pdf.js/src/core/cff_parser.js

/* Copyright 2016 Mozilla Foundation
 *
 * Licensed under the Apache License, Version 2.0 (the "License");
 * you may not use this file except in compliance with the License.
 * You may obtain a copy of the License at
 *
 *     http://www.apache.org/licenses/LICENSE-2.0
 *
 * Unless required by applicable law or agreed to in writing, software
 * distributed under the License is distributed on an "AS IS" BASIS,
 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 * See the License for the specific language governing permissions and
 * limitations under the License.
 */

import {
  bytesToString, FormatError, info, stringToBytes, Util, warn
} from '../shared/util';
import {
  ExpertCharset, ExpertSubsetCharset, ISOAdobeCharset
} from './charsets';
import { ExpertEncoding, StandardEncoding } from './encodings';

// Maximum subroutine call depth of type 2 chartrings. Matches OTS.
var MAX_SUBR_NESTING = 10;

/**
 * The CFF class takes a Type1 file and wrap it into a
 * 'Compact Font Format' which itself embed Type2 charstrings.
 */
var CFFStandardStrings = [
  '.notdef', 'space', 'exclam', 'quotedbl', 'numbersign', 'dollar', 'percent',
  'ampersand', 'quoteright', 'parenleft', 'parenright', 'asterisk', 'plus',
  'comma', 'hyphen', 'period', 'slash', 'zero', 'one', 'two', 'three', 'four',
  'five', 'six', 'seven', 'eight', 'nine', 'colon', 'semicolon', 'less',
  'equal', 'greater', 'question', 'at', 'A', 'B', 'C', 'D', 'E', 'F', 'G', 'H',
  'I', 'J', 'K', 'L', 'M', 'N', 'O', 'P', 'Q', 'R', 'S', 'T', 'U', 'V', 'W',
  'X', 'Y', 'Z', 'bracketleft', 'backslash', 'bracketright', 'asciicircum',
  'underscore', 'quoteleft', 'a', 'b', 'c', 'd', 'e', 'f', 'g', 'h', 'i', 'j',
  'k', 'l', 'm', 'n', 'o', 'p', 'q', 'r', 's', 't', 'u', 'v', 'w', 'x', 'y',
  'z', 'braceleft', 'bar', 'braceright', 'asciitilde', 'exclamdown', 'cent',
  'sterling', 'fraction', 'yen', 'florin', 'section', 'currency',
  'quotesingle', 'quotedblleft', 'guillemotleft', 'guilsinglleft',
  'guilsinglright', 'fi', 'fl', 'endash', 'dagger', 'daggerdbl',
  'periodcentered', 'paragraph', 'bullet', 'quotesinglbase', 'quotedblbase',
  'quotedblright', 'guillemotright', 'ellipsis', 'perthousand', 'questiondown',
  'grave', 'acute', 'circumflex', 'tilde', 'macron', 'breve', 'dotaccent',
  'dieresis', 'ring', 'cedilla', 'hungarumlaut', 'ogonek', 'caron', 'emdash',
  'AE', 'ordfeminine', 'Lslash', 'Oslash', 'OE', 'ordmasculine', 'ae',
  'dotlessi', 'lslash', 'oslash', 'oe', 'germandbls', 'onesuperior',
  'logicalnot', 'mu', 'trademark', 'Eth', 'onehalf', 'plusminus', 'Thorn',
  'onequarter', 'divide', 'brokenbar', 'degree', 'thorn', 'threequarters',
  'twosuperior', 'registered', 'minus', 'eth', 'multiply', 'threesuperior',
  'copyright', 'Aacute', 'Acircumflex', 'Adieresis', 'Agrave', 'Aring',
  'Atilde', 'Ccedilla', 'Eacute', 'Ecircumflex', 'Edieresis', 'Egrave',
  'Iacute', 'Icircumflex', 'Idieresis', 'Igrave', 'Ntilde', 'Oacute',
  'Ocircumflex', 'Odieresis', 'Ograve', 'Otilde', 'Scaron', 'Uacute',
  'Ucircumflex', 'Udieresis', 'Ugrave', 'Yacute', 'Ydieresis', 'Zcaron',
  'aacute', 'acircumflex', 'adieresis', 'agrave', 'aring', 'atilde',
  'ccedilla', 'eacute', 'ecircumflex', 'edieresis', 'egrave', 'iacute',
  'icircumflex', 'idieresis', 'igrave', 'ntilde', 'oacute', 'ocircumflex',
  'odieresis', 'ograve', 'otilde', 'scaron', 'uacute', 'ucircumflex',
  'udieresis', 'ugrave', 'yacute', 'ydieresis', 'zcaron', 'exclamsmall',
  'Hungarumlautsmall', 'dollaroldstyle', 'dollarsuperior', 'ampersandsmall',
  'Acutesmall', 'parenleftsuperior', 'parenrightsuperior', 'twodotenleader',
  'onedotenleader', 'zerooldstyle', 'oneoldstyle', 'twooldstyle',
  'threeoldstyle', 'fouroldstyle', 'fiveoldstyle', 'sixoldstyle',
  'sevenoldstyle', 'eightoldstyle', 'nineoldstyle', 'commasuperior',
  'threequartersemdash', 'periodsuperior', 'questionsmall', 'asuperior',
  'bsuperior', 'centsuperior', 'dsuperior', 'esuperior', 'isuperior',
  'lsuperior', 'msuperior', 'nsuperior', 'osuperior', 'rsuperior', 'ssuperior',
  'tsuperior', 'ff', 'ffi', 'ffl', 'parenleftinferior', 'parenrightinferior',
  'Circumflexsmall', 'hyphensuperior', 'Gravesmall', 'Asmall', 'Bsmall',
  'Csmall', 'Dsmall', 'Esmall', 'Fsmall', 'Gsmall', 'Hsmall', 'Ismall',
  'Jsmall', 'Ksmall', 'Lsmall', 'Msmall', 'Nsmall', 'Osmall', 'Psmall',
  'Qsmall', 'Rsmall', 'Ssmall', 'Tsmall', 'Usmall', 'Vsmall', 'Wsmall',
  'Xsmall', 'Ysmall', 'Zsmall', 'colonmonetary', 'onefitted', 'rupiah',
  'Tildesmall', 'exclamdownsmall', 'centoldstyle', 'Lslashsmall',
  'Scaronsmall', 'Zcaronsmall', 'Dieresissmall', 'Brevesmall', 'Caronsmall',
  'Dotaccentsmall', 'Macronsmall', 'figuredash', 'hypheninferior',
  'Ogoneksmall', 'Ringsmall', 'Cedillasmall', 'questiondownsmall', 'oneeighth',
  'threeeighths', 'fiveeighths', 'seveneighths', 'onethird', 'twothirds',
  'zerosuperior', 'foursuperior', 'fivesuperior', 'sixsuperior',
  'sevensuperior', 'eightsuperior', 'ninesuperior', 'zeroinferior',
  'oneinferior', 'twoinferior', 'threeinferior', 'fourinferior',
  'fiveinferior', 'sixinferior', 'seveninferior', 'eightinferior',
  'nineinferior', 'centinferior', 'dollarinferior', 'periodinferior',
  'commainferior', 'Agravesmall', 'Aacutesmall', 'Acircumflexsmall',
  'Atildesmall', 'Adieresissmall', 'Aringsmall', 'AEsmall', 'Ccedillasmall',
  'Egravesmall', 'Eacutesmall', 'Ecircumflexsmall', 'Edieresissmall',
  'Igravesmall', 'Iacutesmall', 'Icircumflexsmall', 'Idieresissmall',
  'Ethsmall', 'Ntildesmall', 'Ogravesmall', 'Oacutesmall', 'Ocircumflexsmall',
  'Otildesmall', 'Odieresissmall', 'OEsmall', 'Oslashsmall', 'Ugravesmall',
  'Uacutesmall', 'Ucircumflexsmall', 'Udieresissmall', 'Yacutesmall',
  'Thornsmall', 'Ydieresissmall', '001.000', '001.001', '001.002', '001.003',
  'Black', 'Bold', 'Book', 'Light', 'Medium', 'Regular', 'Roman', 'Semibold'
];

