleartext XMP stream with /Flate
LOG.error("Metadata skipped because it could not be read", ex);
}
}
COSDictionary destOCP = destCatalog.getCOSObject().getCOSDictionary(COSName.OCPROPERTIES);
COSDictionary srcOCP = srcCatalog.getCOSObject().getCOSDictionary(COSName.OCPROPERTIES);
if (destOCP == null && srcOCP != null)
{
destCatalog.getCOSObject().setItem(COSName.OCPROPERTIES, cloner.cloneForNewDocument(srcOCP));
}
else if (destOCP != null && srcOCP != null)
{
cloner.cloneMerge(srcOCP, destOCP);
}
mergeOutputIntents(cloner, srcCatalog, destCatalog);
// merge logical structure hierarchy
boolean mergeStructTree = false;
int destParentTreeNextKey = -1;
Map<Integer, COSObjectable> srcNumberTreeAsMap = null;
Map<Integer, COSObjectable> destNumberTreeAsMap = null;
PDStructureTreeRoot srcStructTree = srcCatalog.getStructureTreeRoot();
PDStructureTreeRoot destStructTree = destCatalog.getStructureTreeRoot();
if (destStructTree == null && srcStructTree != null)
{
// create a dummy structure tree in the destination, so that the source
// tree is cloned. (We can't just copy the tree reference due to PDFBOX-3999)
destStructTree = new PDStructureTreeRoot();
destCatalog.setStructureTreeRoot(destStructTree);
destStructTree.setParentTree(new PDNumberTreeNode(PDParentTreeValue.class));
// PDFBOX-4429: remove bogus StructParent(s)
for (PDPage page : destCatalog.getPages())
{
page.getCOSObject().removeItem(COSName.STRUCT_PARENTS);
for (PDAnnotation ann : page.getAnnotations())
{
ann.getCOSObject().removeItem(COSName.STRUCT_PARENT);
}
}
}
if (destStructTree != null)
{
PDNumberTreeNode destParentTree = destStructTree.getParentTree();
destParentTreeNextKey = destStructTree.getParentTreeNextKey();
if (destParentTree != null)
{
destNumberTreeAsMap = getNumberTreeAsMap(destParentTree);
if (destParentTreeNextKey < 0)
{
if (destNumberTreeAsMap.isEmpty())
{
destParentTreeNextKey = 0;
}
else
{
destParentTreeNextKey = Collections.max(destNumberTreeAsMap.keySet()) + 1;
}
}
if (destParentTreeNextKey >= 0 && srcStructTree != null)
{
PDNumberTreeNode srcParentTree = srcStructTree.getParentTree();
if (srcParentTree != null)
{
srcNumberTreeAsMap = getNumberTreeAsMap(srcParentTree);
if (!srcNumberTreeAsMap.isEmpty())
{
mergeStructTree = true;
}
}
}
}
}
Map<COSDictionary, COSDictionary> objMapping = new HashMap<>();
int pageIndex = 0;
for (PDPage page : srcCatalog.getPages())
{
PDPage newPage = new PDPage((COSDictionary) cloner.cloneForNewDocument(page.getCOSObject()));
if (!mergeStructTree)
{
// PDFBOX-4429: remove bogus StructParent(s)
newPage.getCOSObject().removeItem(COSName.STRUCT_PARENTS);
for (PDAnnotation ann : newPage.getAnnotations())
{
ann.getCOSObject().removeItem(COSName.STRUCT_PARENT);
}
}
newPage.setCropBox(page.getCropBox());
newPage.setMediaBox(page.getMediaBox());
newPage.setRotation(page.getRotation());
PDResources resources = page.getResources();
if (resources != null)
{
// this is smart enough to just create references for resources that are used on multiple pages
newPage.setResources(new PDResources((COSDictionary) cloner.cloneForNewDocument(resources)));
}
else
{
newPage.setResources(new PDResources());
}
if (mergeStructTree)
{
// add the value of the destination ParentTreeNextKey to every source element
// StructParent(s) value so that these don't overlap with the existing values
updateStructParentEntries(newPage, destParentTreeNextKey);
objMapping.put(page.getCOSObject(), newPage.getCOSObject());
List<PDAnnotation> oldAnnots = page.getAnnotations();
List<PDAnnotation> newAnnots = newPage.getAnnotations();
for (int i = 0; i < oldAnnots.size(); i++)
{
objMapping.put(oldAnnots.get(i).getCOSObject(), newAnnots.get(i).getCOSObject());
}
// TODO update mapping for XObjects
}
destination.addPage(newPage);
if (pageIndex == pageIndexOpenActionDest)
{
// PDFBOX-3972: reassign the page.
