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

      h1 = -h1;
    }

    // add the characters to the current word
    w1 /= uLen;
    h1 /= uLen;
    for (i = 0; i < uLen; ++i) {
      curWord->addChar(state, x1 + i*w1, y1 + i*h1, w1, h1, u[i]);
    }
  }
  if (curWord) {
    curWord->charLen += nBytes;
  }
  charPos += nBytes;
}

void TextPage::endWord() {
  // 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 (nest > 0) {
    --nest;
    return;
  }

  if (curWord) {
    addWord(curWord);
    curWord = NULL;
  }
}

void TextPage::addWord(TextWord *word) {
  // throw away zero-length words -- they don't have valid xMin/xMax
  // values, and they're useless anyway
  if (word->len == 0) {
    delete word;
    return;
  }

  if (rawOrder) {
    if (rawLastWord) {
      rawLastWord->next = word;
    } else {
      rawWords = word;
    }
    rawLastWord = word;
  } else {
    pools[word->rot]->addWord(word);
  }
}

void TextPage::addUnderline(double x0, double y0, double x1, double y1) {
  underlines->append(new TextUnderline(x0, y0, x1, y1));
}

void TextPage::addLink(int xMin, int yMin, int xMax, int yMax, Link *link) {
  links->append(new TextLink(xMin, yMin, xMax, yMax, link));
}

void TextPage::coalesce(GBool physLayout, GBool doHTML) {
  UnicodeMap *uMap;
  TextPool *pool;
  TextWord *word0, *word1, *word2;
  TextLine *line;
  TextBlock *blkList, *blkStack, *blk, *lastBlk, *blk0, *blk1;
  TextBlock **blkArray;
  TextFlow *flow, *lastFlow;
  TextUnderline *underline;
  TextLink *link;
  int rot, poolMinBaseIdx, baseIdx, startBaseIdx, endBaseIdx;
  double minBase, maxBase, newMinBase, newMaxBase;
  double fontSize, colSpace1, colSpace2, lineSpace, intraLineSpace, blkSpace;
  GBool found;
  int count[4];
  int lrCount;
  int firstBlkIdx, nBlocksLeft;
  int col1, col2;
  int i, j, n;

  if (rawOrder) {
    primaryRot = 0;
    primaryLR = gTrue;
    return;
  }

  uMap = globalParams->getTextEncoding();
  blkList = NULL;
  lastBlk = NULL;
  nBlocks = 0;
  primaryRot = -1;

#if 0 // for debugging
  printf("*** initial words ***\n");
  for (rot = 0; rot < 4; ++rot) {
    pool = pools[rot];
    for (baseIdx = pool->minBaseIdx; baseIdx <= pool->maxBaseIdx; ++baseIdx) {
      for (word0 = pool->getPool(baseIdx); word0; word0 = word0->next) {
	printf("    word: x=%.2f..%.2f y=%.2f..%.2f base=%.2f fontSize=%.2f rot=%d link=%p '",
	       word0->xMin, word0->xMax, word0->yMin, word0->yMax,
	       word0->base, word0->fontSize, rot*90, word0->link);
	for (i = 0; i < word0->len; ++i) {
	  fputc(word0->text[i] & 0xff, stdout);
	}
	printf("'\n");
      }
    }
  }
  printf("\n");
#endif

#if 0 //~ for debugging
  for (i = 0; i < underlines->getLength(); ++i) {
    underline = (TextUnderline *)underlines->get(i);
    printf("underline: x=%g..%g y=%g..%g horiz=%d\n",
	   underline->x0, underline->x1, underline->y0, underline->y1,
	   underline->horiz);
  }
#endif

  if (doHTML) {

    //----- handle underlining
    for (i = 0; i < underlines->getLength(); ++i) {
      underline = (TextUnderline *)underlines->get(i);
      if (underline->horiz) {
	// rot = 0
	if (pools[0]->minBaseIdx <= pools[0]->maxBaseIdx) {
	  startBaseIdx = pools[0]->getBaseIdx(underline->y0 + minUnderlineGap);
	  endBaseIdx = pools[0]->getBaseIdx(underline->y0 + maxUnderlineGap);
	  for (j = startBaseIdx; j <= endBaseIdx; ++j) {
	    for (word0 = pools[0]->getPool(j); word0; word0 = word0->next) {
	      //~ need to check the y value against the word baseline
	      if (underline->x0 < word0->xMin + underlineSlack &&
		  word0->xMax - underlineSlack < underline->x1) {
		word0->underlined = gTrue;
	      }
	    }
	  }
	}

