textoutputdev.cc

这是一个做pdf阅读器的源代码文件,是大家学习阅读器资料的很好参考

    delta = xMin - blk->xMax;
    break;
  case 2:
    delta = yMin - blk->yMax;
    break;
  case 3:
    delta = blk->xMin - xMax;
    break;
  }
  return delta;
}

GBool TextBlock::isBelow(TextBlock *blk) {
  GBool below;

  below = gFalse; // make gcc happy
  switch (page->primaryRot) {
  case 0:
    below = xMin >= blk->priMin && xMax <= blk->priMax &&
            yMin > blk->yMin;
    break;
  case 1:
    below = yMin >= blk->priMin && yMax <= blk->priMax &&
            xMax < blk->xMax;
    break;
  case 2:
    below = xMin >= blk->priMin && xMax <= blk->priMax &&
            yMax < blk->yMax;
    break;
  case 3:
    below = yMin >= blk->priMin && yMax <= blk->priMax &&
            xMin > blk->xMin;
    break;
  }

  return below;
}

//------------------------------------------------------------------------
// TextFlow
//------------------------------------------------------------------------

TextFlow::TextFlow(TextPage *pageA, TextBlock *blk) {
  page = pageA;
  xMin = blk->xMin;
  xMax = blk->xMax;
  yMin = blk->yMin;
  yMax = blk->yMax;
  priMin = blk->priMin;
  priMax = blk->priMax;
  blocks = lastBlk = blk;
  next = NULL;
}

TextFlow::~TextFlow() {
  TextBlock *blk;

  while (blocks) {
    blk = blocks;
    blocks = blocks->next;
    delete blk;
  }
}

void TextFlow::addBlock(TextBlock *blk) {
  if (lastBlk) {
    lastBlk->next = blk;
  } else {
    blocks = blk;
  }
  lastBlk = blk;
  if (blk->xMin < xMin) {
    xMin = blk->xMin;
  }
  if (blk->xMax > xMax) {
    xMax = blk->xMax;
  }
  if (blk->yMin < yMin) {
    yMin = blk->yMin;
  }
  if (blk->yMax > yMax) {
    yMax = blk->yMax;
  }
}

GBool TextFlow::blockFits(TextBlock *blk, TextBlock *prevBlk) {
  GBool fits;

  // lower blocks must use smaller fonts
  if (blk->lines->words->fontSize > lastBlk->lines->words->fontSize) {
    return gFalse;
  }

  fits = gFalse; // make gcc happy
  switch (page->primaryRot) {
  case 0:
    fits = blk->xMin >= priMin && blk->xMax <= priMax;
    break;
  case 1:
    fits = blk->yMin >= priMin && blk->yMax <= priMax;
    break;
  case 2:
    fits = blk->xMin >= priMin && blk->xMax <= priMax;
    break;
  case 3:
    fits = blk->yMin >= priMin && blk->yMax <= priMax;
    break;
  }
  return fits;
}

#if TEXTOUT_WORD_LIST

//------------------------------------------------------------------------
// TextWordList
//------------------------------------------------------------------------

TextWordList::TextWordList(TextPage *text, GBool physLayout) {
  TextFlow *flow;
  TextBlock *blk;
  TextLine *line;
  TextWord *word;
  TextWord **wordArray;
  int nWords, i;

  words = new GList();

  if (text->rawOrder) {
    for (word = text->rawWords; word; word = word->next) {
      words->append(word);
    }

  } else if (physLayout) {
    // this is inefficient, but it's also the least useful of these
    // three cases
    nWords = 0;
    for (flow = text->flows; flow; flow = flow->next) {
      for (blk = flow->blocks; blk; blk = blk->next) {
	for (line = blk->lines; line; line = line->next) {
	  for (word = line->words; word; word = word->next) {
	    ++nWords;
	  }
	}
      }
    }
    wordArray = (TextWord **)gmallocn(nWords, sizeof(TextWord *));
    i = 0;
    for (flow = text->flows; flow; flow = flow->next) {
      for (blk = flow->blocks; blk; blk = blk->next) {
	for (line = blk->lines; line; line = line->next) {
	  for (word = line->words; word; word = word->next) {
	    wordArray[i++] = word;
	  }
	}
      }
    }
    qsort(wordArray, nWords, sizeof(TextWord *), &TextWord::cmpYX);
    for (i = 0; i < nWords; ++i) {
      words->append(wordArray[i]);
    }
    gfree(wordArray);

  } else {
    for (flow = text->flows; flow; flow = flow->next) {
      for (blk = flow->blocks; blk; blk = blk->next) {
	for (line = blk->lines; line; line = line->next) {
	  for (word = line->words; word; word = word->next) {
	    words->append(word);
	  }
	}
      }
    }
  }
}

TextWordList::~TextWordList() {
  delete words;
}

int TextWordList::getLength() {
  return words->getLength();
}

TextWord *TextWordList::get(int idx) {
  if (idx < 0 || idx >= words->getLength()) {
    return NULL;
  }
  return (TextWord *)words->get(idx);
}

