From 535a04fa05f739ec16dd81666e3b0f82dfbd442d Mon Sep 17 00:00:00 2001 From: tknall Date: Wed, 9 Jan 2013 15:41:29 +0000 Subject: pdf-as-lib maven project files moved to pdf-as-lib git-svn-id: https://joinup.ec.europa.eu/svn/pdf-as/pdf-as/trunk@926 7b5415b0-85f9-ee4d-85bd-d5d0c3b42d1c --- .../knowcenter/wag/egov/egiz/pdf/PDFUtilities.java | 148 +++++++++++++++++++++ 1 file changed, 148 insertions(+) create mode 100644 pdf-as-lib/src/main/java/at/knowcenter/wag/egov/egiz/pdf/PDFUtilities.java (limited to 'pdf-as-lib/src/main/java/at/knowcenter/wag/egov/egiz/pdf/PDFUtilities.java') diff --git a/pdf-as-lib/src/main/java/at/knowcenter/wag/egov/egiz/pdf/PDFUtilities.java b/pdf-as-lib/src/main/java/at/knowcenter/wag/egov/egiz/pdf/PDFUtilities.java new file mode 100644 index 0000000..e2a3d06 --- /dev/null +++ b/pdf-as-lib/src/main/java/at/knowcenter/wag/egov/egiz/pdf/PDFUtilities.java @@ -0,0 +1,148 @@ +/** + * Copyright 2006 by Know-Center, Graz, Austria + * PDF-AS has been contracted by the E-Government Innovation Center EGIZ, a + * joint initiative of the Federal Chancellery Austria and Graz University of + * Technology. + * + * Licensed under the EUPL, Version 1.1 or - as soon they will be approved by + * the European Commission - subsequent versions of the EUPL (the "Licence"); + * You may not use this work except in compliance with the Licence. + * You may obtain a copy of the Licence at: + * http://www.osor.eu/eupl/ + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the Licence is distributed on an "AS IS" basis, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the Licence for the specific language governing permissions and + * limitations under the Licence. + * + * This product combines work with different licenses. See the "NOTICE" text + * file for details on the various modules and licenses. + * The "NOTICE" text file is part of the distribution. Any derivative works + * that you distribute must include a readable copy of the "NOTICE" text file. + * + * $Id: PDFUtilities.java,v 1.3 2006/10/31 08:09:33 wprinz Exp $ + */ +package at.knowcenter.wag.egov.egiz.pdf; + +import java.io.ByteArrayInputStream; +import java.io.File; +import java.io.IOException; +import java.util.List; + +import at.gv.egiz.pdfas.framework.input.PdfDataSource; +import at.gv.egiz.pdfas.impl.input.ByteArrayPdfDataSourceImpl; + +import org.pdfbox.pdfparser.PDFParser; +import org.pdfbox.pdmodel.PDDocument; +import org.pdfbox.pdmodel.PDPage; + +import at.knowcenter.wag.egov.egiz.cfg.SettingsReader; +import at.knowcenter.wag.egov.egiz.exceptions.PDFDocumentException; + +import com.lowagie.text.DocumentException; + +/** + * Contains useful helpers for accessing PDF documents. + * + * @author wprinz + * @author mruhmer + */ +public abstract class PDFUtilities +{ + public static float calculatePageLength(final PdfDataSource pdfDataSource,int page ,float footer_line, int pagerotation) throws PDFDocumentException + { + try + { + //ByteArrayInputStream original_bais = new ByteArrayInputStream(pdf); + //byte [] normalized_pdf = TextualSignature.normalizePDF(original_bais); + byte [] normalized_pdf = TextualSignature.normalizePDF(pdfDataSource); + + // PERF: The whole PDF normalization process is costy + ByteArrayInputStream bais = new ByteArrayInputStream(normalized_pdf); + + PDFParser parser = new PDFParser(bais); + File temporary_dir = SettingsReader.getTemporaryDirectory(); + parser.setTempDirectory(temporary_dir); + parser.parse(); + + PDDocument pdfDocument_ = parser.getPDDocument(); + float page_length = calculatePageLength(pdfDocument_,page , footer_line, pagerotation); + pdfDocument_.close(); + return page_length; + } + catch (IOException e) + { + throw new PDFDocumentException(201, e); + } + catch (DocumentException e) + { + throw new PDFDocumentException(201, e); + } + } + public static float calculatePageLength(PDDocument document,int page ,float footer_line, int pagerotation) throws IOException + { + //int last_page_id = document.getNumberOfPages(); + List allPages = document.getDocumentCatalog().getAllPages(); + PDPage pdpage = (PDPage) allPages.get(page); + pdpage.setRotation(pagerotation); + return calculatePageLength(pdpage, footer_line); + } + + /** + * @deprecated + * @param pdf + * @param footer_line + * @return + * @throws PDFDocumentException + */ + public static float calculateLastPageLength(final byte[] pdf, float footer_line) throws PDFDocumentException + { + try + { + //ByteArrayInputStream original_bais = new ByteArrayInputStream(pdf); + PdfDataSource dataSource = new ByteArrayPdfDataSourceImpl(pdf); + byte [] normalized_pdf = TextualSignature.normalizePDF(dataSource); + + ByteArrayInputStream bais = new ByteArrayInputStream(normalized_pdf); + + PDFParser parser = new PDFParser(bais); + File temporary_dir = SettingsReader.getTemporaryDirectory(); + parser.setTempDirectory(temporary_dir); + parser.parse(); + + PDDocument pdfDocument_ = parser.getPDDocument(); + float last_page_length = calculateLastPageLength(pdfDocument_, footer_line); + pdfDocument_.close(); + + return last_page_length; + } + catch (IOException e) + { + throw new PDFDocumentException(201, e); + } + catch (DocumentException e) + { + throw new PDFDocumentException(201, e); + } + } + + public static float calculateLastPageLength(PDDocument document, float footer_line) throws IOException + { + int last_page_id = document.getNumberOfPages(); + List allPages = document.getDocumentCatalog().getAllPages(); + PDPage last_page = (PDPage) allPages.get(last_page_id - 1); + + return calculatePageLength(last_page, footer_line); + } + + public static float calculatePageLength(PDPage page, float footer_line) throws IOException + { + // logger_.debug("Last Page id:" + last_page_id); + // PDPage last_page = (PDPage) allPages.get(0); + PDFPage my_page = new PDFPage(footer_line); + my_page.processStream(page, page.findResources(), page.getContents().getStream()); + return my_page.getMaxPageLength(); + } + +} -- cgit v1.2.3