	// rot = 2
	if (pools[2]->minBaseIdx <= pools[2]->maxBaseIdx) {
	  startBaseIdx = pools[2]->getBaseIdx(underline->y0 - maxUnderlineGap);
	  endBaseIdx = pools[2]->getBaseIdx(underline->y0 - minUnderlineGap);
	  for (j = startBaseIdx; j <= endBaseIdx; ++j) {
	    for (word0 = pools[2]->getPool(j); word0; word0 = word0->next) {
	      if (underline->x0 < word0->xMin + underlineSlack &&
		  word0->xMax - underlineSlack < underline->x1) {
		word0->underlined = gTrue;
	      }
	    }
	  }
	}
      } else {
	// rot = 1
	if (pools[1]->minBaseIdx <= pools[1]->maxBaseIdx) {
	  startBaseIdx = pools[1]->getBaseIdx(underline->x0 - maxUnderlineGap);
	  endBaseIdx = pools[1]->getBaseIdx(underline->x0 - minUnderlineGap);
	  for (j = startBaseIdx; j <= endBaseIdx; ++j) {
	    for (word0 = pools[1]->getPool(j); word0; word0 = word0->next) {
	      if (underline->y0 < word0->yMin + underlineSlack &&
		  word0->yMax - underlineSlack < underline->y1) {
		word0->underlined = gTrue;
	      }
	    }
	  }
	}

	// rot = 3
	if (pools[3]->minBaseIdx <= pools[3]->maxBaseIdx) {
	  startBaseIdx = pools[3]->getBaseIdx(underline->x0 + minUnderlineGap);
	  endBaseIdx = pools[3]->getBaseIdx(underline->x0 + maxUnderlineGap);
	  for (j = startBaseIdx; j <= endBaseIdx; ++j) {
	    for (word0 = pools[3]->getPool(j); word0; word0 = word0->next) {
	      if (underline->y0 < word0->yMin + underlineSlack &&
		  word0->yMax - underlineSlack < underline->y1) {
		word0->underlined = gTrue;
	      }
	    }
	  }
	}
      }
    }

    //----- handle links
    for (i = 0; i < links->getLength(); ++i) {
      link = (TextLink *)links->get(i);

      // rot = 0
      if (pools[0]->minBaseIdx <= pools[0]->maxBaseIdx) {
	startBaseIdx = pools[0]->getBaseIdx(link->yMin);
	endBaseIdx = pools[0]->getBaseIdx(link->yMax);
	for (j = startBaseIdx; j <= endBaseIdx; ++j) {
	  for (word0 = pools[0]->getPool(j); word0; word0 = word0->next) {
	    if (link->xMin < word0->xMin + hyperlinkSlack &&
		word0->xMax - hyperlinkSlack < link->xMax &&
		link->yMin < word0->yMin + hyperlinkSlack &&
		word0->yMax - hyperlinkSlack < link->yMax) {
	      word0->link = link->link;
	    }
	  }
	}
      }

      // rot = 2
      if (pools[2]->minBaseIdx <= pools[2]->maxBaseIdx) {
	startBaseIdx = pools[2]->getBaseIdx(link->yMin);
	endBaseIdx = pools[2]->getBaseIdx(link->yMax);
	for (j = startBaseIdx; j <= endBaseIdx; ++j) {
	  for (word0 = pools[2]->getPool(j); word0; word0 = word0->next) {
	    if (link->xMin < word0->xMin + hyperlinkSlack &&
		word0->xMax - hyperlinkSlack < link->xMax &&
		link->yMin < word0->yMin + hyperlinkSlack &&
		word0->yMax - hyperlinkSlack < link->yMax) {
	      word0->link = link->link;
	    }
	  }
	}
      }

      // rot = 1
      if (pools[1]->minBaseIdx <= pools[1]->maxBaseIdx) {
	startBaseIdx = pools[1]->getBaseIdx(link->xMin);
	endBaseIdx = pools[1]->getBaseIdx(link->xMax);
	for (j = startBaseIdx; j <= endBaseIdx; ++j) {
	  for (word0 = pools[1]->getPool(j); word0; word0 = word0->next) {
	    if (link->yMin < word0->yMin + hyperlinkSlack &&
		word0->yMax - hyperlinkSlack < link->yMax &&
		link->xMin < word0->xMin + hyperlinkSlack &&
		word0->xMax - hyperlinkSlack < link->xMax) {
	      word0->link = link->link;
	    }
	  }
	}
      }

      // rot = 3
      if (pools[3]->minBaseIdx <= pools[3]->maxBaseIdx) {
	startBaseIdx = pools[3]->getBaseIdx(link->xMin);
	endBaseIdx = pools[3]->getBaseIdx(link->xMax);
	for (j = startBaseIdx; j <= endBaseIdx; ++j) {
	  for (word0 = pools[3]->getPool(j); word0; word0 = word0->next) {
	    if (link->yMin < word0->yMin + hyperlinkSlack &&
		word0->yMax - hyperlinkSlack < link->yMax &&
		link->xMin < word0->xMin + hyperlinkSlack &&
		word0->xMax - hyperlinkSlack < link->xMax) {
	      word0->link = link->link;
	    }
	  }
	}
      }
    }
  }