var CFFParser = (function CFFParserClosure() {
  var CharstringValidationData = [
    null,
    { id: 'hstem', min: 2, stackClearing: true, stem: true, },
    null,
    { id: 'vstem', min: 2, stackClearing: true, stem: true, },
    { id: 'vmoveto', min: 1, stackClearing: true, },
    { id: 'rlineto', min: 2, resetStack: true, },
    { id: 'hlineto', min: 1, resetStack: true, },
    { id: 'vlineto', min: 1, resetStack: true, },
    { id: 'rrcurveto', min: 6, resetStack: true, },
    null,
    { id: 'callsubr', min: 1, undefStack: true, },
    { id: 'return', min: 0, undefStack: true, },
    null, // 12
    null,
    { id: 'endchar', min: 0, stackClearing: true, },
    null,
    null,
    null,
    { id: 'hstemhm', min: 2, stackClearing: true, stem: true, },
    { id: 'hintmask', min: 0, stackClearing: true, },
    { id: 'cntrmask', min: 0, stackClearing: true, },
    { id: 'rmoveto', min: 2, stackClearing: true, },
    { id: 'hmoveto', min: 1, stackClearing: true, },
    { id: 'vstemhm', min: 2, stackClearing: true, stem: true, },
    { id: 'rcurveline', min: 8, resetStack: true, },
    { id: 'rlinecurve', min: 8, resetStack: true, },
    { id: 'vvcurveto', min: 4, resetStack: true, },
    { id: 'hhcurveto', min: 4, resetStack: true, },
    null, // shortint
    { id: 'callgsubr', min: 1, undefStack: true, },
    { id: 'vhcurveto', min: 4, resetStack: true, },
    { id: 'hvcurveto', min: 4, resetStack: true, }
  ];
  var CharstringValidationData12 = [
    null,
    null,
    null,
    { id: 'and', min: 2, stackDelta: -1, },
    { id: 'or', min: 2, stackDelta: -1, },
    { id: 'not', min: 1, stackDelta: 0, },
    null,
    null,
    null,
    { id: 'abs', min: 1, stackDelta: 0, },
    { id: 'add', min: 2, stackDelta: -1,
      stackFn: function stack_div(stack, index) {
        stack[index - 2] = stack[index - 2] + stack[index - 1];
      },
    },
    { id: 'sub', min: 2, stackDelta: -1,
      stackFn: function stack_div(stack, index) {
        stack[index - 2] = stack[index - 2] - stack[index - 1];
      },
    },
    { id: 'div', min: 2, stackDelta: -1,
      stackFn: function stack_div(stack, index) {
        stack[index - 2] = stack[index - 2] / stack[index - 1];
      },
    },
    null,
    { id: 'neg', min: 1, stackDelta: 0,
      stackFn: function stack_div(stack, index) {
        stack[index - 1] = -stack[index - 1];
      },
    },
    { id: 'eq', min: 2, stackDelta: -1, },
    null,
    null,
    { id: 'drop', min: 1, stackDelta: -1, },
    null,
    { id: 'put', min: 2, stackDelta: -2, },
    { id: 'get', min: 1, stackDelta: 0, },
    { id: 'ifelse', min: 4, stackDelta: -3, },
    { id: 'random', min: 0, stackDelta: 1, },
    { id: 'mul', min: 2, stackDelta: -1,
      stackFn: function stack_div(stack, index) {
        stack[index - 2] = stack[index - 2] * stack[index - 1];
      },
    },
    null,
    { id: 'sqrt', min: 1, stackDelta: 0, },
    { id: 'dup', min: 1, stackDelta: 1, },
    { id: 'exch', min: 2, stackDelta: 0, },
    { id: 'index', min: 2, stackDelta: 0, },
    { id: 'roll', min: 3, stackDelta: -2, },
    null,
    null,
    null,
    { id: 'hflex', min: 7, resetStack: true, },
    { id: 'flex', min: 13, resetStack: true, },
    { id: 'hflex1', min: 9, resetStack: true, },
    { id: 'flex1', min: 11, resetStack: true, }
  ];

  function CFFParser(file, properties, seacAnalysisEnabled) {
    this.bytes = file.getBytes();
    this.properties = properties;
    this.seacAnalysisEnabled = !!seacAnalysisEnabled;
  }
  CFFParser.prototype = {
    parse: function CFFParser_parse() {
      var properties = this.properties;
      var cff = new CFF();
      this.cff = cff;

      // The first five sections must be in order, all the others are reached
      // via offsets contained in one of the below.
      var header = this.parseHeader();
      var nameIndex = this.parseIndex(header.endPos);
      var topDictIndex = this.parseIndex(nameIndex.endPos);
      var stringIndex = this.parseIndex(topDictIndex.endPos);
      var globalSubrIndex = this.parseIndex(stringIndex.endPos);

      var topDictParsed = this.parseDict(topDictIndex.obj.get(0));
      var topDict = this.createDict(CFFTopDict, topDictParsed, cff.strings);

      cff.header = header.obj;
      cff.names = this.parseNameIndex(nameIndex.obj);
      cff.strings = this.parseStringIndex(stringIndex.obj);
      cff.topDict = topDict;
      cff.globalSubrIndex = globalSubrIndex.obj;

      this.parsePrivateDict(cff.topDict);

      cff.isCIDFont = topDict.hasName('ROS');

      var charStringOffset = topDict.getByName('CharStrings');
      var charStringIndex = this.parseIndex(charStringOffset).obj;

      var fontMatrix = topDict.getByName('FontMatrix');
      if (fontMatrix) {
        properties.fontMatrix = fontMatrix;
      }

      var fontBBox = topDict.getByName('FontBBox');
      if (fontBBox) {
        // adjusting ascent/descent
        properties.ascent = Math.max(fontBBox[3], fontBBox[1]);
        properties.descent = Math.min(fontBBox[1], fontBBox[3]);
        properties.ascentScaled = true;
      }

      var charset, encoding;
      if (cff.isCIDFont) {
        var fdArrayIndex = this.parseIndex(topDict.getByName('FDArray')).obj;
        for (var i = 0, ii = fdArrayIndex.count; i < ii; ++i) {
          var dictRaw = fdArrayIndex.get(i);
          var fontDict = this.createDict(CFFTopDict, this.parseDict(dictRaw),
                                         cff.strings);
          this.parsePrivateDict(fontDict);
          cff.fdArray.push(fontDict);
        }
        // cid fonts don't have an encoding
        encoding = null;
        charset = this.parseCharsets(topDict.getByName('charset'),
                                     charStringIndex.count, cff.strings, true);
        cff.fdSelect = this.parseFDSelect(topDict.getByName('FDSelect'),
                                          charStringIndex.count);
      } else {
        charset = this.parseCharsets(topDict.getByName('charset'),
                                     charStringIndex.count, cff.strings, false);
        encoding = this.parseEncoding(topDict.getByName('Encoding'),
                                      properties,
                                      cff.strings, charset.charset);
      }

      cff.charset = charset;
      cff.encoding = encoding;

      var charStringsAndSeacs = this.parseCharStrings({
        charStrings: charStringIndex,
        localSubrIndex: topDict.privateDict.subrsIndex,
        globalSubrIndex: globalSubrIndex.obj,
        fdSelect: cff.fdSelect,
        fdArray: cff.fdArray,
        privateDict: topDict.privateDict,
      });
      cff.charStrings = charStringsAndSeacs.charStrings;
      cff.seacs = charStringsAndSeacs.seacs;
      cff.widths = charStringsAndSeacs.widths;

      return cff;
    },
    parseHeader: function CFFParser_parseHeader() {
      var bytes = this.bytes;
      var bytesLength = bytes.length;
      var offset = 0;

