C++ PDFDoc::getNumPages方法代码示例

	bool load(const char* fileName)
	{
		delete _pdfDoc;
		delete _outputDev;
		delete _nullOutputDev;

		for (int i=0; i<_bmpCache.size(); i++)	delete _bmpCache[i];
		_bmpCache.resize(0);
		for (int i=0; i<_textCache.size(); i++)	delete _textCache[i];
		_textCache.resize(0);

		_pdfDoc= new PDFDoc(new GooString(fileName), NULL, NULL, NULL);
		if (!_pdfDoc->isOk()) {
			printf("error loading pdf");
			return false;
		}

		GBool bitmapTopDown = gTrue;

		SplashColor white;
		white[0]=0xff;
		white[1]=0xff;
		white[2]=0xff;

		// _outputDev = new SplashOutputDev_mod(splashModeRGB8, 4, gFalse, white, bitmapTopDown);
		_outputDev = new SplashOutputDev(splashModeRGB8, 4, gFalse, white, bitmapTopDown);
		if(!_outputDev)
		{
			printf("error loading pdf");
			return false;
		}
#ifdef USE_NULLOUTPUTDEV
		_nullOutputDev=new NullOutputDev();
#endif
#ifdef POPPLER_VERSION
		_outputDev->startDoc(_pdfDoc);
#else
		_outputDev->startDoc(_pdfDoc->getXRef());
#endif

		_bmpCache.resize(_pdfDoc->getNumPages());
		for (int i=0; i<_bmpCache.size(); i++)	_bmpCache[i]=NULL;
		_textCache.resize(_pdfDoc->getNumPages());
		for (int i=0; i<_textCache.size(); i++)	_textCache[i]=NULL;

		return true;
	}

int
pdf_get_rect(char* filename, int page_num, int pdf_box, realrect* box)
	/* return the box converted to TeX points */
{
	GooString*	name = new GooString(filename);
	PDFDoc*		doc = new PDFDoc(name);

	if (!doc) {
		delete name;
		return -1;
	}

	/* if the doc got created, it now owns name, so we mustn't delete it! */

	if (!doc->isOk()) {
		delete doc;
		return -1;
	}

	int			pages = doc->getNumPages();
	if (page_num > pages)
		page_num = pages;
	if (page_num < 0)
		page_num = pages + 1 + page_num;
	if (page_num < 1)
		page_num = 1;

	Page*		page = doc->getCatalog()->getPage(page_num);

	PDFRectangle*	r;
	switch (pdf_box) {
		default:
		case pdfbox_crop:
			r = page->getCropBox();
			break;
		case pdfbox_media:
			r = page->getMediaBox();
			break;
		case pdfbox_bleed:
			r = page->getBleedBox();
			break;
		case pdfbox_trim:
			r = page->getTrimBox();
			break;
		case pdfbox_art:
			r = page->getArtBox();
			break;
	}

	box->x  = 72.27 / 72 * my_fmin(r->x1, r->x2);
	box->y  = 72.27 / 72 * my_fmin(r->y1, r->y2);
	box->wd = 72.27 / 72 * fabs(r->x2 - r->x1);
	box->ht = 72.27 / 72 * fabs(r->y2 - r->y1);

	delete doc;

	return 0;
}

KoFilter::ConversionStatus PdfImport::convert(const QByteArray& from, const QByteArray& to)
{
    debugPdf << "to:" << to << " from:" << from;

    if (from != "application/pdf" || to != "image/svg+xml") {
        return KoFilter::NotImplemented;
    }

    // read config file
    globalParams = new GlobalParams();
    if (! globalParams)
        return KoFilter::NotImplemented;

    GooString * fname = new GooString(QFile::encodeName(m_chain->inputFile()).data());
    PDFDoc * pdfDoc = new PDFDoc(fname, 0, 0, 0);
    if (! pdfDoc) {
        delete globalParams;
        return KoFilter::StupidError;
    }

    if (! pdfDoc->isOk()) {
        delete globalParams;
        delete pdfDoc;
        return KoFilter::StupidError;
    }

    double hDPI = 72.0;
    double vDPI = 72.0;

    int firstPage = 1;
    int lastPage = pdfDoc->getNumPages();

    debugPdf << "converting pages" << firstPage << "-" << lastPage;

