org.sejda.impl.sambox.MergeTask Maven / Gradle / Ivy
Go to download
Show more of this group Show more artifacts with this name
Show all versions of sejda-sambox Show documentation
Show all versions of sejda-sambox Show documentation
Package containing tasks implemented using sambox.
/*
* Created on 03/set/2015
* Copyright 2015 by Andrea Vacondio ([email protected]).
* This file is part of Sejda.
*
* Sejda is free software: you can redistribute it and/or modify
* it under the terms of the GNU Affero General Public License as published by
* the Free Software Foundation, either version 3 of the License, or
* (at your option) any later version.
*
* Sejda is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU Affero General Public License for more details.
*
* You should have received a copy of the GNU Affero General Public License
* along with Sejda. If not, see {
private static final Logger LOG = LoggerFactory.getLogger(MergeTask.class);
private SingleOutputWriter outputWriter;
private PdfSourceOpener sourceOpener;
private int totalSteps;
private PDDocumentHandler destinationDocument;
private Queue toClose = new LinkedList<>();
private OutlineMerger outlineMerger;
private CatalogPageLabelsMerger catalogPageLabelsMerger;
private AcroFormsMerger acroFormsMerger;
private TableOfContentsCreator tocCreator;
private FilenameFooterWriter footerWriter;
private PDRectangle currentPageSize = PDRectangle.A4;
private int pagesCounter = 0;
private int inputsCounter = 0;
private int firstInputNumberOfPages = 0;
@Override
public void before(MergeParameters parameters, TaskExecutionContext executionContext) throws TaskException {
super.before(parameters, executionContext);
totalSteps = parameters.getInputList().size();
sourceOpener = new DefaultPdfSourceOpener();
outputWriter = OutputWriters.newSingleOutputWriter(parameters.getExistingOutputPolicy(), executionContext);
outlineMerger = new OutlineMerger(parameters.getOutlinePolicy());
}
@Override
public void execute(MergeParameters parameters) throws TaskException {
int currentStep = 0;
File tmpFile = IOUtils.createTemporaryBuffer(parameters.getOutput());
outputWriter.taskOutput(tmpFile);
LOG.debug("Temporary output set to {}", tmpFile);
this.destinationDocument = new PDDocumentHandler();
this.destinationDocument.setCreatorOnPDDocument();
this.destinationDocument.setVersionOnPDDocument(parameters.getVersion());
this.destinationDocument.setCompress(parameters.isCompress());
this.acroFormsMerger = new AcroFormsMerger(parameters.getAcroFormPolicy(),
this.destinationDocument.getUnderlyingPDDocument());
this.tocCreator = new TableOfContentsCreator(parameters, this.destinationDocument.getUnderlyingPDDocument());
this.footerWriter = new FilenameFooterWriter(parameters.isFilenameFooter(),
this.destinationDocument.getUnderlyingPDDocument());
this.catalogPageLabelsMerger = new CatalogPageLabelsMerger(parameters.getCatalogPageLabelsPolicy());
ImagesToPdfDocumentConverter.convertImageMergeInputToPdf(parameters, executionContext());
for (PdfMergeInput input : parameters.getPdfInputList()) {
inputsCounter++;
LOG.debug("Opening {}", input.getSource());
PDDocumentHandler sourceDocumentHandler = input.getSource().open(sourceOpener);
toClose.add(sourceDocumentHandler);
if (inputsCounter == 1) {
firstInputNumberOfPages = sourceDocumentHandler.getNumberOfPages();
}
LOG.debug("Adding pages");
LookupTable pagesLookup = new LookupTable<>();
long relativePagesCounter = 0;
Set pagesToImport = input.getPages(sourceDocumentHandler.getNumberOfPages());
for (Integer currentPage : pagesToImport) {
pagesCounter++;
relativePagesCounter++;
try {
PDPage page = sourceDocumentHandler.getPage(currentPage);
// we keep rotation into account
currentPageSize = page.getMediaBox().rotate(page.getRotation());
// we don't use the original page because once added to the new tree we loose inheritable attributes
// so we use a page duplicate to explicitly assign inheritable resources
PDPage importedPage = destinationDocument.importPage(page);
