htmloutputdev.cc

将pdf文档转换为高质量的html文档

      {
	str1->htext->append(fntFix);
      }
      for (i = 0; i < str2->len; ++i) {
	str1->text[str1->len] = str2->text[i];
	str1->xRight[str1->len] = str2->xRight[i];
	++str1->len;
      }

      if( !hlink1 || !hlink2 || !hlink1->isEqualDest(*hlink2) ) {
	if(hlink1 != NULL )
	  str1->htext->append("</a>");
	if(hlink2 != NULL ) {
	  GString *ls = hlink2->getLinkStart();
	  str1->htext->append(ls);
	  delete ls;
	}
      }

      str1->htext->append(str2->htext);
      sSize = str1->htext2->getLength();      
      pxSize = xoutRoundLower(hfont1->getSize()/scale);
      strSize = (pxSize*(sSize-2));   
      cspace = (diff / strSize);//(strSize-pxSize));
     // we check if the fonts are the same and create a new font to ajust the text
//      double diff = str2->xMin - str1->xMin;
//      printf("%s\n",str1->htext2->getCString());
      // str1 now contains href for link of str2 (if it is defined)
      str1->link = str2->link; 

      HtmlFont *newfnt = new HtmlFont(*hfont1);
      newfnt->setCharSpace(cspace);
      //newfnt->setLineSize(curLineSize);
      str1->fontpos = fonts->AddFont(*newfnt);
      delete newfnt;
      hfont1 = getFont(str1);
      // we have to reget hfont2 because it's location could have
      // changed on resize  GStri;ng *iStr=GString::fromInt(i);
      hfont2 = getFont(str2); 

      hfont1 = hfont2;

      if (str2->xMax > str1->xMax) {
	str1->xMax = str2->xMax;
      }

      if (str2->yMax > str1->yMax) {
	str1->yMax = str2->yMax;
      }

      str1->yxNext = str2->yxNext;

      delete str2;
    } else { 

//     printf("startX = %f, endX = %f, diff = %f, fontsize = %d, pxSize = %f, stringSize = %d, cspace = %f, strSize = %f\n",str1->xMin,str1->xMax,diff,hfont1->getSize(),pxSize,sSize,cspace,strSize);

// keep strings separate
//      printf("no\n"); 
//      if( hfont1->isBold() )
      if(str1->getLink() != NULL )
	str1->htext->append("</a>");  
     
      str1->xMin = curX; str1->yMin = curY; 
      str1 = str2;
      curX = str1->xMin; curY = str1->yMin;
      hfont1 = hfont2;

      if( str1->getLink() != NULL ) {
	GString *ls = str1->getLink()->getLinkStart();
	str1->htext->insert(0, ls);
	delete ls;
      }
    }
  }
  str1->xMin = curX; str1->yMin = curY;

  if(str1->getLink() != NULL )
    str1->htext->append("</a>");

#if 0 //~ for debugging
  for (str1 = yxStrings; str1; str1 = str1->yxNext) {
    printf("x=%3d..%3d  y=%3d..%3d  size=%2d ",
	   (int)str1->xMin, (int)str1->xMax, (int)str1->yMin, (int)str1->yMax,
	   (int)(str1->yMax - str1->yMin));
    printf("'%s'\n", str1->htext->getCString());  
  }
  printf("\n------------------------------------------------------------\n\n");
#endif

}


void HtmlPage::dumpAsXML(FILE* f,int page){  
  fprintf(f, "<page number=\"%d\" position=\"absolute\"", page);
  fprintf(f," top=\"0\" left=\"0\" height=\"%d\" width=\"%d\">\n", pageHeight,pageWidth);
    
  for(int i=fontsPageMarker;i < fonts->size();i++) {
    GString *fontCSStyle = fonts->CSStyle(i);
    fprintf(f,"\t%s\n",fontCSStyle->getCString());
    delete fontCSStyle;
  }
  
  GString *str, *str1;
  for(HtmlString *tmp=yxStrings;tmp;tmp=tmp->yxNext){
    if (tmp->htext){
      str=new GString(tmp->htext);
      fprintf(f,"<text top=\"%d\" left=\"%d\" ",xoutRound(tmp->yMin),xoutRound(tmp->xMin));
      fprintf(f,"width=\"%d\" height=\"%d\" ",xoutRound(tmp->xMax-tmp->xMin),xoutRound(tmp->yMax-tmp->yMin));
      fprintf(f,"font=\"%d\">", tmp->fontpos);
      if (tmp->fontpos!=-1){
	str1=fonts->getCSStyle(tmp->fontpos, str);
      }
      fputs(str1->getCString(),f);
      delete str;
      delete str1;
      fputs("</text>\n",f);
    }
  }
  fputs("</page>\n",f);
}