// The openAction is either a PDActionGoTo or a PDPageDestination
PDDestinationOrAction openAction = destCatalog.getOpenAction();
PDPageDestination pageDestination;
if (openAction instanceof PDActionGoTo)
{
pageDestination = (PDPageDestination) ((PDActionGoTo) openAction).getDestination();
}
else
{
pageDestination = (PDPageDestination) openAction;
}
pageDestination.setPage(newPage);
}
++pageIndex;
}
if (mergeStructTree)
{
updatePageReferences(cloner, srcNumberTreeAsMap, objMapping);
int maxSrcKey = -1;
for (Map.Entry<Integer, COSObjectable> entry : srcNumberTreeAsMap.entrySet())
{
int srcKey = entry.getKey();
maxSrcKey = Math.max(srcKey, maxSrcKey);
destNumberTreeAsMap.put(destParentTreeNextKey + srcKey, cloner.cloneForNewDocument(entry.getValue()));
}
destParentTreeNextKey += maxSrcKey + 1;
PDNumberTreeNode newParentTreeNode = new PDNumberTreeNode(PDParentTreeValue.class);
// Note that all elements are stored flatly. This could become a problem for large files
// when these are opened in a viewer that uses the tagging information.
// If this happens, then ​PDNumberTreeNode should be improved with a convenience method that
// stores the map into a B+Tree, see https://en.wikipedia.org/wiki/B+_tree
newParentTreeNode.setNumbers(destNumberTreeAsMap);
destStructTree.setParentTree(newParentTreeNode);
destStructTree.setParentTreeNextKey(destParentTreeNextKey);
mergeKEntries(cloner, srcStructTree, destStructTree);
mergeRoleMap(srcStructTree, destStructTree);
mergeIDTree(cloner, srcStructTree, destStructTree);
mergeMarkInfo(destCatalog, srcCatalog);
mergeLanguage(destCatalog, srcCatalog);
mergeViewerPreferences(destCatalog, srcCatalog);
}
}
private void mergeViewerPreferences(PDDocumentCatalog destCatalog, PDDocumentCatalog srcCatalog)
{
PDViewerPreferences srcViewerPreferences = srcCatalog.getViewerPreferences();
if (srcViewerPreferences == null)
{
return;
}
PDViewerPreferences destViewerPreferences = destCatalog.getViewerPreferences();
if (destViewerPreferences == null)
{
destViewerPreferences = new PDViewerPreferences(new COSDictionary());
destCatalog.setViewerPreferences(destViewerPreferences);
}
mergeInto(srcViewerPreferences.getCOSObject(), destViewerPreferences.getCOSObject(),
Collections.<COSName>emptySet());
// check the booleans - set to true if one is set and true
if (srcViewerPreferences.hideToolbar() || destViewerPreferences.hideToolbar())
{
destViewerPreferences.setHideToolbar(true);
}
if (srcViewerPreferences.hideMenubar() || destViewerPreferences.hideMenubar())
{
destViewerPreferences.setHideMenubar(true);
}
if (srcViewerPreferences.hideWindowUI() || destViewerPreferences.hideWindowUI())
{
destViewerPreferences.setHideWindowUI(true);
}
if (srcViewerPreferences.fitWindow() || destViewerPreferences.fitWindow())
{
destViewerPreferences.setFitWindow(true);
}
if (srcViewerPreferences.centerWindow() || destViewerPreferences.centerWindow())
{
destViewerPreferences.setCenterWindow(true);
}
if (srcViewerPreferences.displayDocTitle() || destViewerPreferences.displayDocTitle())
{
destViewerPreferences.setDisplayDocTitle(true);
}
}
private void mergeLanguage(PDDocumentCatalog destCatalog, PDDocumentCatalog srcCatalog)
{