  //----- assemble the blocks

  //~ add an outer loop for writing mode (vertical text)

  // build blocks for each rotation value
  for (rot = 0; rot < 4; ++rot) {
    pool = pools[rot];
    poolMinBaseIdx = pool->minBaseIdx;
    count[rot] = 0;

    // add blocks until no more words are left
    while (1) {

      // find the first non-empty line in the pool
      for (;
	   poolMinBaseIdx <= pool->maxBaseIdx &&
	     !pool->getPool(poolMinBaseIdx);
	   ++poolMinBaseIdx) ;
      if (poolMinBaseIdx > pool->maxBaseIdx) {
	break;
      }

      // look for the left-most word in the first four lines of the
      // pool -- this avoids starting with a superscript word
      startBaseIdx = poolMinBaseIdx;
      for (baseIdx = poolMinBaseIdx + 1;
	   baseIdx < poolMinBaseIdx + 4 && baseIdx <= pool->maxBaseIdx;
	   ++baseIdx) {
	if (!pool->getPool(baseIdx)) {
	  continue;
	}
	if (pool->getPool(baseIdx)->primaryCmp(pool->getPool(startBaseIdx))
	    < 0) {
	  startBaseIdx = baseIdx;
	}
      }

      // create a new block
      word0 = pool->getPool(startBaseIdx);
      pool->setPool(startBaseIdx, word0->next);
      word0->next = NULL;
      blk = new TextBlock(this, rot);
      blk->addWord(word0);

      fontSize = word0->fontSize;
      minBase = maxBase = word0->base;
      colSpace1 = minColSpacing1 * fontSize;
      colSpace2 = minColSpacing2 * fontSize;
      lineSpace = maxLineSpacingDelta * fontSize;
      intraLineSpace = maxIntraLineDelta * fontSize;

      // add words to the block
      do {
	found = gFalse;

	// look for words on the line above the current top edge of
	// the block
	newMinBase = minBase;
	for (baseIdx = pool->getBaseIdx(minBase);
	     baseIdx >= pool->getBaseIdx(minBase - lineSpace);
	     --baseIdx) {
	  word0 = NULL;
	  word1 = pool->getPool(baseIdx);
	  while (word1) {
	    if (word1->base < minBase &&
		word1->base >= minBase - lineSpace &&
		((rot == 0 || rot == 2)
		 ? (word1->xMin < blk->xMax && word1->xMax > blk->xMin)
		 : (word1->yMin < blk->yMax && word1->yMax > blk->yMin)) &&
		fabs(word1->fontSize - fontSize) <
		  maxBlockFontSizeDelta1 * fontSize) {
	      word2 = word1;
	      if (word0) {
		word0->next = word1->next;
	      } else {
		pool->setPool(baseIdx, word1->next);
	      }
	      word1 = word1->next;
	      word2->next = NULL;
	      blk->addWord(word2);
	      found = gTrue;
	      newMinBase = word2->base;
	    } else {
	      word0 = word1;
	      word1 = word1->next;
	    }
	  }
	}
	minBase = newMinBase;

	// look for words on the line below the current bottom edge of
	// the block
	newMaxBase = maxBase;
	for (baseIdx = pool->getBaseIdx(maxBase);
	     baseIdx <= pool->getBaseIdx(maxBase + lineSpace);
	     ++baseIdx) {
	  word0 = NULL;
	  word1 = pool->getPool(baseIdx);
	  while (word1) {
	    if (word1->base > maxBase &&
		word1->base <= maxBase + lineSpace &&
		((rot == 0 || rot == 2)
		 ? (word1->xMin < blk->xMax && word1->xMax > blk->xMin)
		 : (word1->yMin < blk->yMax && word1->yMax > blk->yMin)) &&
		fabs(word1->fontSize - fontSize) <
		  maxBlockFontSizeDelta1 * fontSize) {
	      word2 = word1;
	      if (word0) {
		word0->next = word1->next;
	      } else {
		pool->setPool(baseIdx, word1->next);
	      }
	      word1 = word1->next;
	      word2->next = NULL;
	      blk->addWord(word2);
	      found = gTrue;
	      newMaxBase = word2->base;
	    } else {
	      word0 = word1;
	      word1 = word1->next;
	    }
	  }
	}
	maxBase = newMaxBase;

