/*
 * $Id: BaseFont.java 4784 2011-03-15 08:33:00Z blowagie $
 *
 * This file is part of the iText (R) project.
 * Copyright (c) 1998-2011 1T3XT BVBA
 * Authors: Bruno Lowagie, Paulo Soares, et al.
 *
 * This program is free software; you can redistribute it and/or modify
 * it under the terms of the GNU Affero General Public License version 3
 * as published by the Free Software Foundation with the addition of the
 * following permission added to Section 15 as permitted in Section 7(a):
 * FOR ANY PART OF THE COVERED WORK IN WHICH THE COPYRIGHT IS OWNED BY 1T3XT,
 * 1T3XT DISCLAIMS THE WARRANTY OF NON INFRINGEMENT OF THIRD PARTY RIGHTS.
 *
 * This program is distributed in the hope that it will be useful, but
 * WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY
 * or FITNESS FOR A PARTICULAR PURPOSE.
 * See the GNU Affero General Public License for more details.
 * You should have received a copy of the GNU Affero General Public License
 * along with this program; if not, see http://www.gnu.org/licenses or write to
 * the Free Software Foundation, Inc., 51 Franklin Street, Fifth Floor,
 * Boston, MA, 02110-1301 USA, or download the license from the following URL:
 * http://itextpdf.com/terms-of-use/
 *
 * The interactive user interfaces in modified source and object code versions
 * of this program must display Appropriate Legal Notices, as required under
 * Section 5 of the GNU Affero General Public License.
 *
 * In accordance with Section 7(b) of the GNU Affero General Public License,
 * a covered work must retain the producer line in every PDF that is created
 * or manipulated using iText.
 *
 * You can be released from the requirements of the license by purchasing
 * a commercial license. Buying such a license is mandatory as soon as you
 * develop commercial activities involving the iText software without
 * disclosing the source code of your own applications.
 * These activities include: offering paid services to customers as an ASP,
 * serving PDFs on the fly in a web application, shipping iText with a closed
 * source product.
 *
 * For more information, please contact iText Software Corp. at this
 * address: sales@itextpdf.com
 */
package com.itextpdf.text.pdf;
import java.io.IOException;
import java.io.InputStream;
import java.util.ArrayList;
import java.util.HashMap;
import java.util.StringTokenizer;

import com.itextpdf.text.DocumentException;
import com.itextpdf.text.error_messages.MessageLocalization;

/**
 * Base class for the several font types supported
 *
 * @author Paulo Soares
 */

public abstract class BaseFont {

    /** This is a possible value of a base 14 type 1 font */
    public static final String COURIER = "Courier";

    /** This is a possible value of a base 14 type 1 font */
    public static final String COURIER_BOLD = "Courier-Bold";

    /** This is a possible value of a base 14 type 1 font */
    public static final String COURIER_OBLIQUE = "Courier-Oblique";

    /** This is a possible value of a base 14 type 1 font */
    public static final String COURIER_BOLDOBLIQUE = "Courier-BoldOblique";

    /** This is a possible value of a base 14 type 1 font */
    public static final String HELVETICA = "Helvetica";

    /** This is a possible value of a base 14 type 1 font */
    public static final String HELVETICA_BOLD = "Helvetica-Bold";

    /** This is a possible value of a base 14 type 1 font */
    public static final String HELVETICA_OBLIQUE = "Helvetica-Oblique";

    /** This is a possible value of a base 14 type 1 font */
    public static final String HELVETICA_BOLDOBLIQUE = "Helvetica-BoldOblique";

    /** This is a possible value of a base 14 type 1 font */
    public static final String SYMBOL = "Symbol";

    /** This is a possible value of a base 14 type 1 font */
    public static final String TIMES_ROMAN = "Times-Roman";

    /** This is a possible value of a base 14 type 1 font */
    public static final String TIMES_BOLD = "Times-Bold";

    /** This is a possible value of a base 14 type 1 font */
    public static final String TIMES_ITALIC = "Times-Italic";

    /** This is a possible value of a base 14 type 1 font */
    public static final String TIMES_BOLDITALIC = "Times-BoldItalic";

    /** This is a possible value of a base 14 type 1 font */
    public static final String ZAPFDINGBATS = "ZapfDingbats";