if (destCatalog.getLanguage() == null)
{
String srcLanguage = srcCatalog.getLanguage();
if (srcLanguage != null)
{
destCatalog.setLanguage(srcLanguage);
}
}
}
private void mergeMarkInfo(PDDocumentCatalog destCatalog, PDDocumentCatalog srcCatalog)
{
PDMarkInfo destMark = destCatalog.getMarkInfo();
PDMarkInfo srcMark = srcCatalog.getMarkInfo();
if (destMark == null)
{
destMark = new PDMarkInfo();
}
if (srcMark == null)
{
srcMark = new PDMarkInfo();
}
destMark.setMarked(true);
destMark.setSuspect(srcMark.isSuspect() || destMark.isSuspect());
destMark.setSuspect(srcMark.usesUserProperties() || destMark.usesUserProperties());
destCatalog.setMarkInfo(destMark);
}
private void mergeKEntries(PDFCloneUtility cloner,
PDStructureTreeRoot srcStructTree,
PDStructureTreeRoot destStructTree) throws IOException
{
COSBase srcKEntry = srcStructTree.getK();
COSArray srcKArray = new COSArray();
COSBase clonedSrcKEntry = cloner.cloneForNewDocument(srcKEntry);
if (clonedSrcKEntry instanceof COSArray)
{
srcKArray.addAll((COSArray) clonedSrcKEntry);
}
else if (clonedSrcKEntry instanceof COSDictionary)
{
srcKArray.add(clonedSrcKEntry);
}
if (srcKArray.size() == 0)
{
return;
}
COSArray dstKArray = new COSArray();
COSBase dstKEntry = destStructTree.getK();
if (dstKEntry instanceof COSArray)
{
dstKArray.addAll((COSArray) dstKEntry);
}
else if (dstKEntry instanceof COSDictionary)
{
dstKArray.add(dstKEntry);
}
if (dstKArray.size() == 1 && dstKArray.getObject(0) instanceof COSDictionary)
{
// Only one element in the destination. If it is a /Document and its children
// are /Document or /Part, then we can insert there
COSDictionary topKDict = (COSDictionary) dstKArray.getObject(0);
if (COSName.DOCUMENT.equals(topKDict.getCOSName(COSName.S)))
{
COSArray kLevelOneArray = topKDict.getCOSArray(COSName.K);
if (kLevelOneArray != null)
{
boolean onlyDocuments = hasOnlyDocumentsOrParts(kLevelOneArray);
if (onlyDocuments)
{
// insert src elements at level 1
kLevelOneArray.addAll(srcKArray);
updateParentEntry(kLevelOneArray, topKDict, COSName.PART);
return;
}
}
}
}
if (dstKArray.size() == 0)
{
updateParentEntry(srcKArray, destStructTree.getCOSObject(), null);
destStructTree.setK(srcKArray);
return;
}
// whatever this is, merge this under a new /Document element
dstKArray.addAll(srcKArray);
COSDictionary kLevelZeroDict = new COSDictionary();
// If it is all Document, then make it all Part
COSName newStructureType = hasOnlyDocumentsOrParts(dstKArray) ? COSName.PART : null;
updateParentEntry(dstKArray, kLevelZeroDict, newStructureType);
kLevelZeroDict.setItem(COSName.K, dstKArray);
kLevelZeroDict.setItem(COSName.P, destStructTree);
kLevelZeroDict.setItem(COSName.S, COSName.DOCUMENT);
destStructTree.setK(kLevelZeroDict);
}
private boolean hasOnlyDocumentsOrParts(COSArray kLevelOneArray)
{
for (int i = 0; i < kLevelOneArray.size(); ++i)
{
COSBase base = kLevelOneArray.getObject(i);
if (!(base instanceof COSDictionary))
{
return false;
}
COSDictionary dict = (COSDictionary) base;
COSName sEntry = dict.getCOSName(COSName.S);
if (!COSName.DOCUMENT.equals(sEntry) && !COSName.PART.equals(sEntry))
{
return false;
}
}
return true;
}
/**
* Update the P reference to the new parent dictionary.