      // Prevent an infinite loop, by checking that the offset is within the
      // bounds of the bytes array. Necessary in empty, or invalid, font files.
      while (offset < bytesLength && bytes[offset] !== 1) {
        ++offset;
      }
      if (offset >= bytesLength) {
        throw new FormatError('Invalid CFF header');
      }
      if (offset !== 0) {
        info('cff data is shifted');
        bytes = bytes.subarray(offset);
        this.bytes = bytes;
      }
      var major = bytes[0];
      var minor = bytes[1];
      var hdrSize = bytes[2];
      var offSize = bytes[3];
      var header = new CFFHeader(major, minor, hdrSize, offSize);
      return { obj: header, endPos: hdrSize, };
    },
    parseDict: function CFFParser_parseDict(dict) {
      var pos = 0;

      function parseOperand() {
        var value = dict[pos++];
        if (value === 30) {
          return parseFloatOperand();
        } else if (value === 28) {
          value = dict[pos++];
          value = ((value << 24) | (dict[pos++] << 16)) >> 16;
          return value;
        } else if (value === 29) {
          value = dict[pos++];
          value = (value << 8) | dict[pos++];
          value = (value << 8) | dict[pos++];
          value = (value << 8) | dict[pos++];
          return value;
        } else if (value >= 32 && value <= 246) {
          return value - 139;
        } else if (value >= 247 && value <= 250) {
          return ((value - 247) * 256) + dict[pos++] + 108;
        } else if (value >= 251 && value <= 254) {
          return -((value - 251) * 256) - dict[pos++] - 108;
        }
        warn('CFFParser_parseDict: "' + value + '" is a reserved command.');
        return NaN;
      }

      function parseFloatOperand() {
        var str = '';
        var eof = 15;
        var lookup = ['0', '1', '2', '3', '4', '5', '6', '7', '8',
            '9', '.', 'E', 'E-', null, '-'];
        var length = dict.length;
        while (pos < length) {
          var b = dict[pos++];
          var b1 = b >> 4;
          var b2 = b & 15;

          if (b1 === eof) {
            break;
          }
          str += lookup[b1];

          if (b2 === eof) {
            break;
          }
          str += lookup[b2];
        }
        return parseFloat(str);
      }

      var operands = [];
      var entries = [];

      pos = 0;
      var end = dict.length;
      while (pos < end) {
        var b = dict[pos];
        if (b <= 21) {
          if (b === 12) {
            b = (b << 8) | dict[++pos];
          }
          entries.push([b, operands]);
          operands = [];
          ++pos;
        } else {
          operands.push(parseOperand());
        }
      }
      return entries;
    },
    parseIndex: function CFFParser_parseIndex(pos) {
      var cffIndex = new CFFIndex();
      var bytes = this.bytes;
      var count = (bytes[pos++] << 8) | bytes[pos++];
      var offsets = [];
      var end = pos;
      var i, ii;

      if (count !== 0) {
        var offsetSize = bytes[pos++];
        // add 1 for offset to determine size of last object
        var startPos = pos + ((count + 1) * offsetSize) - 1;

        for (i = 0, ii = count + 1; i < ii; ++i) {
          var offset = 0;
          for (var j = 0; j < offsetSize; ++j) {
            offset <<= 8;
            offset += bytes[pos++];
          }
          offsets.push(startPos + offset);
        }
        end = offsets[count];
      }
      for (i = 0, ii = offsets.length - 1; i < ii; ++i) {
        var offsetStart = offsets[i];
        var offsetEnd = offsets[i + 1];
        cffIndex.add(bytes.subarray(offsetStart, offsetEnd));
      }
      return { obj: cffIndex, endPos: end, };
    },
    parseNameIndex: function CFFParser_parseNameIndex(index) {
      var names = [];
      for (var i = 0, ii = index.count; i < ii; ++i) {
        var name = index.get(i);
        names.push(bytesToString(name));
      }
      return names;
    },
    parseStringIndex: function CFFParser_parseStringIndex(index) {
      var strings = new CFFStrings();
      for (var i = 0, ii = index.count; i < ii; ++i) {
        var data = index.get(i);
        strings.add(bytesToString(data));
      }
      return strings;
    },
    createDict: function CFFParser_createDict(Type, dict, strings) {
      var cffDict = new Type(strings);
      for (var i = 0, ii = dict.length; i < ii; ++i) {
        var pair = dict[i];
        var key = pair[0];
        var value = pair[1];
        cffDict.setByKey(key, value);
      }
      return cffDict;
    },
    parseCharString: function CFFParser_parseCharString(state, data,
                                                        localSubrIndex,
                                                        globalSubrIndex) {
      if (!data || state.callDepth > MAX_SUBR_NESTING) {
        return false;
      }
      var stackSize = state.stackSize;
      var stack = state.stack;

      var length = data.length;