#endif // TEXTOUT_WORD_LIST

//------------------------------------------------------------------------
// TextPage
//------------------------------------------------------------------------

TextPage::TextPage(GBool rawOrderA) {
  int rot;

  rawOrder = rawOrderA;
  curWord = NULL;
  charPos = 0;
  curFont = NULL;
  curFontSize = 0;
  nest = 0;
  nTinyChars = 0;
  lastCharOverlap = gFalse;
  if (!rawOrder) {
    for (rot = 0; rot < 4; ++rot) {
      pools[rot] = new TextPool();
    }
  }
  flows = NULL;
  blocks = NULL;
  rawWords = NULL;
  rawLastWord = NULL;
  fonts = new GList();
  lastFindXMin = lastFindYMin = 0;
  haveLastFind = gFalse;
  underlines = new GList();
  links = new GList();
}

TextPage::~TextPage() {
  int rot;

  clear();
  if (!rawOrder) {
    for (rot = 0; rot < 4; ++rot) {
      delete pools[rot];
    }
  }
  delete fonts;
  deleteGList(underlines, TextUnderline);
  deleteGList(links, TextLink);
}

void TextPage::startPage(GfxState *state) {
  clear();
  if (state) {
    pageWidth = state->getPageWidth();
    pageHeight = state->getPageHeight();
  } else {
    pageWidth = pageHeight = 0;
  }
}

void TextPage::endPage() {
  if (curWord) {
    endWord();
  }
}

void TextPage::clear() {
  int rot;
  TextFlow *flow;
  TextWord *word;

  if (curWord) {
    delete curWord;
    curWord = NULL;
  }
  if (rawOrder) {
    while (rawWords) {
      word = rawWords;
      rawWords = rawWords->next;
      delete word;
    }
  } else {
    for (rot = 0; rot < 4; ++rot) {
      delete pools[rot];
    }
    while (flows) {
      flow = flows;
      flows = flows->next;
      delete flow;
    }
    gfree(blocks);
  }
  deleteGList(fonts, TextFontInfo);

  curWord = NULL;
  charPos = 0;
  curFont = NULL;
  curFontSize = 0;
  nest = 0;
  nTinyChars = 0;
  if (!rawOrder) {
    for (rot = 0; rot < 4; ++rot) {
      pools[rot] = new TextPool();
    }
  }
  flows = NULL;
  blocks = NULL;
  rawWords = NULL;
  rawLastWord = NULL;
  fonts = new GList();
}

void TextPage::updateFont(GfxState *state) {
  GfxFont *gfxFont;
  double *fm;
  char *name;
  int code, mCode, letterCode, anyCode;
  double w;
  int i;

  // get the font info object
  curFont = NULL;
  for (i = 0; i < fonts->getLength(); ++i) {
    curFont = (TextFontInfo *)fonts->get(i);
    if (curFont->matches(state)) {
      break;
    }
    curFont = NULL;
  }
  if (!curFont) {
    curFont = new TextFontInfo(state);
    fonts->append(curFont);
  }

  // adjust the font size
  gfxFont = state->getFont();
  curFontSize = state->getTransformedFontSize();
  if (gfxFont && gfxFont->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').
    mCode = letterCode = anyCode = -1;
    for (code = 0; code < 256; ++code) {
      name = ((Gfx8BitFont *)gfxFont)->getCharName(code);
      if (name && name[0] == 'm' && name[1] == '\0') {
	mCode = code;
      }
      if (letterCode < 0 && name && name[1] == '\0' &&
	  ((name[0] >= 'A' && name[0] <= 'Z') ||
	   (name[0] >= 'a' && name[0] <= 'z'))) {
	letterCode = code;
      }
      if (anyCode < 0 && name &&
	  ((Gfx8BitFont *)gfxFont)->getWidth(code) > 0) {
	anyCode = code;
      }
    }
    if (mCode >= 0 &&
	(w = ((Gfx8BitFont *)gfxFont)->getWidth(mCode)) > 0) {
      // 0.6 is a generic average 'm' width -- yes, this is a hack
      curFontSize *= w / 0.6;
    } else if (letterCode >= 0 &&
	       (w = ((Gfx8BitFont *)gfxFont)->getWidth(letterCode)) > 0) {
      // even more of a hack: 0.5 is a generic letter width
      curFontSize *= w / 0.5;
    } else if (anyCode >= 0 &&
	       (w = ((Gfx8BitFont *)gfxFont)->getWidth(anyCode)) > 0) {
      // better than nothing: 0.5 is a generic character width
      curFontSize *= w / 0.5;
    }
    fm = gfxFont->getFontMatrix();
    if (fm[0] != 0) {
      curFontSize *= fabs(fm[3] / fm[0]);
    }
  }
}

void TextPage::beginWord(GfxState *state, double x0, double y0) {
  double *fontm;
  double m[4], m2[4];
  int rot;