*
* @param kArray the kids array
* @param newParent the new parent
* @param newStructureType the new structure type in /S or null so it doesn't get replaced
*/
private void updateParentEntry(COSArray kArray, COSDictionary newParent, COSName newStructureType)
{
for (int i = 0; i < kArray.size(); i++)
{
COSBase subEntry = kArray.getObject(i);
if (subEntry instanceof COSDictionary)
{
COSDictionary dictEntry = (COSDictionary) subEntry;
dictEntry.setItem(COSName.P, newParent);
if (newStructureType != null)
{
dictEntry.setItem(COSName.S, newStructureType);
}
}
}
}
private void mergeIDTree(PDFCloneUtility cloner,
PDStructureTreeRoot srcStructTree,
PDStructureTreeRoot destStructTree) throws IOException
{
PDNameTreeNode<PDStructureElement> srcIDTree = srcStructTree.getIDTree();
if (srcIDTree == null)
{
return;
}
PDNameTreeNode<PDStructureElement> destIDTree = destStructTree.getIDTree();
if (destIDTree == null)
{
destIDTree = new PDStructureElementNameTreeNode();
}
Map<String, PDStructureElement> srcNames = getIDTreeAsMap(srcIDTree);
Map<String, PDStructureElement> destNames = getIDTreeAsMap(destIDTree);
for (Map.Entry<String, PDStructureElement> entry : srcNames.entrySet())
{
if (destNames.containsKey(entry.getKey()))
{
LOG.warn("key " + entry.getKey() + " already exists in destination IDTree");
}
else
{
destNames.put(entry.getKey(),
new PDStructureElement((COSDictionary) cloner.cloneForNewDocument(entry.getValue().getCOSObject())));
}
}
destIDTree = new PDStructureElementNameTreeNode();
destIDTree.setNames(destNames);
destStructTree.setIDTree(destIDTree);
// Note that all elements are stored flatly. This could become a problem for large files
// when these are opened in a viewer that uses the tagging information.
// If this happens, then PDNameTreeNode should be improved with a convenience method that
// stores the map into a B+Tree, see https://en.wikipedia.org/wiki/B+_tree
}
// PDNameTreeNode.getNames() only brings one level, this is why we need this
// might be made public at a later time, or integrated into PDNameTreeNode with template.
static Map<String, PDStructureElement> getIDTreeAsMap(PDNameTreeNode<PDStructureElement> idTree)
throws IOException
{
Map<String, PDStructureElement> names = idTree.getNames();
if (names == null)
{
names = new LinkedHashMap<>();
}
else
{
// must copy because the map is read only
names = new LinkedHashMap<>(names);
}
List<PDNameTreeNode<PDStructureElement>> kids = idTree.getKids();
if (kids != null)
{
for (PDNameTreeNode<PDStructureElement> kid : kids)
{
names.putAll(getIDTreeAsMap(kid));
}
}
return names;
}
// PDNumberTreeNode.getNumbers() only brings one level, this is why we need this
// might be made public at a later time, or integrated into PDNumberTreeNode.
