htmloutputdev.cc

将pdf文档转换为高质量的html文档

//========================================================================
//
// HtmlOutputDev.cc
//
// Copyright 1997-2002 Glyph & Cog, LLC
//
// Changed 1999-2000 by G.Ovtcharov
//
// Changed 2002 by Mikhail Kruk
//
//========================================================================

#ifdef __GNUC__
#pragma implementation
#endif

#include <stdio.h>
#include <stdlib.h>
#include <stdarg.h>
#include <stddef.h>
#include <ctype.h>
#include <math.h>
#include "GString.h"
#include "GList.h"
#include "UnicodeMap.h"
#include "gmem.h"
#include "config.h"
#include "Error.h"
#include "GfxState.h"
#include "GlobalParams.h"
#include "HtmlOutputDev.h"
#include "HtmlFonts.h"


int HtmlPage::pgNum=0;
int HtmlOutputDev::imgNum=1;

extern double scale;
extern GBool complexMode;
extern GBool ignore;
extern GBool printCommands;
extern GBool printHtml;
extern GBool noframes;
extern GBool stout;
extern GBool xml;
extern GBool showHidden;
extern GBool noMerge;

static GString* basename(GString* str){
  
  char *p=str->getCString();
  int len=str->getLength();
  for (int i=len-1;i>=0;i--)
    if (*(p+i)==SLASH) 
      return new GString((p+i+1),len-i-1);
  return new GString(str);
}

static GString* Dirname(GString* str){
  
  char *p=str->getCString();
  int len=str->getLength();
  for (int i=len-1;i>=0;i--)
    if (*(p+i)==SLASH) 
      return new GString(p,i+1);
  return new GString();
} 

//------------------------------------------------------------------------
// HtmlString
//------------------------------------------------------------------------

HtmlString::HtmlString(GfxState *state, double fontSize, double _charspace, HtmlFontAccu* fonts) {
  GfxFont *font;
  double x, y;

  state->transform(state->getCurX(), state->getCurY(), &x, &y);
  if ((font = state->getFont())) {
    yMin = y - font->getAscent() * fontSize;
    yMax = y - font->getDescent() * fontSize;
    GfxRGB rgb;
    state->getFillRGB(&rgb);
    GString *name = state->getFont()->getName();
    if (!name) name = HtmlFont::getDefaultFont(); //new GString("default");
   // HtmlFont hfont=HtmlFont(name, static_cast<int>(fontSize-1),_charspace, rgb);
    HtmlFont hfont=HtmlFont(name, static_cast<int>(fontSize-1),0.0, rgb);
    fontpos = fonts->AddFont(hfont);
  } else {
    // this means that the PDF file draws text without a current font,
    // which should never happen
    yMin = y - 0.95 * fontSize;
    yMax = y + 0.35 * fontSize;
    fontpos=0;
  }
  if (yMin == yMax) {
    // this is a sanity check for a case that shouldn't happen -- but
    // if it does happen, we want to avoid dividing by zero later
    yMin = y;
    yMax = y + 1;
  }
  col = 0;
  text = NULL;
  xRight = NULL;
  link = NULL;
  len = size = 0;
  yxNext = NULL;
  xyNext = NULL;
  strSize = 0;
  htext=new GString();
  htext2=new GString();
  dir = textDirUnknown;
}


HtmlString::~HtmlString() {
  delete text;
  delete htext;
  delete htext2;
//  delete strSize;
  gfree(xRight);
}

void HtmlString::addChar(GfxState *state, double x, double y,
			 double dx, double dy, Unicode u) {
  if (dir == textDirUnknown) {
    dir = UnicodeMap::getDirection(u);
  } 

  if (len == size) {
    size += 16;
    text = (Unicode *)grealloc(text, size * sizeof(Unicode));
    xRight = (double *)grealloc(xRight, size * sizeof(double));
  }
  text[len] = u;
  if (len == 0) {
    xMin = x;
  }
  xMax = xRight[len] = x + dx;
  //xMax = xRight[len] = x;
  ++strSize;
//printf("added char: %f %f xright = %f\n", x, dx, x+dx);
  ++len;
}

void HtmlString::endString()
{
  if( dir == textDirRightLeft && len > 1 )
  {
    //printf("will reverse!\n");
    for (int i = 0; i < len / 2; i++)
    {
      Unicode ch = text[i];
      text[i] = text[len - i - 1];
      text[len - i - 1] = ch;
    }
  }
}

