gfxfont.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

	}
	obj2.free();
      }
    }
    obj1.free();
    if (!fontDict->lookup("CharProcs", &charProcs)->isDict()) {
      error(-1, "Missing or invalid CharProcs dictionary in Type 3 font");
      charProcs.free();
    }
    if (!fontDict->lookup("Resources", &resources)->isDict()) {
      resources.free();
    }
  }

  //----- build the font encoding -----

  // Encodings start with a base encoding, which can come from
  // (in order of priority):
  //   1. FontDict.Encoding or FontDict.Encoding.BaseEncoding
  //        - MacRoman / MacExpert / WinAnsi / Standard
  //   2. embedded or external font file
  //   3. default:
  //        - builtin --> builtin encoding
  //        - TrueType --> WinAnsiEncoding
  //        - others --> StandardEncoding
  // and then add a list of differences (if any) from
  // FontDict.Encoding.Differences.

  // check FontDict for base encoding
  hasEncoding = gFalse;
  usesMacRomanEnc = gFalse;
  baseEnc = NULL;
  baseEncFromFontFile = gFalse;
  fontDict->lookup("Encoding", &obj1);
  if (obj1.isDict()) {
    obj1.dictLookup("BaseEncoding", &obj2);
    if (obj2.isName("MacRomanEncoding")) {
      hasEncoding = gTrue;
      usesMacRomanEnc = gTrue;
      baseEnc = macRomanEncoding;
    } else if (obj2.isName("MacExpertEncoding")) {
      hasEncoding = gTrue;
      baseEnc = macExpertEncoding;
    } else if (obj2.isName("WinAnsiEncoding")) {
      hasEncoding = gTrue;
      baseEnc = winAnsiEncoding;
    }
    obj2.free();
  } else if (obj1.isName("MacRomanEncoding")) {
    hasEncoding = gTrue;
    usesMacRomanEnc = gTrue;
    baseEnc = macRomanEncoding;
  } else if (obj1.isName("MacExpertEncoding")) {
    hasEncoding = gTrue;
    baseEnc = macExpertEncoding;
  } else if (obj1.isName("WinAnsiEncoding")) {
    hasEncoding = gTrue;
    baseEnc = winAnsiEncoding;
  }

  // check embedded or external font file for base encoding
  // (only for Type 1 fonts - trying to get an encoding out of a
  // TrueType font is a losing proposition)
  ffT1 = NULL;
  ffT1C = NULL;
  buf = NULL;
  if (type == fontType1 && (extFontFile || embFontID.num >= 0)) {
    if (extFontFile) {
      ffT1 = FoFiType1::load(extFontFile->getCString());
    } else {
      buf = readEmbFontFile(xref, &len);
      ffT1 = FoFiType1::make(buf, len);
    }
    if (ffT1) {
      if (ffT1->getName()) {
	if (embFontName) {
	  delete embFontName;
	}
	embFontName = new GString(ffT1->getName());
      }
      if (!baseEnc) {
	baseEnc = ffT1->getEncoding();
	baseEncFromFontFile = gTrue;
      }
    }
  } else if (type == fontType1C && (extFontFile || embFontID.num >= 0)) {
    if (extFontFile) {
      ffT1C = FoFiType1C::load(extFontFile->getCString());
    } else {
      buf = readEmbFontFile(xref, &len);
      ffT1C = FoFiType1C::make(buf, len);
    }
    if (ffT1C) {
      if (ffT1C->getName()) {
	if (embFontName) {
	  delete embFontName;
	}
	embFontName = new GString(ffT1C->getName());
      }
      if (!baseEnc) {
	baseEnc = ffT1C->getEncoding();
	baseEncFromFontFile = gTrue;
      }
    }
  }
  if (buf) {
    gfree(buf);
  }

  // get default base encoding
  if (!baseEnc) {
    if (builtinFont && embFontID.num < 0) {
      baseEnc = builtinFont->defaultBaseEnc;
      hasEncoding = gTrue;
    } else if (type == fontTrueType) {
      baseEnc = winAnsiEncoding;
    } else {
      baseEnc = standardEncoding;
    }
  }

  // copy the base encoding
  for (i = 0; i < 256; ++i) {
    enc[i] = baseEnc[i];
    if ((encFree[i] = baseEncFromFontFile) && enc[i]) {
      enc[i] = copyString(baseEnc[i]);
    }
  }

  // some Type 1C font files have empty encodings, which can break the
  // T1C->T1 conversion (since the 'seac' operator depends on having
  // the accents in the encoding), so we fill in any gaps from
  // StandardEncoding
  if (type == fontType1C && (extFontFile || embFontID.num >= 0) &&
      baseEncFromFontFile) {
    for (i = 0; i < 256; ++i) {
      if (!enc[i] && standardEncoding[i]) {
	enc[i] = standardEncoding[i];
	encFree[i] = gFalse;
      }
    }
  }