void HtmlPage::dumpComplex(FILE *file, int page){
  FILE* pageFile;
  GString* tmp;
  char* htmlEncoding;

  if( firstPage == -1 ) firstPage = page; 
  
  if( !noframes )
  {
      GString* pgNum=GString::fromInt(page);
      tmp = new GString(DocName);
      tmp->append('-')->append(pgNum)->append(".html");
      delete pgNum;
  
      if (!(pageFile = fopen(tmp->getCString(), "w"))) {
	  error(-1, "Couldn't open html file '%s'", tmp->getCString());
	  delete tmp;
	  return;
      } 
      delete tmp;

      fprintf(pageFile,"%s\n<HTML>\n<HEAD>\n<TITLE>Page %d</TITLE>\n\n",
	      DOCTYPE, page);

      htmlEncoding = HtmlOutputDev::mapEncodingToHtml
	  (globalParams->getTextEncodingName());
      fprintf(pageFile, "<META http-equiv=\"Content-Type\" content=\"text/html; charset=%s\">\n", htmlEncoding);
  }
  else 
  {
      pageFile = file;
      fprintf(pageFile,"<!-- Page %d -->\n", page);
      fprintf(pageFile,"<a name=\"%d\"></a>\n", page);
  } 
  
  fprintf(pageFile,"<DIV style=\"position:relative;width:%d;height:%d;\">\n",
	pageWidth, pageHeight);

  tmp=basename(DocName);
   
  fputs("<STYLE type=\"text/css\">\n<!--\n",pageFile);
  for(int i=fontsPageMarker;i!=fonts->size();i++) {
    GString *fontCSStyle = fonts->CSStyle(i);
    fprintf(pageFile,"\t%s\n",fontCSStyle->getCString());
    delete fontCSStyle;
  }
 
  fputs("-->\n</STYLE>\n",pageFile);
  
  if( !noframes )
  {  
      fputs("</HEAD>\n<BODY bgcolor=\"#A0A0A0\" vlink=\"blue\" link=\"blue\">\n",pageFile); 
  }
  
  if( !ignore ) 
  {
    fprintf(pageFile,
	    "<IMG width=\"%d\" height=\"%d\" src=\"%s%03d.%s\" alt=\"background image\">\n",
	    pageWidth, pageHeight, tmp->getCString(), 
		(page-firstPage+1), imgExt->getCString());
  }
  
  delete tmp;
  
  GString *str, *str1;
  for(HtmlString *tmp1=yxStrings;tmp1;tmp1=tmp1->yxNext){
    if (tmp1->htext){
      str=new GString(tmp1->htext);
      fprintf(pageFile,
	      "<DIV style=\"position:absolute;top:%d;left:%d\">",
	      xoutRound(tmp1->yMin),
	      xoutRound(tmp1->xMin));
      fputs("<nobr>",pageFile); 
      if (tmp1->fontpos!=-1){
	str1=fonts->getCSStyle(tmp1->fontpos, str);  
      }
      //printf("%s\n", str1->getCString());
      fputs(str1->getCString(),pageFile);
      
      delete str;      
      delete str1;
      fputs("</nobr></DIV>\n",pageFile);
    }
  }

  fputs("</DIV>\n", pageFile);
  
  if( !noframes )
  {
      fputs("</BODY>\n</HTML>\n",pageFile);
      fclose(pageFile);
  }
}


void HtmlPage::dump(FILE *f, int pageNum) 
{
  if (complexMode)
  {
    if (xml) dumpAsXML(f, pageNum);
    if (!xml) dumpComplex(f, pageNum);  
  }
  else
  {
    fprintf(f,"<A name=%d></a>",pageNum);
    GString* fName=basename(DocName); 
    for (int i=1;i<HtmlOutputDev::imgNum;i++)
      fprintf(f,"<IMG src=\"%s-%d_%d.jpg\"><br>\n",fName->getCString(),pageNum,i);
    HtmlOutputDev::imgNum=1;
    delete fName;

    GString* str;
    for(HtmlString *tmp=yxStrings;tmp;tmp=tmp->yxNext){
      if (tmp->htext){
		str=new GString(tmp->htext); 
		fputs(str->getCString(),f);
		delete str;      
		fputs("<br>\n",f);  
      }
    }
	fputs("<hr>\n",f);  
  }
}



void HtmlPage::clear() {
  HtmlString *p1, *p2;

  if (curStr) {
    delete curStr;
    curStr = NULL;
  }
  for (p1 = yxStrings; p1; p1 = p2) {
    p2 = p1->yxNext;
    delete p1;
  }
  yxStrings = NULL;
  xyStrings = NULL;
  yxCur1 = yxCur2 = NULL;

  if( !noframes )
  {
      delete fonts;
      fonts=new HtmlFontAccu();
      fontsPageMarker = 0;
  }
  else
  {
      fontsPageMarker = fonts->size();
  }

  delete links;
  links=new HtmlLinks();
 