  // This check is needed because Type 3 characters can contain
  // text-drawing operations (when TextPage is being used via
  // {X,Win}SplashOutputDev rather than TextOutputDev).
  if (curWord) {
    ++nest;
    return;
  }

  // compute the rotation
  state->getFontTransMat(&m[0], &m[1], &m[2], &m[3]);
  if (state->getFont()->getType() == fontType3) {
    fontm = state->getFont()->getFontMatrix();
    m2[0] = fontm[0] * m[0] + fontm[1] * m[2];
    m2[1] = fontm[0] * m[1] + fontm[1] * m[3];
    m2[2] = fontm[2] * m[0] + fontm[3] * m[2];
    m2[3] = fontm[2] * m[1] + fontm[3] * m[3];
    m[0] = m2[0];
    m[1] = m2[1];
    m[2] = m2[2];
    m[3] = m2[3];
  }
  if (fabs(m[0] * m[3]) > fabs(m[1] * m[2])) {
    rot = (m[3] < 0) ? 0 : 2;
  } else {
    rot = (m[2] > 0) ? 1 : 3;
  }

  curWord = new TextWord(state, rot, x0, y0, charPos, curFont, curFontSize);
}

void TextPage::addChar(GfxState *state, double x, double y,
		       double dx, double dy,
		       CharCode c, int nBytes, Unicode *u, int uLen) {
  double x1, y1, w1, h1, dx2, dy2, base, sp, delta;
  GBool overlap;
  int i;

  // subtract char and word spacing from the dx,dy values
  sp = state->getCharSpace();
  if (c == (CharCode)0x20) {
    sp += state->getWordSpace();
  }
  state->textTransformDelta(sp * state->getHorizScaling(), 0, &dx2, &dy2);
  dx -= dx2;
  dy -= dy2;
  state->transformDelta(dx, dy, &w1, &h1);

  // throw away chars that aren't inside the page bounds
  // (and also do a sanity check on the character size)
  state->transform(x, y, &x1, &y1);
  if (x1 + w1 < 0 || x1 > pageWidth ||
      y1 + h1 < 0 || y1 > pageHeight ||
      w1 > pageWidth || h1 > pageHeight) {
    charPos += nBytes;
    return;
  }

  // check the tiny chars limit
  if (!globalParams->getTextKeepTinyChars() &&
      fabs(w1) < 3 && fabs(h1) < 3) {
    if (++nTinyChars > 50000) {
      charPos += nBytes;
      return;
    }
  }

  // break words at space character
  if (uLen == 1 && u[0] == (Unicode)0x20) {
    if (curWord) {
      ++curWord->charLen;
    }
    charPos += nBytes;
    endWord();
    return;
  }

  // start a new word if:
  // (1) this character doesn't fall in the right place relative to
  //     the end of the previous word (this places upper and lower
  //     constraints on the position deltas along both the primary
  //     and secondary axes), or
  // (2) this character overlaps the previous one (duplicated text), or
  // (3) the previous character was an overlap (we want each duplicated
  //     character to be in a word by itself at this stage),
  // (4) the font size has changed
  if (curWord && curWord->len > 0) {
    base = sp = delta = 0; // make gcc happy
    switch (curWord->rot) {
    case 0:
      base = y1;
      sp = x1 - curWord->xMax;
      delta = x1 - curWord->edge[curWord->len - 1];
      break;
    case 1:
      base = x1;
      sp = y1 - curWord->yMax;
      delta = y1 - curWord->edge[curWord->len - 1];
      break;
    case 2:
      base = y1;
      sp = curWord->xMin - x1;
      delta = curWord->edge[curWord->len - 1] - x1;
      break;
    case 3:
      base = x1;
      sp = curWord->yMin - y1;
      delta = curWord->edge[curWord->len - 1] - y1;
      break;
    }
    overlap = fabs(delta) < dupMaxPriDelta * curWord->fontSize &&
              fabs(base - curWord->base) < dupMaxSecDelta * curWord->fontSize;
    if (overlap || lastCharOverlap ||
	sp < -minDupBreakOverlap * curWord->fontSize ||
	sp > minWordBreakSpace * curWord->fontSize ||
	fabs(base - curWord->base) > 0.5 ||
	curFontSize != curWord->fontSize) {
      endWord();
    }
    lastCharOverlap = overlap;
  } else {
    lastCharOverlap = gFalse;
  }

  if (uLen != 0) {
    // start a new word if needed
    if (!curWord) {
      beginWord(state, x, y);
    }

    // page rotation and/or transform matrices can cause text to be
    // drawn in reverse order -- in this case, swap the begin/end
    // coordinates and break text into individual chars
    if ((curWord->rot == 0 && w1 < 0) ||
	(curWord->rot == 1 && h1 < 0) ||
	(curWord->rot == 2 && w1 > 0) ||
	(curWord->rot == 3 && h1 > 0)) {
      endWord();
      beginWord(state, x + dx, y + dy);
      x1 += w1;
      y1 += h1;
      w1 = -w1;