  // merge differences into encoding
  if (obj1.isDict()) {
    obj1.dictLookup("Differences", &obj2);
    if (obj2.isArray()) {
      hasEncoding = gTrue;
      code = 0;
      for (i = 0; i < obj2.arrayGetLength(); ++i) {
	obj2.arrayGet(i, &obj3);
	if (obj3.isInt()) {
	  code = obj3.getInt();
	} else if (obj3.isName()) {
	  if (code >= 0 && code < 256) {
	    if (encFree[code]) {
	      gfree(enc[code]);
	    }
	    enc[code] = copyString(obj3.getName());
	    encFree[code] = gTrue;
	  }
	  ++code;
	} else {
	  error(-1, "Wrong type in font encoding resource differences (%s)",
		obj3.getTypeName());
	}
	obj3.free();
      }
    }
    obj2.free();
  }
  obj1.free();
  if (ffT1) {
    delete ffT1;
  }
  if (ffT1C) {
    delete ffT1C;
  }

  //----- build the mapping to Unicode -----

  // pass 1: use the name-to-Unicode mapping table
  missing = hex = gFalse;
  for (code = 0; code < 256; ++code) {
    if ((charName = enc[code])) {
      if (!(toUnicode[code] = globalParams->mapNameToUnicode(charName)) &&
	  strcmp(charName, ".notdef")) {
	// if it wasn't in the name-to-Unicode table, check for a
	// name that looks like 'Axx' or 'xx', where 'A' is any letter
	// and 'xx' is two hex digits
	if ((strlen(charName) == 3 &&
	     isalpha(charName[0]) &&
	     isxdigit(charName[1]) && isxdigit(charName[2]) &&
	     ((charName[1] >= 'a' && charName[1] <= 'f') ||
	      (charName[1] >= 'A' && charName[1] <= 'F') ||
	      (charName[2] >= 'a' && charName[2] <= 'f') ||
	      (charName[2] >= 'A' && charName[2] <= 'F'))) ||
	    (strlen(charName) == 2 &&
	     isxdigit(charName[0]) && isxdigit(charName[1]) &&
	     ((charName[0] >= 'a' && charName[0] <= 'f') ||
	      (charName[0] >= 'A' && charName[0] <= 'F') ||
	      (charName[1] >= 'a' && charName[1] <= 'f') ||
	      (charName[1] >= 'A' && charName[1] <= 'F')))) {
	  hex = gTrue;
	}
	missing = gTrue;
      }
    } else {
      toUnicode[code] = 0;
    }
  }

  // pass 2: try to fill in the missing chars, looking for names of
  // the form 'Axx', 'xx', 'Ann', 'ABnn', or 'nn', where 'A' and 'B'
  // are any letters, 'xx' is two hex digits, and 'nn' is 2-4
  // decimal digits
  if (missing && globalParams->getMapNumericCharNames()) {
    for (code = 0; code < 256; ++code) {
      if ((charName = enc[code]) && !toUnicode[code] &&
	  strcmp(charName, ".notdef")) {
	n = strlen(charName);
	code2 = -1;
	if (hex && n == 3 && isalpha(charName[0]) &&
	    isxdigit(charName[1]) && isxdigit(charName[2])) {
	  sscanf(charName+1, "%x", &code2);
	} else if (hex && n == 2 &&
		   isxdigit(charName[0]) && isxdigit(charName[1])) {
	  sscanf(charName, "%x", &code2);
	} else if (!hex && n >= 2 && n <= 4 &&
		   isdigit(charName[0]) && isdigit(charName[1])) {
	  code2 = atoi(charName);
	} else if (n >= 3 && n <= 5 &&
		   isdigit(charName[1]) && isdigit(charName[2])) {
	  code2 = atoi(charName+1);
	} else if (n >= 4 && n <= 6 &&
		   isdigit(charName[2]) && isdigit(charName[3])) {
	  code2 = atoi(charName+2);
	}
	if (code2 >= 0 && code2 <= 0xff) {
	  toUnicode[code] = (Unicode)code2;
	}
      }
    }

  // if the 'mapUnknownCharNames' flag is set, do a simple pass-through
  // mapping for unknown character names
  } else if (missing && globalParams->getMapUnknownCharNames()) {
    for (code = 0; code < 256; ++code) {
      if (!toUnicode[code]) {
	toUnicode[code] = code;
      }
    }
  }

  // construct the char code -> Unicode mapping object
  ctu = CharCodeToUnicode::make8BitToUnicode(toUnicode);