static Map<Integer, COSObjectable> getNumberTreeAsMap(PDNumberTreeNode tree)
throws IOException
{
Map<Integer, COSObjectable> numbers = tree.getNumbers();
if (numbers == null)
{
numbers = new LinkedHashMap<>();
}
else
{
// must copy because the map is read only
numbers = new LinkedHashMap<>(numbers);
}
List<PDNumberTreeNode> kids = tree.getKids();
if (kids != null)
{
for (PDNumberTreeNode kid : kids)
{
numbers.putAll(getNumberTreeAsMap(kid));
}
}
return numbers;
}
private void mergeRoleMap(PDStructureTreeRoot srcStructTree, PDStructureTreeRoot destStructTree)
{
COSDictionary srcDict = srcStructTree.getCOSObject().getCOSDictionary(COSName.ROLE_MAP);
COSDictionary destDict = destStructTree.getCOSObject().getCOSDictionary(COSName.ROLE_MAP);
if (srcDict == null)
{
return;
}
if (destDict == null)
{
destStructTree.getCOSObject().setItem(COSName.ROLE_MAP, srcDict); // clone not needed
return;
}
for (Map.Entry<COSName, COSBase> entry : srcDict.entrySet())
{
COSBase destValue = destDict.getDictionaryObject(entry.getKey());
if (destValue != null && destValue.equals(entry.getValue()))
{
// already exists, but identical
continue;
}
if (destDict.containsKey(entry.getKey()))
{
LOG.warn("key " + entry.getKey() + " already exists in destination RoleMap");
}
else
{
destDict.setItem(entry.getKey(), entry.getValue());
}
}
}
private void mergeAcroForm(PDFCloneUtility cloner, PDDocumentCatalog destCatalog,
PDDocumentCatalog srcCatalog ) throws IOException
{
try
{
PDAcroForm destAcroForm = destCatalog.getAcroForm();
PDAcroForm srcAcroForm = srcCatalog.getAcroForm();
if (destAcroForm == null && srcAcroForm != null)
{
destCatalog.getCOSObject().setItem(COSName.ACRO_FORM,
cloner.cloneForNewDocument(srcAcroForm.getCOSObject()));
}
else
{
if (srcAcroForm != null)
{
if (acroFormMergeMode == AcroFormMergeMode.PDFBOX_LEGACY_MODE)
{
acroFormLegacyMode(cloner, destAcroForm, srcAcroForm);
}
else if (acroFormMergeMode == AcroFormMergeMode.JOIN_FORM_FIELDS_MODE)
{
acroFormJoinFieldsMode(cloner, destAcroForm, srcAcroForm);
}
}
}
}
catch (IOException e)
{
// if we are not ignoring exceptions, we'll re-throw this
if (!ignoreAcroFormErrors)
{
throw new IOException(e);
}
}
}
/*
* Merge the contents of the source form into the destination form for the
* destination file.
*
* @param cloner the object cloner for the destination document
* @param destAcroForm the destination form
* @param srcAcroForm the source form
* @throws IOException If an error occurs while adding the field.
*/
private void acroFormJoinFieldsMode(PDFCloneUtility cloner, PDAcroForm destAcroForm, PDAcroForm srcAcroForm)
throws IOException
{
acroFormLegacyMode(cloner, destAcroForm, srcAcroForm);
}
/*
* Merge the contents of the source form into the destination form for the
* destination file.
*
* @param cloner the object cloner for the destination document
* @param destAcroForm the destination form
* @param srcAcroForm the source form
* @throws IOException If an error occurs while adding the field.
*/
private void acroFormLegacyMode(PDFCloneUtility cloner, PDAcroForm destAcroForm, PDAcroForm srcAcroForm)
throws IOException
{
List<PDField> srcFields = srcAcroForm.getFields();
COSArray destFields;
if (srcFields != null && !srcFields.isEmpty())
{
// if a form is merged multiple times using PDFBox the newly generated
// fields starting with dummyFieldName may already exist. We need to determine the last unique
// number used and increment that.