//------------------------------------------------------------------------
// HtmlPage
//------------------------------------------------------------------------

HtmlPage::HtmlPage(GBool rawOrder, char *imgExtVal) {
  this->rawOrder = rawOrder;
  curStr = NULL;
  yxStrings = NULL;
  xyStrings = NULL;
  yxCur1 = yxCur2 = NULL;
  fonts=new HtmlFontAccu();
  links=new HtmlLinks();
  pageWidth=0;
  pageHeight=0;
  fontsPageMarker = 0;
  DocName=NULL;
  firstPage = -1;
  imgExt = new GString(imgExtVal);
}

HtmlPage::~HtmlPage() {
  clear();
  if (DocName) delete DocName;
  if (fonts) delete fonts;
  if (links) delete links;
  if (imgExt) delete imgExt;  
}

void HtmlPage::updateFont(GfxState *state) {
  GfxFont *font;
  double *fm;
  char *name;
  int code;
  double w;
  
  // adjust the font size
  fontSize = state->getTransformedFontSize();
  if ((font = state->getFont()) && font->getType() == fontType3) {
    // This is a hack which makes it possible to deal with some Type 3
    // fonts.  The problem is that it's impossible to know what the
    // base coordinate system used in the font is without actually
    // rendering the font.  This code tries to guess by looking at the
    // width of the character 'm' (which breaks if the font is a
    // subset that doesn't contain 'm').
    for (code = 0; code < 256; ++code) {
      if ((name = ((Gfx8BitFont *)font)->getCharName(code)) &&
	  name[0] == 'm' && name[1] == '\0') {
	break;
      }
    }
    if (code < 256) {
      w = ((Gfx8BitFont *)font)->getWidth(code);
      if (w != 0) {
	// 600 is a generic average 'm' width -- yes, this is a hack
	fontSize *= w / 0.6;
      }
    }
    fm = font->getFontMatrix();
    if (fm[0] != 0) {
      fontSize *= fabs(fm[3] / fm[0]);
    }
  }
}

void HtmlPage::beginString(GfxState *state, GString *s) {
  curStr = new HtmlString(state, fontSize,charspace, fonts);
}


void HtmlPage::conv(){
  HtmlString *tmp;

  int linkIndex = 0;
  HtmlFont* h;
  for(tmp=yxStrings;tmp;tmp=tmp->yxNext){
     int pos=tmp->fontpos;
     //  printf("%d\n",pos);
     h=fonts->Get(pos);

     if (tmp->htext) delete tmp->htext; 
     tmp->htext=HtmlFont::simple(h,tmp->text,tmp->len);
     tmp->htext2=HtmlFont::simple(h,tmp->text,tmp->len);

     if (links->inLink(tmp->xMin,tmp->yMin,tmp->xMax,tmp->yMax, linkIndex)){
       tmp->link = links->getLink(linkIndex);
       /*GString *t=tmp->htext;
       tmp->htext=links->getLink(k)->Link(tmp->htext);
       delete t;*/
     }
  }

}


void HtmlPage::addChar(GfxState *state, double x, double y,
		       double dx, double dy, 
			double ox, double oy, Unicode *u, int uLen) {
  double x1, y1, w1, h1, dx2, dy2;
  int n, i, d;
  state->transform(x, y, &x1, &y1);
  n = curStr->len;
  d = 0;
 
  // check that new character is in the same direction as current string
  // and is not too far away from it before adding 
/*  if ((UnicodeMap::getDirection(u[0]) != curStr->dir) || 
     (n > 0 && 
      fabs(x1 - curStr->xRight[n-1]) > 0.1 * (curStr->yMax - curStr->yMin))) {
    endString();
    beginString(state, NULL);
  }*/
  state->textTransformDelta(state->getCharSpace() * state->getHorizScaling(),
			    0, &dx2, &dy2);
  dx -= dx2;
  dy -= dy2;
  state->transformDelta(dx, dy, &w1, &h1);
  if (uLen != 0) {
    w1 /= uLen;
    h1 /= uLen;
  }
/* if (d != 3)
 {
 endString();
 beginString(state, NULL);
 }
*/


  for (i = 0; i < uLen; ++i) 
  {
	if (u[i] == ' ')
        {
	    endString();
	    beginString(state, NULL);
	} else {
	    curStr->addChar(state, x1 + i*w1, y1 + i*h1, w1, h1, u[i]);
        }
   }