  // merge in a ToUnicode CMap, if there is one -- this overwrites
  // existing entries in ctu, i.e., the ToUnicode CMap takes
  // precedence, but the other encoding info is allowed to fill in any
  // holes
  readToUnicodeCMap(fontDict, 8, ctu);

  // look for a Unicode-to-Unicode mapping
  if (name && (utu = globalParams->getUnicodeToUnicode(name))) {
    for (i = 0; i < 256; ++i) {
      toUnicode[i] = 0;
    }
    ctu2 = CharCodeToUnicode::make8BitToUnicode(toUnicode);
    for (i = 0; i < 256; ++i) {
      n = ctu->mapToUnicode((CharCode)i, uBuf, 8);
      if (n >= 1) {
	n = utu->mapToUnicode((CharCode)uBuf[0], uBuf, 8);
	if (n >= 1) {
	  ctu2->setMapping((CharCode)i, uBuf, n);
	}
      }
    }
    utu->decRefCnt();
    delete ctu;
    ctu = ctu2;
  }

  //----- get the character widths -----

  // initialize all widths
  for (code = 0; code < 256; ++code) {
    widths[code] = missingWidth * 0.001;
  }

  // use widths from font dict, if present
  fontDict->lookup("FirstChar", &obj1);
  firstChar = obj1.isInt() ? obj1.getInt() : 0;
  obj1.free();
  if (firstChar < 0 || firstChar > 255) {
    firstChar = 0;
  }
  fontDict->lookup("LastChar", &obj1);
  lastChar = obj1.isInt() ? obj1.getInt() : 255;
  obj1.free();
  if (lastChar < 0 || lastChar > 255) {
    lastChar = 255;
  }
  mul = (type == fontType3) ? fontMat[0] : 0.001;
  fontDict->lookup("Widths", &obj1);
  if (obj1.isArray()) {
    flags |= fontFixedWidth;
    if (obj1.arrayGetLength() < lastChar - firstChar + 1) {
      lastChar = firstChar + obj1.arrayGetLength() - 1;
    }
    for (code = firstChar; code <= lastChar; ++code) {
      obj1.arrayGet(code - firstChar, &obj2);
      if (obj2.isNum()) {
	widths[code] = obj2.getNum() * mul;
	if (widths[code] != widths[firstChar]) {
	  flags &= ~fontFixedWidth;
	}
      }
      obj2.free();
    }

  // use widths from built-in font
  } else if (builtinFont) {
    // this is a kludge for broken PDF files that encode char 32
    // as .notdef
    if (builtinFont->widths->getWidth("space", &w)) {
      widths[32] = 0.001 * w;
    }
    for (code = 0; code < 256; ++code) {
      if (enc[code] && builtinFont->widths->getWidth(enc[code], &w)) {
	widths[code] = 0.001 * w;
      }
    }

  // couldn't find widths -- use defaults 
  } else {
    // this is technically an error -- the Widths entry is required
    // for all but the Base-14 fonts -- but certain PDF generators
    // apparently don't include widths for Arial and TimesNewRoman
    if (isFixedWidth()) {
      i = 0;
    } else if (isSerif()) {
      i = 8;
    } else {
      i = 4;
    }
    if (isBold()) {
      i += 2;
    }
    if (isItalic()) {
      i += 1;
    }
    builtinFont = builtinFontSubst[i];
    // this is a kludge for broken PDF files that encode char 32
    // as .notdef
    if (builtinFont->widths->getWidth("space", &w)) {
      widths[32] = 0.001 * w;
    }
    for (code = 0; code < 256; ++code) {
      if (enc[code] && builtinFont->widths->getWidth(enc[code], &w)) {
	widths[code] = 0.001 * w;
      }
    }
  }
  obj1.free();

  ok = gTrue;
}

Gfx8BitFont::~Gfx8BitFont() {
  int i;

  for (i = 0; i < 256; ++i) {
    if (encFree[i] && enc[i]) {
      gfree(enc[i]);
    }
  }
  ctu->decRefCnt();
  if (charProcs.isDict()) {
    charProcs.free();
  }
  if (resources.isDict()) {
    resources.free();
  }
}

int Gfx8BitFont::getNextChar(char *s, int len, CharCode *code,
			     Unicode *u, int uSize, int *uLen,
			     double *dx, double *dy, double *ox, double *oy) {
  CharCode c;

  *code = c = (CharCode)(*s & 0xff);
  *uLen = ctu->mapToUnicode(c, u, uSize);
  *dx = widths[c];
  *dy = *ox = *oy = 0;
  return 1;
}

CharCodeToUnicode *Gfx8BitFont::getToUnicode() {
  ctu->incRefCnt();
  return ctu;
}