}

void HtmlPage::setDocName(char *fname){
  DocName=new GString(fname);
}

void HtmlPage::updateCharSpace(GfxState *state)
{
	charspace = state->getCharSpace();
}

//------------------------------------------------------------------------
// HtmlMetaVar
//------------------------------------------------------------------------

HtmlMetaVar::HtmlMetaVar(char *_name, char *_content)
{
    name = new GString(_name);
    content = new GString(_content);
}

HtmlMetaVar::~HtmlMetaVar()
{
   delete name;
   delete content;
} 
    
GString* HtmlMetaVar::toString()	
{
    GString *result = new GString("<META name=\"");
    result->append(name);
    result->append("\" content=\"");
    result->append(content);
    result->append("\">"); 
    return result;
}

//------------------------------------------------------------------------
// HtmlOutputDev
//------------------------------------------------------------------------

static char* HtmlEncodings[][2] = {
    {"Latin1", "ISO-8859-1"},
    {NULL, NULL}
};


char* HtmlOutputDev::mapEncodingToHtml(GString* encoding)
{
    char* enc = encoding->getCString();
    for(int i = 0; HtmlEncodings[i][0] != NULL; i++)
    {
	if( strcmp(enc, HtmlEncodings[i][0]) == 0 )
	{
	    return HtmlEncodings[i][1];
	}
    }
    return enc; 
}

void HtmlOutputDev::doFrame(int firstPage){
  GString* fName=new GString(Docname);
  char* htmlEncoding;
  fName->append(".html");

  if (!(fContentsFrame = fopen(fName->getCString(), "w"))){
    delete fName;
    error(-1, "Couldn't open html file '%s'", fName->getCString());
    return;
  }
  
  delete fName;
    
  fName=basename(Docname);
  fputs(DOCTYPE_FRAMES, fContentsFrame);
  fputs("\n<HTML>",fContentsFrame);
  fputs("\n<HEAD>",fContentsFrame);
  fprintf(fContentsFrame,"\n<TITLE>%s</TITLE>",docTitle->getCString());
  htmlEncoding = mapEncodingToHtml(globalParams->getTextEncodingName());
  fprintf(fContentsFrame, "\n<META http-equiv=\"Content-Type\" content=\"text/html; charset=%s\">\n", htmlEncoding);
  dumpMetaVars(fContentsFrame);
  fprintf(fContentsFrame, "</HEAD>\n");
  fputs("<FRAMESET cols=\"100,*\">\n",fContentsFrame);
  fprintf(fContentsFrame,"<FRAME name=\"links\" src=\"%s_ind.html\">\n",fName->getCString());
  fputs("<FRAME name=\"contents\" src=",fContentsFrame); 
  if (complexMode) 
      fprintf(fContentsFrame,"\"%s-%d.html\"",fName->getCString(), firstPage);
  else
      fprintf(fContentsFrame,"\"%ss.html\"",fName->getCString());
  
  fputs(">\n</FRAMESET>\n</HTML>\n",fContentsFrame);
 
  delete fName;
  fclose(fContentsFrame);  
}

HtmlOutputDev::HtmlOutputDev(char *fileName, char *title, 
	char *author, char *keywords, char *subject, char *date,
	char *extension,
	GBool rawOrder, int firstPage, GBool outline) 
{
  char *htmlEncoding;
  
  fContentsFrame = NULL;
  docTitle = new GString(title);
  pages = NULL;
  dumpJPEG=gTrue;
  //write = gTrue;
  this->rawOrder = rawOrder;
  this->doOutline = outline;
  ok = gFalse;
  imgNum=1;
  //this->firstPage = firstPage;
  //pageNum=firstPage;
  // open file
  needClose = gFalse;
  pages = new HtmlPage(rawOrder, extension);
  
  glMetaVars = new GList();
  glMetaVars->append(new HtmlMetaVar("generator", "pdftohtml 0.40"));  
  if( author ) glMetaVars->append(new HtmlMetaVar("author", author));  
  if( keywords ) glMetaVars->append(new HtmlMetaVar("keywords", keywords));  
  if( date ) glMetaVars->append(new HtmlMetaVar("date", date));  
  if( subject ) glMetaVars->append(new HtmlMetaVar("subject", subject));
 
  maxPageWidth = 0;
  maxPageHeight = 0;

  pages->setDocName(fileName);
  Docname=new GString (fileName);