    /** The maximum height above the baseline reached by glyphs in this
     * font, excluding the height of glyphs for accented characters.
     */
    public static final int ASCENT = 1;
    /** The y coordinate of the top of flat capital letters, measured from
     * the baseline.
     */
    public static final int CAPHEIGHT = 2;
    /** The maximum depth below the baseline reached by glyphs in this
     * font. The value is a negative number.
     */
    public static final int DESCENT = 3;
    /** The angle, expressed in degrees counterclockwise from the vertical,
     * of the dominant vertical strokes of the font. The value is
     * negative for fonts that slope to the right, as almost all italic fonts do.
     */
    public static final int ITALICANGLE = 4;
    /** The lower left x glyph coordinate.
     */
    public static final int BBOXLLX = 5;
    /** The lower left y glyph coordinate.
     */
    public static final int BBOXLLY = 6;
    /** The upper right x glyph coordinate.
     */
    public static final int BBOXURX = 7;
    /** The upper right y glyph coordinate.
     */
    public static final int BBOXURY = 8;

    /** java.awt.Font property */
    public static final int AWT_ASCENT = 9;
    /** java.awt.Font property */
    public static final int AWT_DESCENT = 10;
    /** java.awt.Font property */
    public static final int AWT_LEADING = 11;
    /** java.awt.Font property */
    public static final int AWT_MAXADVANCE = 12;
    /**
     * The underline position. Usually a negative value.
     */
    public static final int UNDERLINE_POSITION = 13;
    /**
     * The underline thickness.
     */
    public static final int UNDERLINE_THICKNESS = 14;
    /**
     * The strikethrough position.
     */
    public static final int STRIKETHROUGH_POSITION = 15;
    /**
     * The strikethrough thickness.
     */
    public static final int STRIKETHROUGH_THICKNESS = 16;
    /**
     * The recommended vertical size for subscripts for this font.
     */
    public static final int SUBSCRIPT_SIZE = 17;
    /**
     * The recommended vertical offset from the baseline for subscripts for this font. Usually a negative value.
     */
    public static final int SUBSCRIPT_OFFSET = 18;
    /**
     * The recommended vertical size for superscripts for this font.
     */
    public static final int SUPERSCRIPT_SIZE = 19;
    /**
     * The recommended vertical offset from the baseline for superscripts for this font.
     */
    public static final int SUPERSCRIPT_OFFSET = 20;
    /**
     * The weight class of the font, as defined by the font author
     * @since 5.0.2
     */
    public static final int WEIGHT_CLASS = 21;
    /**
     * The width class of the font, as defined by the font author
     * @since 5.0.2
     */
    public static final int WIDTH_CLASS = 22;
    /** The font is Type 1.
     */
    public static final int FONT_TYPE_T1 = 0;
    /** The font is True Type with a standard encoding.
     */
    public static final int FONT_TYPE_TT = 1;
    /** The font is CJK.
     */
    public static final int FONT_TYPE_CJK = 2;
    /** The font is True Type with a Unicode encoding.
     */
    public static final int FONT_TYPE_TTUNI = 3;
    /** A font already inside the document.
     */
    public static final int FONT_TYPE_DOCUMENT = 4;
    /** A Type3 font.
     */
    public static final int FONT_TYPE_T3 = 5;
    /** The Unicode encoding with horizontal writing.
     */
    public static final String IDENTITY_H = "Identity-H";
    /** The Unicode encoding with vertical writing.
     */
    public static final String IDENTITY_V = "Identity-V";

    /** A possible encoding. */
    public static final String CP1250 = "Cp1250";

    /** A possible encoding. */
    public static final String CP1252 = "Cp1252";

    /** A possible encoding. */
    public static final String CP1257 = "Cp1257";

    /** A possible encoding. */
    public static final String WINANSI = "Cp1252";

    /** A possible encoding. */
    public static final String MACROMAN = "MacRoman";

    public static final int[] CHAR_RANGE_LATIN = {0, 0x17f, 0x2000, 0x206f, 0x20a0, 0x20cf, 0xfb00, 0xfb06};
    public static final int[] CHAR_RANGE_ARABIC = {0, 0x7f, 0x0600, 0x067f, 0x20a0, 0x20cf, 0xfb50, 0xfbff, 0xfe70, 0xfeff};
    public static final int[] CHAR_RANGE_HEBREW = {0, 0x7f, 0x0590, 0x05ff, 0x20a0, 0x20cf, 0xfb1d, 0xfb4f};
    public static final int[] CHAR_RANGE_CYRILLIC = {0, 0x7f, 0x0400, 0x052f, 0x2000, 0x206f, 0x20a0, 0x20cf};

/** if the font has to be embedded */
    public static final boolean EMBEDDED = true;