      for (var j = 0; j < length;) {
        var value = data[j++];
        var validationCommand = null;
        if (value === 12) {
          var q = data[j++];
          if (q === 0) {
            // The CFF specification state that the 'dotsection' command
            // (12, 0) is deprecated and treated as a no-op, but all Type2
            // charstrings processors should support them. Unfortunately
            // the font sanitizer don't. As a workaround the sequence (12, 0)
            // is replaced by a useless (0, hmoveto).
            data[j - 2] = 139;
            data[j - 1] = 22;
            stackSize = 0;
          } else {
            validationCommand = CharstringValidationData12[q];
          }
        } else if (value === 28) { // number (16 bit)
          stack[stackSize] = ((data[j] << 24) | (data[j + 1] << 16)) >> 16;
          j += 2;
          stackSize++;
        } else if (value === 14) {
          if (stackSize >= 4) {
            stackSize -= 4;
            if (this.seacAnalysisEnabled) {
              state.seac = stack.slice(stackSize, stackSize + 4);
              return false;
            }
          }
          validationCommand = CharstringValidationData[value];
        } else if (value >= 32 && value <= 246) {  // number
          stack[stackSize] = value - 139;
          stackSize++;
        } else if (value >= 247 && value <= 254) {  // number (+1 bytes)
          stack[stackSize] = (value < 251 ?
                              ((value - 247) << 8) + data[j] + 108 :
                              -((value - 251) << 8) - data[j] - 108);
          j++;
          stackSize++;
        } else if (value === 255) {  // number (32 bit)
          stack[stackSize] = ((data[j] << 24) | (data[j + 1] << 16) |
                              (data[j + 2] << 8) | data[j + 3]) / 65536;
          j += 4;
          stackSize++;
        } else if (value === 19 || value === 20) {
          state.hints += stackSize >> 1;
          // skipping right amount of hints flag data
          j += (state.hints + 7) >> 3;
          stackSize %= 2;
          validationCommand = CharstringValidationData[value];
        } else if (value === 10 || value === 29) {
          var subrsIndex;
          if (value === 10) {
            subrsIndex = localSubrIndex;
          } else {
            subrsIndex = globalSubrIndex;
          }
          if (!subrsIndex) {
            validationCommand = CharstringValidationData[value];
            warn('Missing subrsIndex for ' + validationCommand.id);
            return false;
          }
          var bias = 32768;
          if (subrsIndex.count < 1240) {
            bias = 107;
          } else if (subrsIndex.count < 33900) {
            bias = 1131;
          }
          var subrNumber = stack[--stackSize] + bias;
          if (subrNumber < 0 || subrNumber >= subrsIndex.count ||
              isNaN(subrNumber)) {
            validationCommand = CharstringValidationData[value];
            warn('Out of bounds subrIndex for ' + validationCommand.id);
            return false;
          }
          state.stackSize = stackSize;
          state.callDepth++;
          var valid = this.parseCharString(state, subrsIndex.get(subrNumber),
                                           localSubrIndex, globalSubrIndex);
          if (!valid) {
            return false;
          }
          state.callDepth--;
          stackSize = state.stackSize;
          continue;
        } else if (value === 11) {
          state.stackSize = stackSize;
          return true;
        } else {
          validationCommand = CharstringValidationData[value];
        }
        if (validationCommand) {
          if (validationCommand.stem) {
            state.hints += stackSize >> 1;
            if (value === 3 || value === 23) {
              // vstem or vstemhm.
              state.hasVStems = true;
            } else if (state.hasVStems && (value === 1 || value === 18)) {
              // Some browsers don't draw glyphs that specify vstems before
              // hstems. As a workaround, replace hstem (1) and hstemhm (18)
              // with a pointless vstem (3) or vstemhm (23).
              warn('CFF stem hints are in wrong order');
              data[j - 1] = (value === 1) ? 3 : 23;
            }
          }
          if ('min' in validationCommand) {
            if (!state.undefStack && stackSize < validationCommand.min) {
              warn('Not enough parameters for ' + validationCommand.id +
                   '; actual: ' + stackSize +
                   ', expected: ' + validationCommand.min);
              return false;
            }
          }
          if (state.firstStackClearing && validationCommand.stackClearing) {
            state.firstStackClearing = false;
            // the optional character width can be found before the first
            // stack-clearing command arguments
            stackSize -= validationCommand.min;
            if (stackSize >= 2 && validationCommand.stem) {
              // there are even amount of arguments for stem commands
              stackSize %= 2;
            } else if (stackSize > 1) {
              warn('Found too many parameters for stack-clearing command');
            }
            if (stackSize > 0 && stack[stackSize - 1] >= 0) {
              state.width = stack[stackSize - 1];
            }
          }
          if ('stackDelta' in validationCommand) {
            if ('stackFn' in validationCommand) {
              validationCommand.stackFn(stack, stackSize);
            }
            stackSize += validationCommand.stackDelta;
          } else if (validationCommand.stackClearing) {
            stackSize = 0;
          } else if (validationCommand.resetStack) {
            stackSize = 0;
            state.undefStack = false;
          } else if (validationCommand.undefStack) {
            stackSize = 0;
            state.undefStack = true;
            state.firstStackClearing = false;
          }
        }
      }
      state.stackSize = stackSize;
      return true;
    },
    parseCharStrings({ charStrings, localSubrIndex, globalSubrIndex, fdSelect,
                       fdArray, privateDict, }) {
      var seacs = [];
      var widths = [];
      var count = charStrings.count;
      for (var i = 0; i < count; i++) {
        var charstring = charStrings.get(i);
        var state = {
          callDepth: 0,
          stackSize: 0,
          stack: [],
          undefStack: true,
          hints: 0,
          firstStackClearing: true,
          seac: null,
          width: null,
          hasVStems: false,
        };
        var valid = true;
        var localSubrToUse = null;
        var privateDictToUse = privateDict;
        if (fdSelect && fdArray.length) {
          var fdIndex = fdSelect.getFDIndex(i);
          if (fdIndex === -1) {
            warn('Glyph index is not in fd select.');
            valid = false;
          }
          if (fdIndex >= fdArray.length) {
            warn('Invalid fd index for glyph index.');
            valid = false;
          }
          if (valid) {
            privateDictToUse = fdArray[fdIndex].privateDict;
            localSubrToUse = privateDictToUse.subrsIndex;
          }
        } else if (localSubrIndex) {
          localSubrToUse = localSubrIndex;
        }
        if (valid) {
          valid = this.parseCharString(state, charstring, localSubrToUse,
                                       globalSubrIndex);
        }
        if (state.width !== null) {
          const nominalWidth = privateDictToUse.getByName('nominalWidthX');
          widths[i] = nominalWidth + state.width;
        } else {
          const defaultWidth = privateDictToUse.getByName('defaultWidthX');
          widths[i] = defaultWidth;
        }
        if (state.seac !== null) {
          seacs[i] = state.seac;
        }
        if (!valid) {
          // resetting invalid charstring to single 'endchar'
          charStrings.set(i, new Uint8Array([14]));
        }
      }
      return { charStrings, seacs, widths, };
    },
    emptyPrivateDictionary:
      function CFFParser_emptyPrivateDictionary(parentDict) {
      var privateDict = this.createDict(CFFPrivateDict, [],
                                        parentDict.strings);
      parentDict.setByKey(18, [0, 0]);
      parentDict.privateDict = privateDict;
    },
    parsePrivateDict: function CFFParser_parsePrivateDict(parentDict) {
      // no private dict, do nothing
      if (!parentDict.hasName('Private')) {
        this.emptyPrivateDictionary(parentDict);
        return;
      }
      var privateOffset = parentDict.getByName('Private');
      // make sure the params are formatted correctly
      if (!Array.isArray(privateOffset) || privateOffset.length !== 2) {
        parentDict.removeByName('Private');
        return;
      }
      var size = privateOffset[0];
      var offset = privateOffset[1];
      // remove empty dicts or ones that refer to invalid location
      if (size === 0 || offset >= this.bytes.length) {
        this.emptyPrivateDictionary(parentDict);
        return;
      }

      var privateDictEnd = offset + size;
      var dictData = this.bytes.subarray(offset, privateDictEnd);
      var dict = this.parseDict(dictData);
      var privateDict = this.createDict(CFFPrivateDict, dict,
                                        parentDict.strings);
      parentDict.privateDict = privateDict;

      // Parse the Subrs index also since it's relative to the private dict.
      if (!privateDict.getByName('Subrs')) {
        return;
      }
      var subrsOffset = privateDict.getByName('Subrs');
      var relativeOffset = offset + subrsOffset;
      // Validate the offset.
      if (subrsOffset === 0 || relativeOffset >= this.bytes.length) {
        this.emptyPrivateDictionary(parentDict);
        return;
      }
      var subrsIndex = this.parseIndex(relativeOffset);
      privateDict.subrsIndex = subrsIndex.obj;
    },
    parseCharsets: function CFFParser_parseCharsets(pos, length, strings, cid) {
      if (pos === 0) {
        return new CFFCharset(true, CFFCharsetPredefinedTypes.ISO_ADOBE,
                              ISOAdobeCharset);
      } else if (pos === 1) {
        return new CFFCharset(true, CFFCharsetPredefinedTypes.EXPERT,
                              ExpertCharset);
      } else if (pos === 2) {
        return new CFFCharset(true, CFFCharsetPredefinedTypes.EXPERT_SUBSET,
                              ExpertSubsetCharset);
      }

      var bytes = this.bytes;
      var start = pos;
      var format = bytes[pos++];
      var charset = ['.notdef'];
      var id, count, i;

      // subtract 1 for the .notdef glyph
      length -= 1;

      switch (format) {
        case 0:
          for (i = 0; i < length; i++) {
            id = (bytes[pos++] << 8) | bytes[pos++];
            charset.push(cid ? id : strings.get(id));
          }
          break;
        case 1:
          while (charset.length <= length) {
            id = (bytes[pos++] << 8) | bytes[pos++];
            count = bytes[pos++];
            for (i = 0; i <= count; i++) {
              charset.push(cid ? id++ : strings.get(id++));
            }
          }
          break;
        case 2:
          while (charset.length <= length) {
            id = (bytes[pos++] << 8) | bytes[pos++];
            count = (bytes[pos++] << 8) | bytes[pos++];
            for (i = 0; i <= count; i++) {
              charset.push(cid ? id++ : strings.get(id++));
            }
          }
          break;
        default:
          throw new FormatError('Unknown charset format');
      }
      // Raw won't be needed if we actually compile the charset.
      var end = pos;
      var raw = bytes.subarray(start, end);

      return new CFFCharset(false, format, charset, raw);
    },
    parseEncoding: function CFFParser_parseEncoding(pos,
                                                    properties,
                                                    strings,
                                                    charset) {
      var encoding = Object.create(null);
      var bytes = this.bytes;
      var predefined = false;
      var format, i, ii;
      var raw = null;

      function readSupplement() {
        var supplementsCount = bytes[pos++];
        for (i = 0; i < supplementsCount; i++) {
          var code = bytes[pos++];
          var sid = (bytes[pos++] << 8) + (bytes[pos++] & 0xff);
          encoding[code] = charset.indexOf(strings.get(sid));
        }
      }