    SvgOutputDev * dev = new SvgOutputDev(m_chain->outputFile());
    if (dev->isOk()) {
        int rotate = 0;
        GBool useMediaBox = gTrue;
        GBool crop = gFalse;
        GBool printing = gFalse;
        pdfDoc->displayPages(dev, firstPage, lastPage, hDPI, vDPI, rotate, useMediaBox, crop, printing);
        dev->dumpContent();
    }

    debugPdf << "wrote file to" << m_chain->outputFile();

    delete dev;
    delete pdfDoc;
    delete globalParams;
    globalParams = 0;

    // check for memory leaks
    Object::memCheck(stderr);

    return KoFilter::OK;
}

gboolean
import_pdf(const gchar *filename, DiagramData *dia, DiaContext *ctx, void* user_data)
{
  PDFDoc *doc;
  GooString *fileName = new GooString(filename);
  // no passwords yet
  GooString *ownerPW = NULL;
  GooString *userPW = NULL;
  gboolean ret = FALSE;

  // without this we will get strange crashes (at least with /O2 build)
  globalParams = new GlobalParams();

  doc = PDFDocFactory().createPDFDoc(*fileName, ownerPW, userPW);
  if (!doc->isOk()) {
    dia_context_add_message (ctx, _("PDF document not OK.\n%s"),
			     dia_context_get_filename (ctx));
  } else {
    DiaOutputDev *diaOut = new DiaOutputDev(dia, doc->getNumPages());

    for (int pg = 1; pg <= doc->getNumPages(); ++pg) {
      Page *page = doc->getPage (pg);
      if (!page || !page->isOk())
        continue;
      doc->displayPage(diaOut, pg,
		       72.0, 72.0, /* DPI, scaling elsewhere */
		       0, /* rotate */
		       gTrue, /* useMediaBox */
		       gTrue, /* Crop */
		       gFalse /* printing */
		       );
    }
    delete diaOut;
    ret = TRUE;
  }
  delete doc;
  delete globalParams;
  delete fileName;

  return ret;
}

bool PdfPlug::convert(QString fn)
{
	bool firstPg = true;
	int currentLayer = m_Doc->activeLayer();
	int baseLayer = m_Doc->activeLayer();
	importedColors.clear();
	if(progressDialog)
	{
		progressDialog->setOverallProgress(2);
		progressDialog->setLabel("GI", tr("Generating Items"));
		qApp->processEvents();
	}
	QFile f(fn);
	oldDocItemCount = m_Doc->Items->count();
	if (progressDialog)
	{
		progressDialog->setBusyIndicator("GI");
		qApp->processEvents();
	}

	globalParams = new GlobalParams();
	if (globalParams)
	{
		GooString *fname = new GooString(QFile::encodeName(fn).data());
		globalParams->setErrQuiet(gTrue);
		GBool hasOcg = gFalse;
		QList<OptionalContentGroup*> ocgGroups;
//		globalParams->setPrintCommands(gTrue);
		PDFDoc *pdfDoc = new PDFDoc(fname, 0, 0, 0);
		if (pdfDoc)
		{
			if (pdfDoc->isOk())
			{
				double hDPI = 72.0;
				double vDPI = 72.0;
				int firstPage = 1;
				int lastPage = pdfDoc->getNumPages();
				SlaOutputDev *dev = new SlaOutputDev(m_Doc, &Elements, &importedColors, importerFlags);
				if (dev->isOk())
				{
					OCGs* ocg = pdfDoc->getOptContentConfig();
					if (ocg)
					{
						hasOcg = ocg->hasOCGs();
						if (hasOcg)
						{