Gushort *Gfx8BitFont::getCodeToGIDMap(FoFiTrueType *ff) {
  Gushort *map;
  int cmapPlatform, cmapEncoding;
  int unicodeCmap, macRomanCmap, msSymbolCmap, cmap;
  GBool useMacRoman, useUnicode;
  char *charName;
  Unicode u;
  int code, i, n;

  map = (Gushort *)gmallocn(256, sizeof(Gushort));
  for (i = 0; i < 256; ++i) {
    map[i] = 0;
  }

  // To match up with the Adobe-defined behaviour, we choose a cmap
  // like this:
  // 1. If the PDF font has an encoding:
  //    1a. If the PDF font specified MacRomanEncoding and the
  //        TrueType font has a Macintosh Roman cmap, use it, and
  //        reverse map the char names through MacRomanEncoding to
  //        get char codes.
  //    1b. If the TrueType font has a Microsoft Unicode cmap or a
  //        non-Microsoft Unicode cmap, use it, and use the Unicode
  //        indexes, not the char codes.
  //    1c. If the PDF font is symbolic and the TrueType font has a
  //        Microsoft Symbol cmap, use it, and use char codes
  //        directly (possibly with an offset of 0xf000).
  //    1d. If the TrueType font has a Macintosh Roman cmap, use it,
  //        as in case 1a.
  // 2. If the PDF font does not have an encoding or the PDF font is
  //    symbolic:
  //    2a. If the TrueType font has a Macintosh Roman cmap, use it,
  //        and use char codes directly (possibly with an offset of
  //        0xf000).
  //    2b. If the TrueType font has a Microsoft Symbol cmap, use it,
  //        and use char codes directly (possible with an offset of
  //        0xf000).
  // 3. If none of these rules apply, use the first cmap and hope for
  //    the best (this shouldn't happen).
  unicodeCmap = macRomanCmap = msSymbolCmap = -1;
  for (i = 0; i < ff->getNumCmaps(); ++i) {
    cmapPlatform = ff->getCmapPlatform(i);
    cmapEncoding = ff->getCmapEncoding(i);
    if ((cmapPlatform == 3 && cmapEncoding == 1) ||
	cmapPlatform == 0) {
      unicodeCmap = i;
    } else if (cmapPlatform == 1 && cmapEncoding == 0) {
      macRomanCmap = i;
    } else if (cmapPlatform == 3 && cmapEncoding == 0) {
      msSymbolCmap = i;
    }
  }
  cmap = 0;
  useMacRoman = gFalse;
  useUnicode = gFalse;
  if (hasEncoding) {
    if (usesMacRomanEnc && macRomanCmap >= 0) {
      cmap = macRomanCmap;
      useMacRoman = gTrue;
    } else if (unicodeCmap >= 0) {
      cmap = unicodeCmap;
      useUnicode = gTrue;
    } else if ((flags & fontSymbolic) && msSymbolCmap >= 0) {
      cmap = msSymbolCmap;
    } else if ((flags & fontSymbolic) && macRomanCmap >= 0) {
      cmap = macRomanCmap;
    } else if (macRomanCmap >= 0) {
      cmap = macRomanCmap;
      useMacRoman = gTrue;
    }
  } else {
    if (msSymbolCmap >= 0) {
      cmap = msSymbolCmap;
    } else if (macRomanCmap >= 0) {
      cmap = macRomanCmap;
    }
  }

  // reverse map the char names through MacRomanEncoding, then map the
  // char codes through the cmap
  if (useMacRoman) {
    for (i = 0; i < 256; ++i) {
      if ((charName = enc[i])) {
	if ((code = globalParams->getMacRomanCharCode(charName))) {
	  map[i] = ff->mapCodeToGID(cmap, code);
	}
      }
    }

  // map Unicode through the cmap
  } else if (useUnicode) {
    for (i = 0; i < 256; ++i) {
      if (((charName = enc[i]) &&
	   (u = globalParams->mapNameToUnicode(charName))) ||
	  (n = ctu->mapToUnicode((CharCode)i, &u, 1))) {
	map[i] = ff->mapCodeToGID(cmap, u);
      }
    }

  // map the char codes through the cmap, possibly with an offset of
  // 0xf000
  } else {
    for (i = 0; i < 256; ++i) {
      if (!(map[i] = ff->mapCodeToGID(cmap, i))) {
	map[i] = ff->mapCodeToGID(cmap, 0xf000 + i);
      }
    }
  }

  // try the TrueType 'post' table to handle any unmapped characters
  for (i = 0; i < 256; ++i) {
    if (!map[i] && (charName = enc[i])) {
      map[i] = (Gushort)(int)ff->mapNameToGID(charName);
    }
  }

  return map;
}

Dict *Gfx8BitFont::getCharProcs() {