      if (pos === 0 || pos === 1) {
        predefined = true;
        format = pos;
        var baseEncoding = pos ? ExpertEncoding : StandardEncoding;
        for (i = 0, ii = charset.length; i < ii; i++) {
          var index = baseEncoding.indexOf(charset[i]);
          if (index !== -1) {
            encoding[index] = i;
          }
        }
      } else {
        var dataStart = pos;
        format = bytes[pos++];
        switch (format & 0x7f) {
          case 0:
            var glyphsCount = bytes[pos++];
            for (i = 1; i <= glyphsCount; i++) {
              encoding[bytes[pos++]] = i;
            }
            break;

          case 1:
            var rangesCount = bytes[pos++];
            var gid = 1;
            for (i = 0; i < rangesCount; i++) {
              var start = bytes[pos++];
              var left = bytes[pos++];
              for (var j = start; j <= start + left; j++) {
                encoding[j] = gid++;
              }
            }
            break;

          default:
            throw new FormatError(`Unknown encoding format: ${format} in CFF`);
        }
        var dataEnd = pos;
        if (format & 0x80) { // hasSupplement
          // The font sanitizer does not support CFF encoding with a
          // supplement, since the encoding is not really used to map
          // between gid to glyph, let's overwrite what is declared in
          // the top dictionary to let the sanitizer think the font use
          // StandardEncoding, that's a lie but that's ok.
          bytes[dataStart] &= 0x7f;
          readSupplement();
        }
        raw = bytes.subarray(dataStart, dataEnd);
      }
      format = format & 0x7f;
      return new CFFEncoding(predefined, format, encoding, raw);
    },
    parseFDSelect: function CFFParser_parseFDSelect(pos, length) {
      var bytes = this.bytes;
      var format = bytes[pos++];
      var fdSelect = [];
      var i;

      switch (format) {
        case 0:
          for (i = 0; i < length; ++i) {
            var id = bytes[pos++];
            fdSelect.push(id);
          }
          break;
        case 3:
          var rangesCount = (bytes[pos++] << 8) | bytes[pos++];
          for (i = 0; i < rangesCount; ++i) {
            var first = (bytes[pos++] << 8) | bytes[pos++];
            if (i === 0 && first !== 0) {
              warn('parseFDSelect: The first range must have a first GID of 0' +
                   ' -- trying to recover.');
              first = 0;
            }
            var fdIndex = bytes[pos++];
            var next = (bytes[pos] << 8) | bytes[pos + 1];
            for (var j = first; j < next; ++j) {
              fdSelect.push(fdIndex);
            }
          }
          // Advance past the sentinel(next).
          pos += 2;
          break;
        default:
          throw new FormatError(`parseFDSelect: Unknown format "${format}".`);
      }
      if (fdSelect.length !== length) {
        throw new FormatError('parseFDSelect: Invalid font data.');
      }

      return new CFFFDSelect(format, fdSelect);
    },
  };
  return CFFParser;
})();

// Compact Font Format
var CFF = (function CFFClosure() {
  function CFF() {
    this.header = null;
    this.names = [];
    this.topDict = null;
    this.strings = new CFFStrings();
    this.globalSubrIndex = null;

    // The following could really be per font, but since we only have one font
    // store them here.
    this.encoding = null;
    this.charset = null;
    this.charStrings = null;
    this.fdArray = [];
    this.fdSelect = null;

    this.isCIDFont = false;
  }
  CFF.prototype = {
    duplicateFirstGlyph: function CFF_duplicateFirstGlyph() {
      // Browsers will not display a glyph at position 0. Typically glyph 0 is
      // notdef, but a number of fonts put a valid glyph there so it must be
      // duplicated and appended.
      if (this.charStrings.count >= 65535) {
        warn('Not enough space in charstrings to duplicate first glyph.');
        return;
      }
      var glyphZero = this.charStrings.get(0);
      this.charStrings.add(glyphZero);
      if (this.isCIDFont) {
        this.fdSelect.fdSelect.push(this.fdSelect.fdSelect[0]);
      }
    },
    hasGlyphId: function CFF_hasGlyphID(id) {
      if (id < 0 || id >= this.charStrings.count) {
        return false;
      }
      var glyph = this.charStrings.get(id);
      return glyph.length > 0;
    },
  };

  return CFF;
})();

var CFFHeader = (function CFFHeaderClosure() {
  function CFFHeader(major, minor, hdrSize, offSize) {
    this.major = major;
    this.minor = minor;
    this.hdrSize = hdrSize;
    this.offSize = offSize;
  }
  return CFFHeader;
})();

var CFFStrings = (function CFFStringsClosure() {
  function CFFStrings() {
    this.strings = [];
  }
  CFFStrings.prototype = {
    get: function CFFStrings_get(index) {
      if (index >= 0 && index <= 390) {
        return CFFStandardStrings[index];
      }
      if (index - 391 <= this.strings.length) {
        return this.strings[index - 391];
      }
      return CFFStandardStrings[0];
    },
    add: function CFFStrings_add(value) {
      this.strings.push(value);
    },
    get count() {
      return this.strings.length;
    },
  };
  return CFFStrings;
})();

var CFFIndex = (function CFFIndexClosure() {
  function CFFIndex() {
    this.objects = [];
    this.length = 0;
  }
  CFFIndex.prototype = {
    add: function CFFIndex_add(data) {
      this.length += data.length;
      this.objects.push(data);
    },
    set: function CFFIndex_set(index, data) {
      this.length += data.length - this.objects[index].length;
      this.objects[index] = data;
    },
    get: function CFFIndex_get(index) {
      return this.objects[index];
    },
    get count() {
      return this.objects.length;
    },
  };
  return CFFIndex;
})();

var CFFDict = (function CFFDictClosure() {
  function CFFDict(tables, strings) {
    this.keyToNameMap = tables.keyToNameMap;
    this.nameToKeyMap = tables.nameToKeyMap;
    this.defaults = tables.defaults;
    this.types = tables.types;
    this.opcodes = tables.opcodes;
    this.order = tables.order;
    this.strings = strings;
    this.values = Object.create(null);
  }
  CFFDict.prototype = {
    // value should always be an array
    setByKey: function CFFDict_setByKey(key, value) {
      if (!(key in this.keyToNameMap)) {
        return false;
      }
      var valueLength = value.length;
      // ignore empty values
      if (valueLength === 0) {
        return true;
      }
      // Ignore invalid values (fixes bug1068432.pdf and bug1308536.pdf).
      for (var i = 0; i < valueLength; i++) {
        if (isNaN(value[i])) {
          warn('Invalid CFFDict value: "' + value + '" for key "' + key + '".');
          return true;
        }
      }
      var type = this.types[key];
      // remove the array wrapping these types of values
      if (type === 'num' || type === 'sid' || type === 'offset') {
        value = value[0];
      }
      this.values[key] = value;
      return true;
    },
    setByName: function CFFDict_setByName(name, value) {
      if (!(name in this.nameToKeyMap)) {
        throw new FormatError(`Invalid dictionary name "${name}"`);
      }
      this.values[this.nameToKeyMap[name]] = value;
    },
    hasName: function CFFDict_hasName(name) {
      return this.nameToKeyMap[name] in this.values;
    },
    getByName: function CFFDict_getByName(name) {
      if (!(name in this.nameToKeyMap)) {
        throw new FormatError(`Invalid dictionary name ${name}"`);
      }
      var key = this.nameToKeyMap[name];
      if (!(key in this.values)) {
        return this.defaults[key];
      }
      return this.values[key];
    },
    removeByName: function CFFDict_removeByName(name) {
      delete this.values[this.nameToKeyMap[name]];
    },
  };
  CFFDict.createTables = function CFFDict_createTables(layout) {
    var tables = {
      keyToNameMap: {},
      nameToKeyMap: {},
      defaults: {},
      types: {},
      opcodes: {},
      order: [],
    };
    for (var i = 0, ii = layout.length; i < ii; ++i) {
      var entry = layout[i];
      var key = Array.isArray(entry[0]) ?
                (entry[0][0] << 8) + entry[0][1] : entry[0];
      tables.keyToNameMap[key] = entry[1];
      tables.nameToKeyMap[entry[1]] = key;
      tables.types[key] = entry[2];
      tables.defaults[key] = entry[3];
      tables.opcodes[key] = Array.isArray(entry[0]) ? entry[0] : [entry[0]];
      tables.order.push(key);
    }
    return tables;
  };
  return CFFDict;
})();

