
org.jpedal.parser.PdfStreamDecoderForSampling Maven / Gradle / Ivy
/*
* ===========================================
* Java Pdf Extraction Decoding Access Library
* ===========================================
*
* Project Info: http://www.idrsolutions.com
* Help section for developers at http://www.idrsolutions.com/support/
*
* (C) Copyright 1997-2015 IDRsolutions and Contributors.
*
* This file is part of JPedal/JPDF2HTML5
*
This library is free software; you can redistribute it and/or
modify it under the terms of the GNU Lesser General Public
License as published by the Free Software Foundation; either
version 2.1 of the License, or (at your option) any later version.
This library 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
Lesser General Public License for more details.
You should have received a copy of the GNU Lesser General Public
License along with this library; if not, write to the Free Software
Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
*
* ---------------
* PdfStreamDecoderForSampling.java
* ---------------
*/
package org.jpedal.parser;
import org.jpedal.external.ExternalHandlers;
import org.jpedal.io.PdfObjectReader;
import org.jpedal.objects.GraphicsState;
import org.jpedal.objects.raw.PdfDictionary;
import org.jpedal.objects.raw.PdfObject;
public class PdfStreamDecoderForSampling extends PdfStreamDecoder{
PdfStreamDecoderForSampling(final PdfObjectReader currentPdfFile){
super(currentPdfFile);
}
/**
*
* just scan for DO and CM to get image sizes so we can work out sampling used
*/
public final float decodePageContentForImageSampling(final PdfObject pdfObject) {/* take out min's%%*/
try{
parserOptions.setRenderDirectly(true);
//check switched off
parserOptions.imagesProcessedFully=true;
//reset count
imageCount=0;
gs = new GraphicsState(0,0);/* take out min's%%*/
//get the binary data from the file
final byte[] b_data;
byte[][] pageContents= null;
if(pdfObject !=null){
pageContents= pdfObject.getKeyArray(PdfDictionary.Contents);
isDataValid= pdfObject.streamMayBeCorrupt();
}
if(pdfObject !=null && pageContents==null) {
b_data = currentPdfFile.readStream(pdfObject, true, true, false, false, false, pdfObject.getCacheName(currentPdfFile.getObjectReader()));
} else if(pageStream!=null) {
b_data = pageStream;
} else {
b_data = currentPdfFile.getObjectReader().readPageIntoStream(pdfObject);
}
//if page data found, turn it into a set of commands
//and decode the stream of commands
if (b_data!=null && b_data.length > 0) {
getSamplingOnly=true;
decodeStreamIntoObjects(b_data, false);
}
//flush fonts
cache.resetFonts();
return parserOptions.getSamplingUsed();
}catch(final Error err){
errorTracker.addPageFailureMessage("Problem decoding page " + err);
if (ExternalHandlers.throwMissingCIDError && err.getMessage()!=null && err.getMessage().contains("kochi")) {
throw err;
}
}
return -1;
}
}
© 2015 - 2025 Weber Informatics LLC | Privacy Policy