/*
  for (i = 0; i < uLen; ++i) {
    curStr->addChar(state, x1 + i*w1, y1 + i*h1, w1, h1, u[i]);
  }
*/
}

void HtmlPage::endString() {
  HtmlString *p1, *p2;
  double h, y1, y2;

  // throw away zero-length strings -- they don't have valid xMin/xMax
  // values, and they're useless anyway
  if (curStr->len == 0) {
    delete curStr;
    curStr = NULL;
    return;
  }

  curStr->endString();

#if 0 //~tmp
  if (curStr->yMax - curStr->yMin > 20) {
    delete curStr;
    curStr = NULL;
    return;
  }
#endif

  // insert string in y-major list
  h = curStr->yMax - curStr->yMin;
  y1 = curStr->yMin + 0.5 * h;
  y2 = curStr->yMin + 0.8 * h;
  if (rawOrder) {
    p1 = yxCur1;
    p2 = NULL;
  } else if ((!yxCur1 ||
              (y1 >= yxCur1->yMin &&
               (y2 >= yxCur1->yMax || curStr->xMax >= yxCur1->xMin))) &&
             (!yxCur2 ||
              (y1 < yxCur2->yMin ||
               (y2 < yxCur2->yMax && curStr->xMax < yxCur2->xMin)))) {
    p1 = yxCur1;
    p2 = yxCur2;
  } else {
    for (p1 = NULL, p2 = yxStrings; p2; p1 = p2, p2 = p2->yxNext) {
      if (y1 < p2->yMin || (y2 < p2->yMax && curStr->xMax < p2->xMin))
        break;
    }
    yxCur2 = p2;
  }
  yxCur1 = curStr;
  if (p1)
    p1->yxNext = curStr;
  else
    yxStrings = curStr;
  curStr->yxNext = p2;
  curStr = NULL;
}

void HtmlPage::coalesce() {
  HtmlString *str1, *str2;
  HtmlFont *hfont1, *hfont2;
  double space, horSpace, vertSpace, vertOverlap;
  GBool addSpace, addLineBreak;
  int n, i;
  double curX, curY, lastX, lastY;
  int sSize = 0;      
  double diff = 0.0;
  double pxSize = 0.0;
  double strSize = 0.0;
  double cspace = 0.0;

#if 0 //~ for debugging
  for (str1 = yxStrings; str1; str1 = str1->yxNext) {
    printf("x=%f..%f  y=%f..%f  size=%2d '",
	   str1->xMin, str1->xMax, str1->yMin, str1->yMax,
	   (int)(str1->yMax - str1->yMin));
    for (i = 0; i < str1->len; ++i) {
      fputc(str1->text[i] & 0xff, stdout);
    }
    printf("'\n");
  }
  printf("\n------------------------------------------------------------\n\n");
#endif
  str1 = yxStrings;

  if( !str1 ) return;

  //----- discard duplicated text (fake boldface, drop shadows)
  if( !complexMode )
  {
	HtmlString *str3;
	GBool found;
  	while (str1)
	{
		double size = str1->yMax - str1->yMin;
		double xLimit = str1->xMin + size * 0.2;
		found = gFalse;
		for (str2 = str1, str3 = str1->yxNext;
			str3 && str3->xMin < xLimit;
			str2 = str3, str3 = str2->yxNext)
		{
			if (str3->len == str1->len &&
				!memcmp(str3->text, str1->text, str1->len * sizeof(Unicode)) &&
				fabs(str3->yMin - str1->yMin) < size * 0.2 &&
				fabs(str3->yMax - str1->yMax) < size * 0.2 &&
				fabs(str3->xMax - str1->xMax) < size * 0.2)
			{
				found = gTrue;
				//printf("found duplicate!\n");
				break;
			}
		}
		if (found)
		{
			str2->xyNext = str3->xyNext;
			str2->yxNext = str3->yxNext;
			delete str3;
		}
		else
		{
			str1 = str1->yxNext;
		}
	}		
  }
  