pagesLookup.addLookupEntry(page, importedPage);
// rotate
Rotation rotation = parameters.getRotation(inputsCounter - 1);
if (rotation != Rotation.DEGREES_0) {
PdfRotator.rotate(importedPage, rotation);
}
String sourceBaseName = FilenameUtils.getBaseName(input.getSource().getName());
// processing the first page of the source
if (tocCreator.shouldGenerateToC() && relativePagesCounter == 1) {
if (parameters.isFirstInputCoverTitle() && inputsCounter == 1) {
// skip the cover/title document, don't add it to the ToC
} else {
tocCreator.pageSizeIfNotSet(currentPageSize);
String tocText = sourceBaseName;
if (ToCPolicy.DOC_TITLES == parameters.getTableOfContentsPolicy()) {
tocText = ofNullable(
sourceDocumentHandler.getUnderlyingPDDocument().getDocumentInformation())
.map(i -> i.getTitle()).filter(StringUtils::isNotBlank)
.orElse(sourceBaseName);
}
tocCreator.appendItem(tocText, pagesCounter, importedPage);
}
}
long currentPageNumber = pagesCounter + tocCreator.tocNumberOfPages();
if (parameters.isFirstInputCoverTitle() && inputsCounter == 1) {
// the toc will be added after the cover/title pages
currentPageNumber = pagesCounter;
}
this.footerWriter.addFooter(importedPage, sourceBaseName, currentPageNumber);
LOG.trace("Added imported page");
} catch (PageNotFoundException e) {
executionContext().assertTaskIsLenient(e);
notifyEvent(executionContext().notifiableTaskMetadata())
.taskWarning(String.format("Page %d was skipped, could not be processed", currentPage), e);
}
}
relativePagesCounter = 0;
outlineMerger.updateOutline(sourceDocumentHandler.getUnderlyingPDDocument(), input.getSource().getName(),
pagesLookup);
LookupTable annotationsLookup = new AnnotationsDistiller(
sourceDocumentHandler.getUnderlyingPDDocument()).retainRelevantAnnotations(pagesLookup);
clipSignatures(annotationsLookup.values());
acroFormsMerger.mergeForm(
sourceDocumentHandler.getUnderlyingPDDocument().getDocumentCatalog().getAcroForm(),
annotationsLookup);
if (parameters.isBlankPageIfOdd()) {
ofNullable(destinationDocument.addBlankPageIfOdd(currentPageSize)).ifPresent(p -> pagesCounter++);
}
catalogPageLabelsMerger.add(sourceDocumentHandler.getUnderlyingPDDocument(), pagesToImport);
notifyEvent(executionContext().notifiableTaskMetadata()).stepsCompleted(++currentStep).outOf(totalSteps);
}
if (outlineMerger.hasOutline()) {
LOG.debug("Adding generated outline");
destinationDocument.setDocumentOutline(outlineMerger.getOutline());
}
ofNullable(acroFormsMerger.getForm()).filter(f -> !f.getFields().isEmpty()).ifPresent(f -> {
LOG.debug("Adding generated AcroForm");
destinationDocument.setDocumentAcroForm(f);
});
if (parameters.isNormalizePageSizes()) {
LOG.debug("Normalizing page widths to match width of first page");
// Do this before generating TOC, so the first page is from content.
new PdfScaler(ScaleType.PAGE).scalePages(destinationDocument.getUnderlyingPDDocument());
}
if (tocCreator.hasToc()) {
LOG.debug("Adding generated ToC");
try {
// add ToC as first page or after the cover/title pages
int beforePageNumber = parameters.isFirstInputCoverTitle() ? firstInputNumberOfPages : 0;
tocCreator.addToC(beforePageNumber);
} catch (TaskException e) {
notifyEvent(executionContext().notifiableTaskMetadata())
.taskWarning("Unable to create the Table of Contents", e);
}
}
if (catalogPageLabelsMerger.hasPageLabels()) {
LOG.debug("Adding merged /Catalog /PageLabels");
destinationDocument.getUnderlyingPDDocument().getDocumentCatalog()
.setPageLabels(catalogPageLabelsMerger.getMergedPageLabels());
}
destinationDocument.savePDDocument(tmpFile, parameters.getOutput().getEncryptionAtRestPolicy());
closeResources();
parameters.getOutput().accept(outputWriter);
LOG.debug("Input documents merged correctly and written to {}", parameters.getOutput());
}
private void closeResources() {
Closeable current;
while ((current = toClose.poll()) != null) {
closeQuietly(current);
}
closeQuietly(destinationDocument);
}
@Override
public void after() {
closeResources();
outputWriter = null;
}
}
© 2015 - 2025 Weber Informatics LLC | Privacy Policy