var CFFTopDict = (function CFFTopDictClosure() {
  var layout = [
    [[12, 30], 'ROS', ['sid', 'sid', 'num'], null],
    [[12, 20], 'SyntheticBase', 'num', null],
    [0, 'version', 'sid', null],
    [1, 'Notice', 'sid', null],
    [[12, 0], 'Copyright', 'sid', null],
    [2, 'FullName', 'sid', null],
    [3, 'FamilyName', 'sid', null],
    [4, 'Weight', 'sid', null],
    [[12, 1], 'isFixedPitch', 'num', 0],
    [[12, 2], 'ItalicAngle', 'num', 0],
    [[12, 3], 'UnderlinePosition', 'num', -100],
    [[12, 4], 'UnderlineThickness', 'num', 50],
    [[12, 5], 'PaintType', 'num', 0],
    [[12, 6], 'CharstringType', 'num', 2],
    [[12, 7], 'FontMatrix', ['num', 'num', 'num', 'num', 'num', 'num'],
                            [0.001, 0, 0, 0.001, 0, 0]],
    [13, 'UniqueID', 'num', null],
    [5, 'FontBBox', ['num', 'num', 'num', 'num'], [0, 0, 0, 0]],
    [[12, 8], 'StrokeWidth', 'num', 0],
    [14, 'XUID', 'array', null],
    [15, 'charset', 'offset', 0],
    [16, 'Encoding', 'offset', 0],
    [17, 'CharStrings', 'offset', 0],
    [18, 'Private', ['offset', 'offset'], null],
    [[12, 21], 'PostScript', 'sid', null],
    [[12, 22], 'BaseFontName', 'sid', null],
    [[12, 23], 'BaseFontBlend', 'delta', null],
    [[12, 31], 'CIDFontVersion', 'num', 0],
    [[12, 32], 'CIDFontRevision', 'num', 0],
    [[12, 33], 'CIDFontType', 'num', 0],
    [[12, 34], 'CIDCount', 'num', 8720],
    [[12, 35], 'UIDBase', 'num', null],
    // XXX: CID Fonts on DirectWrite 6.1 only seem to work if FDSelect comes
    // before FDArray.
    [[12, 37], 'FDSelect', 'offset', null],
    [[12, 36], 'FDArray', 'offset', null],
    [[12, 38], 'FontName', 'sid', null]
  ];
  var tables = null;
  function CFFTopDict(strings) {
    if (tables === null) {
      tables = CFFDict.createTables(layout);
    }
    CFFDict.call(this, tables, strings);
    this.privateDict = null;
  }
  CFFTopDict.prototype = Object.create(CFFDict.prototype);
  return CFFTopDict;
})();

var CFFPrivateDict = (function CFFPrivateDictClosure() {
  var layout = [
    [6, 'BlueValues', 'delta', null],
    [7, 'OtherBlues', 'delta', null],
    [8, 'FamilyBlues', 'delta', null],
    [9, 'FamilyOtherBlues', 'delta', null],
    [[12, 9], 'BlueScale', 'num', 0.039625],
    [[12, 10], 'BlueShift', 'num', 7],
    [[12, 11], 'BlueFuzz', 'num', 1],
    [10, 'StdHW', 'num', null],
    [11, 'StdVW', 'num', null],
    [[12, 12], 'StemSnapH', 'delta', null],
    [[12, 13], 'StemSnapV', 'delta', null],
    [[12, 14], 'ForceBold', 'num', 0],
    [[12, 17], 'LanguageGroup', 'num', 0],
    [[12, 18], 'ExpansionFactor', 'num', 0.06],
    [[12, 19], 'initialRandomSeed', 'num', 0],
    [20, 'defaultWidthX', 'num', 0],
    [21, 'nominalWidthX', 'num', 0],
    [19, 'Subrs', 'offset', null]
  ];
  var tables = null;
  function CFFPrivateDict(strings) {
    if (tables === null) {
      tables = CFFDict.createTables(layout);
    }
    CFFDict.call(this, tables, strings);
    this.subrsIndex = null;
  }
  CFFPrivateDict.prototype = Object.create(CFFDict.prototype);
  return CFFPrivateDict;
})();

var CFFCharsetPredefinedTypes = {
  ISO_ADOBE: 0,
  EXPERT: 1,
  EXPERT_SUBSET: 2,
};
var CFFCharset = (function CFFCharsetClosure() {
  function CFFCharset(predefined, format, charset, raw) {
    this.predefined = predefined;
    this.format = format;
    this.charset = charset;
    this.raw = raw;
  }
  return CFFCharset;
})();

var CFFEncoding = (function CFFEncodingClosure() {
  function CFFEncoding(predefined, format, encoding, raw) {
    this.predefined = predefined;
    this.format = format;
    this.encoding = encoding;
    this.raw = raw;
  }
  return CFFEncoding;
})();

var CFFFDSelect = (function CFFFDSelectClosure() {
  function CFFFDSelect(format, fdSelect) {
    this.format = format;
    this.fdSelect = fdSelect;
  }
  CFFFDSelect.prototype = {
    getFDIndex: function CFFFDSelect_get(glyphIndex) {
      if (glyphIndex < 0 || glyphIndex >= this.fdSelect.length) {
        return -1;
      }
      return this.fdSelect[glyphIndex];
    },
  };
  return CFFFDSelect;
})();

// Helper class to keep track of where an offset is within the data and helps
// filling in that offset once it's known.
var CFFOffsetTracker = (function CFFOffsetTrackerClosure() {
  function CFFOffsetTracker() {
    this.offsets = Object.create(null);
  }
  CFFOffsetTracker.prototype = {
    isTracking: function CFFOffsetTracker_isTracking(key) {
      return key in this.offsets;
    },
    track: function CFFOffsetTracker_track(key, location) {
      if (key in this.offsets) {
        throw new FormatError(`Already tracking location of ${key}`);
      }
      this.offsets[key] = location;
    },
    offset: function CFFOffsetTracker_offset(value) {
      for (var key in this.offsets) {
        this.offsets[key] += value;
      }
    },
    setEntryLocation: function CFFOffsetTracker_setEntryLocation(key,
                                                                 values,
                                                                 output) {
      if (!(key in this.offsets)) {
        throw new FormatError(`Not tracking location of ${key}`);
      }
      var data = output.data;
      var dataOffset = this.offsets[key];
      var size = 5;
      for (var i = 0, ii = values.length; i < ii; ++i) {
        var offset0 = i * size + dataOffset;
        var offset1 = offset0 + 1;
        var offset2 = offset0 + 2;
        var offset3 = offset0 + 3;
        var offset4 = offset0 + 4;
        // It's easy to screw up offsets so perform this sanity check.
        if (data[offset0] !== 0x1d || data[offset1] !== 0 ||
            data[offset2] !== 0 || data[offset3] !== 0 || data[offset4] !== 0) {
          throw new FormatError('writing to an offset that is not empty');
        }
        var value = values[i];
        data[offset0] = 0x1d;
        data[offset1] = (value >> 24) & 0xFF;
        data[offset2] = (value >> 16) & 0xFF;
        data[offset3] = (value >> 8) & 0xFF;
        data[offset4] = value & 0xFF;
      }
    },
  };
  return CFFOffsetTracker;
})();

// Takes a CFF and converts it to the binary representation.
var CFFCompiler = (function CFFCompilerClosure() {
  function CFFCompiler(cff) {
    this.cff = cff;
  }
  CFFCompiler.prototype = {
    compile: function CFFCompiler_compile() {
      var cff = this.cff;
      var output = {
        data: [],
        length: 0,
        add: function CFFCompiler_add(data) {
          this.data = this.data.concat(data);
          this.length = this.data.length;
        },
      };