  // for non-xml output (complex or simple) with frames generate the left frame
  if(!xml && !noframes)
  {
     GString* left=new GString(fileName);
     left->append("_ind.html");

     doFrame(firstPage);
   
     if (!(fContentsFrame = fopen(left->getCString(), "w")))
	 {
        error(-1, "Couldn't open html file '%s'", left->getCString());
		delete left;
        return;
     }
     delete left;
     fputs(DOCTYPE, fContentsFrame);
     fputs("<HTML>\n<HEAD>\n<TITLE></TITLE>\n</HEAD>\n<BODY>\n",fContentsFrame);
     
  	if (doOutline)
	{
		GString *str = basename(Docname);
		fprintf(fContentsFrame, "<A href=\"%s%s\" target=\"contents\">Outline</a><br>", str->getCString(), complexMode ? "-outline.html" : "s.html#outline");
		delete str;
	}
  	
	if (!complexMode)
	{	/* not in complex mode */
		
       GString* right=new GString(fileName);
       right->append("s.html");

       if (!(page=fopen(right->getCString(),"w"))){
        error(-1, "Couldn't open html file '%s'", right->getCString());
        delete right;
		return;
       }
       delete right;
       fputs(DOCTYPE, page);
       fputs("<HTML>\n<HEAD>\n<TITLE></TITLE>\n</HEAD>\n<BODY>\n",page);
     }
  }

  if (noframes) {
    if (stout) page=stdout;
    else {
      GString* right=new GString(fileName);
      if (!xml) right->append(".html");
      if (xml) right->append(".xml");
      if (!(page=fopen(right->getCString(),"w"))){
	delete right;
	error(-1, "Couldn't open html file '%s'", right->getCString());
	return;
      }  
      delete right;
    }

    htmlEncoding = mapEncodingToHtml(globalParams->getTextEncodingName()); 
    if (xml) 
    {
      fprintf(page, "<?xml version=\"1.0\" encoding=\"%s\"?>\n", htmlEncoding);
      fputs("<!DOCTYPE pdf2xml SYSTEM \"pdf2xml.dtd\">\n\n", page);
      fputs("<pdf2xml>\n",page);
    } 
    else 
    {
      fprintf(page,"%s\n<HTML>\n<HEAD>\n<TITLE>%s</TITLE>\n",
	      DOCTYPE, docTitle->getCString());
      
      fprintf(page, "<META http-equiv=\"Content-Type\" content=\"text/html; charset=%s\">\n", htmlEncoding);
      
      dumpMetaVars(page);
      fprintf(page,"</HEAD>\n");
      fprintf(page,"<BODY bgcolor=\"#A0A0A0\" vlink=\"blue\" link=\"blue\">\n");
    }
  }
  ok = gTrue; 
}

HtmlOutputDev::~HtmlOutputDev() {
  /*if (mode&&!xml){
    int h=xoutRound(pages->pageHeight/scale);
    int w=xoutRound(pages->pageWidth/scale);
    fprintf(tin,"%s=%03d\n","PAPER_WIDTH",w);
    fprintf(tin,"%s=%03d\n","PAPER_HEIGHT",h);
    fclose(tin);
    }*/

    HtmlFont::clear(); 
    
    delete Docname;
    delete docTitle;

    deleteGList(glMetaVars, HtmlMetaVar);

    if (fContentsFrame){
      fputs("</BODY>\n</HTML>\n",fContentsFrame);  
      fclose(fContentsFrame);
    }
    if (xml) {
      fputs("</pdf2xml>\n",page);  
      fclose(page);
    } else
    if ( !complexMode || xml || noframes )
    { 
      fputs("</BODY>\n</HTML>\n",page);  
      fclose(page);
    }
    if (pages)
      delete pages;
}



void HtmlOutputDev::startPage(int pageNum, GfxState *state) {
  /*if (mode&&!xml){
    if (write){
      write=gFalse;
      GString* fname=Dirname(Docname);
      fname->append("image.log");
      if((tin=fopen(fname->getCString(),"w"))==NULL){
	printf("Error : can not open %s",fname);
	exit(1);
      }
      delete fname;
    // if(state->getRotation()!=0) 
    //  fprintf(tin,"ROTATE=%d rotate %d neg %d neg translate\n",state->getRotation(),state->getX1(),-state->getY1());
    // else 
      fprintf(tin,"ROTATE=%d neg %d neg translate\n",state->getX1(),state->getY1());  
    }
  }*/

  this->pageNum = pageNum;
  GString *str=basename(Docname);