final String prefix = "dummyFieldName";
final int prefixLength = prefix.length();
for (PDField destField : destAcroForm.getFieldTree())
{
String fieldName = destField.getPartialName();
if (fieldName.startsWith(prefix))
{
String suffix = fieldName.substring(prefixLength);
if (suffix.matches("\\d+"))
{
nextFieldNum = Math.max(nextFieldNum, Integer.parseInt(suffix) + 1);
}
}
}
// get the destinations root fields. Could be that the entry doesn't exist
// or is of wrong type
COSBase base = destAcroForm.getCOSObject().getItem(COSName.FIELDS);
if (base instanceof COSArray)
{
destFields = (COSArray) base;
}
else
{
destFields = new COSArray();
}
for (PDField srcField : srcAcroForm.getFields())
{
COSDictionary dstField = (COSDictionary) cloner.cloneForNewDocument(srcField.getCOSObject());
// if the form already has a field with this name then we need to rename this field
// to prevent merge conflicts.
if (destAcroForm.getField(srcField.getFullyQualifiedName()) != null)
{
dstField.setString(COSName.T, prefix + nextFieldNum++);
}
destFields.add(dstField);
}
destAcroForm.getCOSObject().setItem(COSName.FIELDS,destFields);
}
}
// copy outputIntents to destination, but avoid duplicate OutputConditionIdentifier,
// except when it is missing or is named "Custom".
private void mergeOutputIntents(PDFCloneUtility cloner,
PDDocumentCatalog srcCatalog, PDDocumentCatalog destCatalog) throws IOException
{
List<PDOutputIntent> srcOutputIntents = srcCatalog.getOutputIntents();
List<PDOutputIntent> dstOutputIntents = destCatalog.getOutputIntents();
for (PDOutputIntent srcOI : srcOutputIntents)
{
String srcOCI = srcOI.getOutputConditionIdentifier();
if (srcOCI != null && !"Custom".equals(srcOCI))
{
// is that identifier already there?
boolean skip = false;
for (PDOutputIntent dstOI : dstOutputIntents)
{
if (dstOI.getOutputConditionIdentifier().equals(srcOCI))
{
skip = true;
break;
}
}
if (skip)
{
continue;
}
}
destCatalog.addOutputIntent(new PDOutputIntent((COSDictionary) cloner.cloneForNewDocument(srcOI)));
dstOutputIntents.add(srcOI);
}
}
private int nextFieldNum = 1;
/**
* Indicates if acroform errors are ignored or not.
*
* @return true if acroform errors are ignored
*/
public boolean isIgnoreAcroFormErrors()
{
return ignoreAcroFormErrors;
}
/**
* Set to true to ignore acroform errors.
*
* @param ignoreAcroFormErrorsValue true if acroform errors should be
* ignored
*/
public void setIgnoreAcroFormErrors(boolean ignoreAcroFormErrorsValue)
{
ignoreAcroFormErrors = ignoreAcroFormErrorsValue;
}
/**
* Update the Pg and Obj references to the new (merged) page.
*/
private void updatePageReferences(PDFCloneUtility cloner,
Map<Integer, COSObjectable> numberTreeAsMap,
Map<COSDictionary, COSDictionary> objMapping) throws IOException
{
for (COSObjectable obj : numberTreeAsMap.values())
{
if (obj == null)
{
continue;
}
PDParentTreeValue val = (PDParentTreeValue) obj;
COSBase base = val.getCOSObject();
if (base instanceof COSArray)
{
updatePageReferences(cloner, (COSArray) base, objMapping);
}
else
{
updatePageReferences(cloner, (COSDictionary) base, objMapping);
}
}
}
/**
* Update the Pg and Obj references to the new (merged) page.