							QStringList ocgNames;
							Array *order = ocg->getOrderArray();
							if (order)
							{
								for (int i = 0; i < order->getLength (); ++i)
								{
									Object orderItem;
									order->get(i, &orderItem);
									if (orderItem.isDict())
									{
										Object ref;
										order->getNF(i, &ref);
										if (ref.isRef())
										{
											OptionalContentGroup *oc = ocg->findOcgByRef(ref.getRef());
											QString ocgName = UnicodeParsedString(oc->getName());
											if (!ocgNames.contains(ocgName))
											{
												ocgGroups.prepend(oc);
												ocgNames.append(ocgName);
											}
										}
										ref.free();
									}
									else
									{
										GooList *ocgs;
										int i;
										ocgs = ocg->getOCGs ();
										for (i = 0; i < ocgs->getLength (); ++i)
										{
											OptionalContentGroup *oc = (OptionalContentGroup *)ocgs->get(i);
											QString ocgName = UnicodeParsedString(oc->getName());
											if (!ocgNames.contains(ocgName))
											{
												ocgGroups.prepend(oc);
												ocgNames.append(ocgName);
											}
										}
									}
								}
							}
							else
							{
								GooList *ocgs;
								int i;
								ocgs = ocg->getOCGs ();
								for (i = 0; i < ocgs->getLength (); ++i)
								{
									OptionalContentGroup *oc = (OptionalContentGroup *)ocgs->get(i);
									QString ocgName = UnicodeParsedString(oc->getName());
									if (!ocgNames.contains(ocgName))
									{
//.........这里部分代码省略.........

int main(int argc, char *argv[])
{
  // parse args
  bool ok = parseArgs(argDesc, &argc, argv);
  if (!ok || argc < 2 || argc > 3 || printHelp) {
    fprintf(stderr, "pdftoipe version %s\n", PDFTOIPE_VERSION);
    printUsage("pdftoipe", "<PDF-file> [<XML-file>]", argDesc);
    return 1;
  }

  GooString *fileName = new GooString(argv[1]);

  globalParams = new GlobalParams();
  if (quiet)
    globalParams->setErrQuiet(quiet);

  GooString *ownerPW, *userPW;
  if (ownerPassword[0]) {
    ownerPW = new GooString(ownerPassword);
  } else {
    ownerPW = 0;
  }
  if (userPassword[0]) {
    userPW = new GooString(userPassword);
  } else {
    userPW = 0;
  }

  // open PDF file
  PDFDoc *doc = new PDFDoc(fileName, ownerPW, userPW);
  delete userPW;
  delete ownerPW;

  if (!doc->isOk())
    return 1;
  
  // construct XML file name
  std::string xmlFileName;
  if (argc == 3) {
    xmlFileName = argv[2];
  } else {
    const char *p = fileName->c_str() + fileName->getLength() - 4;
    if (!strcmp(p, ".pdf") || !strcmp(p, ".PDF")) {
        xmlFileName = std::string(fileName->c_str(),
                                  fileName->getLength() - 4);
    } else {
      xmlFileName = fileName->c_str();
    }
    xmlFileName += ".ipe";
  }

  // get page range
  if (firstPage < 1)
    firstPage = 1;

  if (lastPage < 1 || lastPage > doc->getNumPages())
    lastPage = doc->getNumPages();

  // write XML file
  XmlOutputDev *xmlOut = 
    new XmlOutputDev(xmlFileName, doc->getXRef(),
                     doc->getCatalog(), firstPage, lastPage);

  // tell output device about text handling
  xmlOut->setTextHandling(math, notext, literal, mergeLevel, unicodeLevel);
  
  int exitCode = 2;
  if (xmlOut->isOk()) {
    doc->displayPages(xmlOut, firstPage, lastPage, 
		      // double hDPI, double vDPI, int rotate,
		      // bool useMediaBox, bool crop, bool printing,
		      72.0, 72.0, 0, false, false, false);
    exitCode = 0;
  }

  if (xmlOut->hasUnicode()) {
    fprintf(stderr, "The document contains Unicode (non-ASCII) text.\n");
    if (unicodeLevel <= 1)
      fprintf(stderr, "Unknown Unicode characters were replaced by [U+XXX].\n");
    else
      fprintf(stderr, "UTF-8 was set as document encoding in the preamble.\n");
  }