	// look for words that are on lines already in the block, and
	// that overlap the block horizontally
	for (baseIdx = pool->getBaseIdx(minBase - intraLineSpace);
	     baseIdx <= pool->getBaseIdx(maxBase + intraLineSpace);
	     ++baseIdx) {
	  word0 = NULL;
	  word1 = pool->getPool(baseIdx);
	  while (word1) {
	    if (word1->base >= minBase - intraLineSpace &&
		word1->base <= maxBase + intraLineSpace &&
		((rot == 0 || rot == 2)
		 ? (word1->xMin < blk->xMax + colSpace1 &&
		    word1->xMax > blk->xMin - colSpace1)
		 : (word1->yMin < blk->yMax + colSpace1 &&
		    word1->yMax > blk->yMin - colSpace1)) &&
		fabs(word1->fontSize - fontSize) <
		  maxBlockFontSizeDelta2 * fontSize) {
	      word2 = word1;
	      if (word0) {
		word0->next = word1->next;
	      } else {
		pool->setPool(baseIdx, word1->next);
	      }
	      word1 = word1->next;
	      word2->next = NULL;
	      blk->addWord(word2);
	      found = gTrue;
	    } else {
	      word0 = word1;
	      word1 = word1->next;
	    }
	  }
	}

	// only check for outlying words (the next two chunks of code)
	// if we didn't find anything else
	if (found) {
	  continue;
	}

	// scan down the left side of the block, looking for words
	// that are near (but not overlapping) the block; if there are
	// three or fewer, add them to the block
	n = 0;
	for (baseIdx = pool->getBaseIdx(minBase - intraLineSpace);
	     baseIdx <= pool->getBaseIdx(maxBase + intraLineSpace);
	     ++baseIdx) {
	  word1 = pool->getPool(baseIdx);
	  while (word1) {
	    if (word1->base >= minBase - intraLineSpace &&
		word1->base <= maxBase + intraLineSpace &&
		((rot == 0 || rot == 2)
		 ? (word1->xMax <= blk->xMin &&
		    word1->xMax > blk->xMin - colSpace2)
		 : (word1->yMax <= blk->yMin &&
		    word1->yMax > blk->yMin - colSpace2)) &&
		fabs(word1->fontSize - fontSize) <
		  maxBlockFontSizeDelta3 * fontSize) {
	      ++n;
	      break;
	    }
	    word1 = word1->next;
	  }
	}
	if (n > 0 && n <= 3) {
	  for (baseIdx = pool->getBaseIdx(minBase - intraLineSpace);
	       baseIdx <= pool->getBaseIdx(maxBase + intraLineSpace);
	       ++baseIdx) {
	    word0 = NULL;
	    word1 = pool->getPool(baseIdx);
	    while (word1) {
	      if (word1->base >= minBase - intraLineSpace &&
		  word1->base <= maxBase + intraLineSpace &&
		  ((rot == 0 || rot == 2)
		   ? (word1->xMax <= blk->xMin &&
		      word1->xMax > blk->xMin - colSpace2)
		   : (word1->yMax <= blk->yMin &&
		      word1->yMax > blk->yMin - colSpace2)) &&
		  fabs(word1->fontSize - fontSize) <
		    maxBlockFontSizeDelta3 * fontSize) {
		word2 = word1;
		if (word0) {
		  word0->next = word1->next;
		} else {
		  pool->setPool(baseIdx, word1->next);
		}
		word1 = word1->next;
		word2->next = NULL;
		blk->addWord(word2);
		if (word2->base < minBase) {
		  minBase = word2->base;
		} else if (word2->base > maxBase) {
		  maxBase = word2->base;
		}
		found = gTrue;
		break;
	      } else {
		word0 = word1;
		word1 = word1->next;
	      }
	    }
	  }
	}

	// scan down the right side of the block, looking for words
	// that are near (but not overlapping) the block; if there are
	// three or fewer, add them to the block
	n = 0;
	for (baseIdx = pool->getBaseIdx(minBase - intraLineSpace);
	     baseIdx <= pool->getBaseIdx(maxBase + intraLineSpace);
	     ++baseIdx) {
	  word1 = pool->getPool(baseIdx);
	  while (word1) {
	    if (word1->base >= minBase - intraLineSpace &&
		word1->base <= maxBase + intraLineSpace &&
		((rot == 0 || rot == 2)
		 ? (word1->xMin >= blk->xMax &&
		    word1->xMin < blk->xMax + colSpace2)
		 : (word1->yMin >= blk->yMax &&
		    word1->yMin < blk->yMax + colSpace2)) &&
		fabs(word1->fontSize - fontSize) <
		  maxBlockFontSizeDelta3 * fontSize) {
	      ++n;
	      break;
	    }
	    word1 = word1->next;
	  }
	}
	if (n > 0 && n <= 3) {
	  for (baseIdx = pool->getBaseIdx(minBase - intraLineSpace);
	       baseIdx <= pool->getBaseIdx(maxBase + intraLineSpace);
	       ++baseIdx) {
	    word0 = NU