*
* @param parentTreeEntry
* @param objMapping mapping between old and new references
*/
private void updatePageReferences(PDFCloneUtility cloner,
COSDictionary parentTreeEntry, Map<COSDictionary, COSDictionary> objMapping)
throws IOException
{
COSDictionary pageDict = parentTreeEntry.getCOSDictionary(COSName.PG);
if (objMapping.containsKey(pageDict))
{
parentTreeEntry.setItem(COSName.PG, objMapping.get(pageDict));
}
COSDictionary objDict = parentTreeEntry.getCOSDictionary(COSName.OBJ);
if (objDict != null)
{
if (objMapping.containsKey(objDict))
{
parentTreeEntry.setItem(COSName.OBJ, objMapping.get(objDict));
}
else
{
// PDFBOX-3999: clone objects that are not in mapping to make sure that
// these don't remain attached to the source document
COSBase item = parentTreeEntry.getItem(COSName.OBJ);
if (item instanceof COSObject)
{
LOG.debug("clone potential orphan object in structure tree: " + item +
", Type: " + objDict.getNameAsString(COSName.TYPE) +
", Subtype: " + objDict.getNameAsString(COSName.SUBTYPE) +
", T: " + objDict.getNameAsString(COSName.T));
}
else
{
// don't display in full because of stack overflow
LOG.debug("clone potential orphan object in structure tree" +
", Type: " + objDict.getNameAsString(COSName.TYPE) +
", Subtype: " + objDict.getNameAsString(COSName.SUBTYPE) +
", T: " + objDict.getNameAsString(COSName.T));
}
parentTreeEntry.setItem(COSName.OBJ, cloner.cloneForNewDocument(objDict));
}
}
COSBase kSubEntry = parentTreeEntry.getDictionaryObject(COSName.K);
if (kSubEntry instanceof COSArray)
{
updatePageReferences(cloner, (COSArray) kSubEntry, objMapping);
}
else if (kSubEntry instanceof COSDictionary)
{
updatePageReferences(cloner, (COSDictionary) kSubEntry, objMapping);
}
}
private void updatePageReferences(PDFCloneUtility cloner,
COSArray parentTreeEntry, Map<COSDictionary, COSDictionary> objMapping)
throws IOException
{
for (int i = 0; i < parentTreeEntry.size(); i++)
{
COSBase subEntry = parentTreeEntry.getObject(i);
if (subEntry instanceof COSArray)
{
updatePageReferences(cloner, (COSArray) subEntry, objMapping);
}
else if (subEntry instanceof COSDictionary)
{
updatePageReferences(cloner, (COSDictionary) subEntry, objMapping);
}
}
}
/**
* Update the StructParents and StructParent values in a PDPage.
*
* @param page the new page
* @param structParentOffset the offset which should be applied
*/
private void updateStructParentEntries(PDPage page, int structParentOffset) throws IOException
{
int structParents = page.getStructParents();
if (structParents >= 0)
{
page.setStructParents(structParents + structParentOffset);
}
List<PDAnnotation> annots = page.getAnnotations();
List<PDAnnotation> newannots = new ArrayList<>(annots.size());
annots.forEach(annot ->
{
int structParent = annot.getStructParent();
if (structParent >= 0)
{
annot.setStructParent(structParent + structParentOffset);
}
newannots.add(annot);
});
page.setAnnotations(newannots);
}
/**
* Test for dynamic XFA content.
*
* @param acroForm the AcroForm
* @return true if there is a dynamic XFA form.
*/
private boolean isDynamicXfa(PDAcroForm acroForm)
{
return acroForm != null && acroForm.xfaIsDynamic();
}
/**
* This will add all of the dictionaries keys/values to this dictionary, but
* only if they are not in an exclusion list and if they don't already
* exist. If a key already exists in this dictionary then nothing is
* changed.
*
* @param src The source dictionary to get the keys/values from.
* @param dst The destination dictionary to merge the keys/values into.
* @param exclude Names of keys that shall be skipped.
*/
private void mergeInto(COSDictionary src, COSDictionary dst, Set<COSName> exclude)
{
for (Map.Entry<COSName, COSBase> entry : src.entrySet())
{
if (!exclude.contains(entry.getKey()) && !dst.containsKey(entry.getKey()))
{
dst.setItem(entry.getKey(), entry.getValue());
}
}
}
}