  // clean up
  delete xmlOut;
  delete doc;
  delete globalParams;

  return exitCode;
}

int main(int argc, char *argv[]) {
  PDFDoc *doc;
  GString *fileName;
  GString *ownerPW, *userPW;
  SplashColor paperColor;
  SplashOutputDev *splashOut;
  GBool ok;
  int exitCode;
  int pg;

  exitCode = 99;

  // parse args
  ok = parseArgs(argDesc, &argc, argv);
  
  if (!ok || argc != 2 || printVersion || printHelp) {
    fprintf(stderr, "pdf2jpeg version %s\n", xpdfVersion);
    fprintf(stderr, "%s\n", xpdfCopyright);
    if (!printVersion) {
      printUsage("pdf2jpeg", "<PDF-file> -o <jpegfile>", argDesc);
    }
    goto err0;
  }
  fileName = new GString(argv[1]);

  // read config file
  globalParams = new GlobalParams(cfgFileName);
  globalParams->setupBaseFonts(NULL);
  
  // open PDF file
  if (ownerPassword[0]) {
    ownerPW = new GString(ownerPassword);
  } else {
    ownerPW = NULL;
  }
  if (userPassword[0]) {
    userPW = new GString(userPassword);
  } else {
    userPW = NULL;
  }
  doc = new PDFDoc(fileName, ownerPW, userPW);
  if (userPW) {
    delete userPW;
  }
  if (ownerPW) {
    delete ownerPW;
  }
  if (!doc->isOk()) {
    exitCode = 1;
    goto err1;
  }

  paperColor[0] = paperColor[1] = paperColor[2] = 0xff;
  splashOut = new SplashOutputDev(splashModeRGB8, 1, gFalse, paperColor);
  
  splashOut->startDoc(doc->getXRef());

  if(page>=1 && page<=doc->getNumPages()) {
      double r = resolution;
      if(width) {
	int old_width = doc->getPageCropWidth(page);
	r = 72.0*width/old_width;
      }
      doc->displayPage(splashOut, page, r, r, 0, gFalse, gTrue, gFalse);
      SplashBitmap*bitmap = splashOut->getBitmap();
      if(bitmap) {
	Guchar*rgb = bitmap->getDataPtr();
	int width = bitmap->getWidth();
	int height = bitmap->getHeight();
	jpeg_save(rgb, width, height, quality, output);
      }
  }
  delete splashOut;

  exitCode = 0;

  // clean up
 err1:
  delete doc;
  delete globalParams;
 err0:

  // check for memory leaks
  Object::memCheck(stderr);
  gMemReport(stderr);

  return exitCode;
}

int extract_images_from_pdf(char* filename,
                            char* target,
                            char* owner_password,
                            char* user_password,
                            char* range,
                            char* format,
                            int jpg_quality,
                            GBool dump_jpg,
                            GBool tiff_jpg)
{
    if (user_cancelled)
        return gpret_user_cancelled;

    // load config
    xpdf_rc xrc;

    // open file
    xpdf_doc xdoc(filename, owner_password, user_password);
    PDFDoc* doc = xdoc.get_doc();
    if (!doc->isOk())
        return doc->getErrorCode() == errEncrypted ?
               gpret_pdf_encrypted :
               gpret_cant_open_pdf;

    // check for copy permission
    // if (!doc->okToCopy())
    //	 return gpret_dont_allow_copy;

    // get page range
    page_range range_list(range);
    if (*range == '\0')
    {
        range_list.add_item(range_item(1, doc->getNumPages()));
    }

    if (user_cancelled)
        return gpret_user_cancelled;

    // write image files
    fi_loader fi;
    int progress = 0;
    image_extractor img_out(target, dump_jpg, format, jpg_quality, tiff_jpg);
    for (int i = 0; i < range_list.item_count(); i++)
    {
        range_item& item = range_list.get_item(i);
        for (int pg = item.first;
                pg <= min(item.last, doc->getNumPages());
                pg++)
        {
            if (user_cancelled)
                return gpret_user_cancelled;
            doc->displayPage(&img_out, pg, 72, 72, 0, gFalse, gTrue, gFalse);
            printf("progress: %d\n",
                   ++progress * 100 / range_list.page_count());
        }
    }

    printf("image count: %d\n", img_out.get_image_number());

    return gpret_success;
}

indri::parse::UnparsedDocument* indri::parse::PDFDocumentExtractor::nextDocument() {
  if( !_documentPath.length() )
    return 0;