      // Compile the five entries that must be in order.
      var header = this.compileHeader(cff.header);
      output.add(header);

      var nameIndex = this.compileNameIndex(cff.names);
      output.add(nameIndex);

      if (cff.isCIDFont) {
        // The spec is unclear on how font matrices should relate to each other
        // when there is one in the main top dict and the sub top dicts.
        // Windows handles this differently than linux and osx so we have to
        // normalize to work on all.
        // Rules based off of some mailing list discussions:
        // - If main font has a matrix and subfont doesn't, use the main matrix.
        // - If no main font matrix and there is a subfont matrix, use the
        //   subfont matrix.
        // - If both have matrices, concat together.
        // - If neither have matrices, use default.
        // To make this work on all platforms we move the top matrix into each
        // sub top dict and concat if necessary.
        if (cff.topDict.hasName('FontMatrix')) {
          var base = cff.topDict.getByName('FontMatrix');
          cff.topDict.removeByName('FontMatrix');
          for (var i = 0, ii = cff.fdArray.length; i < ii; i++) {
            var subDict = cff.fdArray[i];
            var matrix = base.slice(0);
            if (subDict.hasName('FontMatrix')) {
              matrix = Util.transform(matrix, subDict.getByName('FontMatrix'));
            }
            subDict.setByName('FontMatrix', matrix);
          }
        }
      }

      cff.topDict.setByName('charset', 0);
      var compiled = this.compileTopDicts([cff.topDict],
                                          output.length,
                                          cff.isCIDFont);
      output.add(compiled.output);
      var topDictTracker = compiled.trackers[0];

      var stringIndex = this.compileStringIndex(cff.strings.strings);
      output.add(stringIndex);

      var globalSubrIndex = this.compileIndex(cff.globalSubrIndex);
      output.add(globalSubrIndex);

      // Now start on the other entries that have no specific order.
      if (cff.encoding && cff.topDict.hasName('Encoding')) {
        if (cff.encoding.predefined) {
          topDictTracker.setEntryLocation('Encoding', [cff.encoding.format],
                                          output);
        } else {
          var encoding = this.compileEncoding(cff.encoding);
          topDictTracker.setEntryLocation('Encoding', [output.length], output);
          output.add(encoding);
        }
      }
      var charset = this.compileCharset(cff.charset);
      topDictTracker.setEntryLocation('charset', [output.length], output);
      output.add(charset);

      var charStrings = this.compileCharStrings(cff.charStrings);
      topDictTracker.setEntryLocation('CharStrings', [output.length], output);
      output.add(charStrings);

      if (cff.isCIDFont) {
        // For some reason FDSelect must be in front of FDArray on windows. OSX
        // and linux don't seem to care.
        topDictTracker.setEntryLocation('FDSelect', [output.length], output);
        var fdSelect = this.compileFDSelect(cff.fdSelect);
        output.add(fdSelect);
        // It is unclear if the sub font dictionary can have CID related
        // dictionary keys, but the sanitizer doesn't like them so remove them.
        compiled = this.compileTopDicts(cff.fdArray, output.length, true);
        topDictTracker.setEntryLocation('FDArray', [output.length], output);
        output.add(compiled.output);
        var fontDictTrackers = compiled.trackers;

        this.compilePrivateDicts(cff.fdArray, fontDictTrackers, output);
      }

      this.compilePrivateDicts([cff.topDict], [topDictTracker], output);

      // If the font data ends with INDEX whose object data is zero-length,
      // the sanitizer will bail out. Add a dummy byte to avoid that.
      output.add([0]);

      return output.data;
    },
    encodeNumber: function CFFCompiler_encodeNumber(value) {
      if (parseFloat(value) === parseInt(value, 10) && !isNaN(value)) { // isInt
        return this.encodeInteger(value);
      }
      return this.encodeFloat(value);
    },
    encodeFloat: function CFFCompiler_encodeFloat(num) {
      var value = num.toString();

      // rounding inaccurate doubles
      var m = /\.(\d*?)(?:9{5,20}|0{5,20})\d{0,2}(?:e(.+)|$)/.exec(value);
      if (m) {
        var epsilon = parseFloat('1e' + ((m[2] ? +m[2] : 0) + m[1].length));
        value = (Math.round(num * epsilon) / epsilon).toString();
      }

      var nibbles = '';
      var i, ii;
      for (i = 0, ii = value.length; i < ii; ++i) {
        var a = value[i];
        if (a === 'e') {
          nibbles += value[++i] === '-' ? 'c' : 'b';
        } else if (a === '.') {
          nibbles += 'a';
        } else if (a === '-') {
          nibbles += 'e';
        } else {
          nibbles += a;
        }
      }
      nibbles += (nibbles.length & 1) ? 'f' : 'ff';
      var out = [30];
      for (i = 0, ii = nibbles.length; i < ii; i += 2) {
        out.push(parseInt(nibbles.substring(i, i + 2), 16));
      }
      return out;
    },
    encodeInteger: function CFFCompiler_encodeInteger(value) {
      var code;
      if (value >= -107 && value <= 107) {
        code = [value + 139];
      } else if (value >= 108 && value <= 1131) {
        value = value - 108;
        code = [(value >> 8) + 247, value & 0xFF];
      } else if (value >= -1131 && value <= -108) {
        value = -value - 108;
        code = [(value >> 8) + 251, value & 0xFF];
      } else if (value >= -32768 && value <= 32767) {
        code = [0x1c, (value >> 8) & 0xFF, value & 0xFF];
      } else {
        code = [0x1d,
                (value >> 24) & 0xFF,
                (value >> 16) & 0xFF,
                (value >> 8) & 0xFF,
                 value & 0xFF];
      }
      return code;
    },
    compileHeader: function CFFCompiler_compileHeader(header) {
      return [
        header.major,
        header.minor,
        header.hdrSize,
        header.offSize
      ];
    },
    compileNameIndex: function CFFCompiler_compileNameIndex(names) {
      var nameIndex = new CFFIndex();
      for (var i = 0, ii = names.length; i < ii; ++i) {
        var name = names[i];
        // OTS doesn't allow names to be over 127 characters.
        var length = Math.min(name.length, 127);
        var sanitizedName = new Array(length);
        for (var j = 0; j < length; j++) {
          // OTS requires chars to be between a range and not certain other
          // chars.
          var char = name[j];
          if (char < '!' || char > '~' || char === '[' || char === ']' ||
              char === '(' || char === ')' || char === '{' || char === '}' ||
              char === '<' || char === '>' || char === '/' || char === '%') {
            char = '_';
          }
          sanitizedName[j] = char;
        }
        sanitizedName = sanitizedName.join('');

        if (sanitizedName === '') {
          sanitizedName = 'Bad_Font_Name';
        }
        nameIndex.add(stringToBytes(sanitizedName));
      }
      return this.compileIndex(nameIndex);
    },
    compileTopDicts: function CFFCompiler_compileTopDicts(dicts,
                                                          length,
                                                          removeCidKeys) {
      var fontDictTrackers = [];
      var fdArrayIndex = new CFFIndex();
      for (var i = 0, ii = dicts.length; i < ii; ++i) {
        var fontDict = dicts[i];
        if (removeCidKeys) {
          fontDict.removeByName('CIDFontVersion');
          fontDict.removeByName('CIDFontRevision');
          fontDict.removeByName('CIDFontType');
          fontDict.removeByName('CIDCount');
          fontDict.removeByName('UIDBase');
        }
        var fontDictTracker = new CFFOffsetTracker();
        var fontDictData = this.compileDict(fontDict, fontDictTracker);
        fontDictTrackers.push(fontDictTracker);
        fdArrayIndex.add(fontDictData);
        fontDictTracker.offset(length);
      }
      fdArrayIndex = this.compileIndex(fdArrayIndex, fontDictTrackers);
      return {
        trackers: fontDictTrackers,
        output: fdArrayIndex,
      };
    },
    compilePrivateDicts: function CFFCompiler_compilePrivateDicts(dicts,
                                                                  trackers,
                                                                  output) {
      for (var i = 0, ii = dicts.length; i < ii; ++i) {
        var fontDict = dicts[i];
        var privateDict = fontDict.privateDict;
        if (!privateDict || !fontDict.hasName('Private')) {
          throw new FormatError('There must be a private dictionary.');
        }
        var privateDictTracker = new CFFOffsetTracker();
        var privateDictData = this.compileDict(privateDict, privateDictTracker);