  str1 = yxStrings;
  
  hfont1 = getFont(str1);

  str1->htext2->append(str1->htext);
  if( str1->getLink() != NULL ) {
    GString *ls = str1->getLink()->getLinkStart();
    str1->htext->insert(0, ls);
    delete ls;
  }
  curX = str1->xMin; curY = str1->yMin;
  lastX = str1->xMin; lastY = str1->yMin;

  while (str1 && (str2 = str1->yxNext)) {
    hfont2 = getFont(str2);
    space = str1->yMax - str1->yMin;
    horSpace = str2->xMin - str1->xMax;
    addLineBreak = !noMerge && (fabs(str1->xMin - str2->xMin) < 0.4);
    vertSpace = str2->yMin - str1->yMax;

//printf("coalesce %d %d %f? ", str1->dir, str2->dir, d);

    if (str2->yMin >= str1->yMin && str2->yMin <= str1->yMax)
    {
	vertOverlap = str1->yMax - str2->yMin;
    } else
    if (str2->yMax >= str1->yMin && str2->yMax <= str1->yMax)
    {
	vertOverlap = str2->yMax - str1->yMin;
    } else
    {
    	vertOverlap = 0;
    } 
    
    if (
	(
	 (
	  (
	   (rawOrder && vertOverlap > 0.5 * space) 
	   ||
	   (!rawOrder && str2->yMin < str1->yMax)
	  ) &&
	  (horSpace > -0.5 * space && horSpace < space)
	 ) ||
       	 (vertSpace >= 0 && vertSpace < 0.5 * space && addLineBreak)
	) &&
 // in complex mode fonts must be the same, in other modes fonts do not metter
	str1->dir == str2->dir // text direction the same
       ) 
    {
     diff = str2->xMax - str1->xMin;

     n = str1->len + str2->len;
     if ((addSpace = horSpace > 0.1 * space)) {
        ++n;
      }
    
      if (addLineBreak) {
        ++n;
      }
  
      str1->size = (n + 15) & ~15;
      str1->text = (Unicode *)grealloc(str1->text,
				       str1->size * sizeof(Unicode));
      str1->xRight = (double *)grealloc(str1->xRight,
					str1->size * sizeof(double));
      if (addSpace) {
		/*  if (addSpace > (xoutRoundLower(hfont1->getSize()/scale)))
		  {
		  	str1->text[str1->len] = 0x20;
			str1->htext->append(" ");
			str1->htext2->append(" ");
			str1->xRight[str1->len] = str2->xMin;
			++str1->len;
			++str1->strSize;
		 } */
  	   	 str1->text[str1->len] = 0x20;
                 str1->htext->append(" ");
                 str1->htext2->append(" ");
                 str1->xRight[str1->len] = str2->xMin;
                 ++str1->len;
                ++str1->strSize;
      }
      if (addLineBreak) {
	  str1->text[str1->len] = '\n';
	  str1->htext->append("<br>");
	  str1->htext2->append(" ");
	  str1->xRight[str1->len] = str2->xMin;
	  ++str1->len;
	  str1->yMin = str2->yMin;
	  str1->yMax = str2->yMax;
	  str1->xMax = str2->xMax;
	  int fontLineSize = hfont1->getLineSize();
	  int curLineSize = (int)(vertSpace + space); 

	  if( curLineSize != fontLineSize )
	  {
	      HtmlFont *newfnt = new HtmlFont(*hfont1);
	      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); 
	  }

      }
      str1->htext2->append(str2->htext2);

      HtmlLink *hlink1 = str1->getLink();
      HtmlLink *hlink2 = str2->getLink();

      GString *fntFix;
      GString *iStr=GString::fromInt(str2->fontpos);     
      fntFix = new GString("</span><span class=\"ft");
      fntFix->append(iStr);
      fntFix->append("\">");
      if (((hlink1 == NULL) && (hlink2 == NULL)) && (hfont1->isEqualIgnoreBold(*hfont2) == gFalse))