  PDFDoc* doc = 0;
  TextOutputDev* textOut = 0;
  GString* gfilename = new GString(_documentPath.c_str());
  doc = new PDFDoc( gfilename );
  // if the doc is not ok, or ok to copy, it
  // will be a document of length 0.
  if( doc->isOk() && doc->okToCopy() ) {
    void* stream = &_documentTextBuffer;
    textOut = new TextOutputDev( buffer_write, stream, gFalse, gFalse);
    if ( textOut->isOk() ) {
      int firstPage = 1;
      int lastPage = doc->getNumPages();
	  double hDPI=72.0;
	  double vDPI=72.0;
	  int rotate=0;
	  GBool useMediaBox=gFalse;
	  GBool crop=gTrue; 
	  GBool printing=gFalse; 
	  if(doc->readMetadata()!=NULL)
	  {
		  GString rawMetaData = doc->readMetadata();
		  GString preparedMetaData="";

		  //zoek <rdf:RDF  en eindig bij </rdf:RDF>!! 
		  for(int x=0; x<rawMetaData.getLength(); x++) {
			  if(rawMetaData.getChar(x)!='?' && rawMetaData.getChar(x)!=':') {
				  //skip characters which the XMLReader doesn't understand
				  preparedMetaData.append(rawMetaData.getChar(x));
			  }
		  }
		  std::string metaData(preparedMetaData.getCString());
		  int startbegin = metaData.find("<rdf");
		  int stopend = metaData.find(">", metaData.rfind("</rdf") );
		  metaData = metaData.substr(startbegin, (stopend-startbegin)+1 );
	  

     	  indri::xml::XMLReader reader;

		  try {
			  std::auto_ptr<indri::xml::XMLNode> result( reader.read( metaData.c_str() ) );
			  appendPdfMetaData( result.get() );
		  } catch( lemur::api::Exception& e ) {
			LEMUR_RETHROW( e, "Had trouble reading PDF metadata" );
		  } 
		  if( _author.length()>0 || _title.length()>0 )
		  {
			std::string createdPdfHeader;
			createdPdfHeader="<head>\n";
			if(_title.length()>0) {
				createdPdfHeader+="<title>";
				createdPdfHeader+=_title;
				createdPdfHeader+="</title>\n";
			}
			if(_author.length()>0) {
				createdPdfHeader+="<author>";
				createdPdfHeader+=_author;
				createdPdfHeader+="</author>\n";
			}
			createdPdfHeader+="</head>\n";
			char *metastream = _documentTextBuffer.write( createdPdfHeader.length()+1 );
			strcpy(metastream, createdPdfHeader.c_str());
		  }
	  }
      doc->displayPages(textOut, firstPage, lastPage, hDPI, vDPI, rotate, useMediaBox, crop, printing);
    }
  }
  

  delete textOut;
  delete doc;

  _unparsedDocument.textLength = _documentTextBuffer.position();
  _unparsedDocument.contentLength = _unparsedDocument.textLength ? _documentTextBuffer.position() - 1 : 0 ; // no null 0 if text is empty.
  char* docnoPoint = _documentTextBuffer.write( _documentPath.length()+1 );
  strcpy( docnoPoint, _documentPath.c_str() );
  _unparsedDocument.text = _documentTextBuffer.front();
  _unparsedDocument.content = _documentTextBuffer.front();
  _unparsedDocument.metadata.clear();

  indri::parse::MetadataPair pair;

  pair.key = "path";
  pair.value = docnoPoint;
  pair.valueLength = _documentPath.length()+1;
  _unparsedDocument.metadata.push_back( pair );

  _docnostring.assign(_documentPath.c_str() );
  cleanDocno();
  pair.value = _docnostring.c_str();
  pair.valueLength = _docnostring.length()+1;
  pair.key = "docno";
  _unparsedDocument.metadata.push_back( pair );

  _documentPath = "";

  return &_unparsedDocument;
//.........这里部分代码省略.........