        var outputLength = output.length;
        privateDictTracker.offset(outputLength);
        if (!privateDictData.length) {
          // The private dictionary was empty, set the output length to zero to
          // ensure the offset length isn't out of bounds in the eyes of the
          // sanitizer.
          outputLength = 0;
        }

        trackers[i].setEntryLocation('Private',
                                     [privateDictData.length, outputLength],
                                     output);
        output.add(privateDictData);

        if (privateDict.subrsIndex && privateDict.hasName('Subrs')) {
          var subrs = this.compileIndex(privateDict.subrsIndex);
          privateDictTracker.setEntryLocation('Subrs', [privateDictData.length],
                                              output);
          output.add(subrs);
        }
      }
    },
    compileDict: function CFFCompiler_compileDict(dict, offsetTracker) {
      var out = [];
      // The dictionary keys must be in a certain order.
      var order = dict.order;
      for (var i = 0; i < order.length; ++i) {
        var key = order[i];
        if (!(key in dict.values)) {
          continue;
        }
        var values = dict.values[key];
        var types = dict.types[key];
        if (!Array.isArray(types)) {
          types = [types];
        }
        if (!Array.isArray(values)) {
          values = [values];
        }

        // Remove any empty dict values.
        if (values.length === 0) {
          continue;
        }

        for (var j = 0, jj = types.length; j < jj; ++j) {
          var type = types[j];
          var value = values[j];
          switch (type) {
            case 'num':
            case 'sid':
              out = out.concat(this.encodeNumber(value));
              break;
            case 'offset':
              // For offsets we just insert a 32bit integer so we don't have to
              // deal with figuring out the length of the offset when it gets
              // replaced later on by the compiler.
              var name = dict.keyToNameMap[key];
              // Some offsets have the offset and the length, so just record the
              // position of the first one.
              if (!offsetTracker.isTracking(name)) {
                offsetTracker.track(name, out.length);
              }
              out = out.concat([0x1d, 0, 0, 0, 0]);
              break;
            case 'array':
            case 'delta':
              out = out.concat(this.encodeNumber(value));
              for (var k = 1, kk = values.length; k < kk; ++k) {
                out = out.concat(this.encodeNumber(values[k]));
              }
              break;
            default:
              throw new FormatError(`Unknown data type of ${type}`);
          }
        }
        out = out.concat(dict.opcodes[key]);
      }
      return out;
    },
    compileStringIndex: function CFFCompiler_compileStringIndex(strings) {
      var stringIndex = new CFFIndex();
      for (var i = 0, ii = strings.length; i < ii; ++i) {
        stringIndex.add(stringToBytes(strings[i]));
      }
      return this.compileIndex(stringIndex);
    },
    compileGlobalSubrIndex: function CFFCompiler_compileGlobalSubrIndex() {
      var globalSubrIndex = this.cff.globalSubrIndex;
      this.out.writeByteArray(this.compileIndex(globalSubrIndex));
    },
    compileCharStrings: function CFFCompiler_compileCharStrings(charStrings) {
      var charStringsIndex = new CFFIndex();
      for (var i = 0; i < charStrings.count; i++) {
        var glyph = charStrings.get(i);
        // If the CharString outline is empty, replace it with .notdef to
        // prevent OTS from rejecting the font (fixes bug1252420.pdf).
        if (glyph.length === 0) {
          charStringsIndex.add(new Uint8Array([0x8B, 0x0E]));
          continue;
        }
        charStringsIndex.add(glyph);
      }
      return this.compileIndex(charStringsIndex);
    },
    compileCharset: function CFFCompiler_compileCharset(charset) {
      let length = 1 + (this.cff.charStrings.count - 1) * 2;
      // The contents of the charset doesn't matter, it's just there to make
      // freetype happy.
      let out = new Uint8Array(length);
      return this.compileTypedArray(out);
    },
    compileEncoding: function CFFCompiler_compileEncoding(encoding) {
      return this.compileTypedArray(encoding.raw);
    },
    compileFDSelect: function CFFCompiler_compileFDSelect(fdSelect) {
      let format = fdSelect.format;
      let out, i;
      switch (format) {
        case 0:
          out = new Uint8Array(1 + fdSelect.fdSelect.length);
          out[0] = format;
          for (i = 0; i < fdSelect.fdSelect.length; i++) {
            out[i + 1] = fdSelect.fdSelect[i];
          }
          break;
        case 3:
          let start = 0;
          let lastFD = fdSelect.fdSelect[0];
          let ranges = [
            format,
            0, // nRanges place holder
            0, // nRanges place holder
            (start >> 8) & 0xFF,
            start & 0xFF,
            lastFD
          ];
          for (i = 1; i < fdSelect.fdSelect.length; i++) {
            let currentFD = fdSelect.fdSelect[i];
            if (currentFD !== lastFD) {
              ranges.push((i >> 8) & 0xFF, i & 0xFF, currentFD);
              lastFD = currentFD;
            }
          }
          // 3 bytes are pushed for every range and there are 3 header bytes.
          let numRanges = (ranges.length - 3) / 3;
          ranges[1] = (numRanges >> 8) & 0xFF;
          ranges[2] = numRanges & 0xFF;
          // sentinel
          ranges.push((i >> 8) & 0xFF, i & 0xFF);
          out = new Uint8Array(ranges);
          break;
      }
      return this.compileTypedArray(out);
    },
    compileTypedArray: function CFFCompiler_compileTypedArray(data) {
      var out = [];
      for (var i = 0, ii = data.length; i < ii; ++i) {
        out[i] = data[i];
      }
      return out;
    },
    compileIndex: function CFFCompiler_compileIndex(index, trackers) {
      trackers = trackers || [];
      var objects = index.objects;
      // First 2 bytes contains the number of objects contained into this index
      var count = objects.length;

      // If there is no object, just create an index. This technically
      // should just be [0, 0] but OTS has an issue with that.
      if (count === 0) {
        return [0, 0, 0];
      }

      var data = [(count >> 8) & 0xFF, count & 0xff];

      var lastOffset = 1, i;
      for (i = 0; i < count; ++i) {
        lastOffset += objects[i].length;
      }

      var offsetSize;
      if (lastOffset < 0x100) {
        offsetSize = 1;
      } else if (lastOffset < 0x10000) {
        offsetSize = 2;
      } else if (lastOffset < 0x1000000) {
        offsetSize = 3;
      } else {
        offsetSize = 4;
      }

      // Next byte contains the offset size use to reference object in the file
      data.push(offsetSize);

      // Add another offset after this one because we need a new offset
      var relativeOffset = 1;
      for (i = 0; i < count + 1; i++) {
        if (offsetSize === 1) {
          data.push(relativeOffset & 0xFF);
        } else if (offsetSize === 2) {
          data.push((relativeOffset >> 8) & 0xFF,
                     relativeOffset & 0xFF);
        } else if (offsetSize === 3) {
          data.push((relativeOffset >> 16) & 0xFF,
                    (relativeOffset >> 8) & 0xFF,
                     relativeOffset & 0xFF);
        } else {
          data.push((relativeOffset >>> 24) & 0xFF,
                    (relativeOffset >> 16) & 0xFF,
                    (relativeOffset >> 8) & 0xFF,
                     relativeOffset & 0xFF);
        }

        if (objects[i]) {
          relativeOffset += objects[i].length;
        }
      }

      for (i = 0; i < count; i++) {
        // Notify the tracker where the object will be offset in the data.
        if (trackers[i]) {
          trackers[i].offset(data.length);
        }
        for (var j = 0, jj = objects[i].length; j < jj; j++) {
          data.push(objects[i][j]);
        }
      }
      return data;
    },
  };
  return CFFCompiler;
})();

export {
  CFFStandardStrings,
  CFFParser,
  CFF,
  CFFHeader,
  CFFStrings,
  CFFIndex,
  CFFCharset,
  CFFTopDict,
  CFFPrivateDict,
  CFFCompiler,
  CFFFDSelect,
};