/** if the font doesn't have to be embedded */
    public static final boolean NOT_EMBEDDED = false;
/** if the font has to be cached */
    public static final boolean CACHED = true;
/** if the font doesn't have to be cached */
    public static final boolean NOT_CACHED = false;

    /** The path to the font resources. */
    public static final String RESOURCE_PATH = "com/itextpdf/text/pdf/fonts/";
    /** The fake CID code that represents a newline. */
    public static final char CID_NEWLINE = '\u7fff';

    protected ArrayList<int[]> subsetRanges;
    /** The font type.
     */
    int fontType;
/** a not defined character in a custom PDF encoding */
    public static final String notdef = ".notdef";

/** table of characters widths for this encoding */
    protected int widths[] = new int[256];

/** encoding names */
    protected String differences[] = new String[256];
/** same as differences but with the unicode codes */
    protected char unicodeDifferences[] = new char[256];

    protected int charBBoxes[][] = new int[256][];
/** encoding used with this font */
    protected String encoding;

/** true if the font is to be embedded in the PDF */
    protected boolean embedded;

    /**
     * The compression level for the font stream.
     * @since   2.1.3
     */
    protected int compressionLevel = PdfStream.DEFAULT_COMPRESSION;

/**
 * true if the font must use its built in encoding. In that case the
 * <CODE>encoding</CODE> is only used to map a char to the position inside
 * the font, not to the expected char name.
 */
    protected boolean fontSpecific = true;

/** cache for the fonts already used. */
    protected static HashMap<String, BaseFont> fontCache = new HashMap<String, BaseFont>();

/** list of the 14 built in fonts. */
    protected static final HashMap<String, PdfName> BuiltinFonts14 = new HashMap<String, PdfName>();

    /** Forces the output of the width array. Only matters for the 14
     * built-in fonts.
     */
    protected boolean forceWidthsOutput = false;

    /** Converts <CODE>char</CODE> directly to <CODE>byte</CODE>
     * by casting.
     */
    protected boolean directTextToByte = false;

    /** Indicates if all the glyphs and widths for that particular
     * encoding should be included in the document.
     */
    protected boolean subset = true;

    protected boolean fastWinansi = false;

    /**
     * Custom encodings use this map to key the Unicode character
     * to the single byte code.
     */
    protected IntHashtable specialMap;

    static {
        BuiltinFonts14.put(COURIER, PdfName.COURIER);
        BuiltinFonts14.put(COURIER_BOLD, PdfName.COURIER_BOLD);
        BuiltinFonts14.put(COURIER_BOLDOBLIQUE, PdfName.COURIER_BOLDOBLIQUE);
        BuiltinFonts14.put(COURIER_OBLIQUE, PdfName.COURIER_OBLIQUE);
        BuiltinFonts14.put(HELVETICA, PdfName.HELVETICA);
        BuiltinFonts14.put(HELVETICA_BOLD, PdfName.HELVETICA_BOLD);
        BuiltinFonts14.put(HELVETICA_BOLDOBLIQUE, PdfName.HELVETICA_BOLDOBLIQUE);
        BuiltinFonts14.put(HELVETICA_OBLIQUE, PdfName.HELVETICA_OBLIQUE);
        BuiltinFonts14.put(SYMBOL, PdfName.SYMBOL);
        BuiltinFonts14.put(TIMES_ROMAN, PdfName.TIMES_ROMAN);
        BuiltinFonts14.put(TIMES_BOLD, PdfName.TIMES_BOLD);
        BuiltinFonts14.put(TIMES_BOLDITALIC, PdfName.TIMES_BOLDITALIC);
        BuiltinFonts14.put(TIMES_ITALIC, PdfName.TIMES_ITALIC);
        BuiltinFonts14.put(ZAPFDINGBATS, PdfName.ZAPFDINGBATS);
    }

    /** Generates the PDF stream with the Type1 and Truetype fonts returning
     * a PdfStream.
     */
    static class StreamFont extends PdfStream {

        /** Generates the PDF stream with the Type1 and Truetype fonts returning
         * a PdfStream.
         * @param contents the content of the stream
         * @param lengths an array of int that describes the several lengths of each part of the font
         * @param compressionLevel      the compression level of the Stream
         * @throws DocumentException error in the stream compression
         * @since       2.1.3 (replaces the constructor without param compressionLevel)
         */
        public StreamFont(byte contents[], int lengths[], int compressionLevel) throws DocumentException {
            try {
                bytes = contents;
                put(PdfName.LENGTH, new PdfNumber(bytes.length));
                for (int k = 0; k < lengths.length; ++k) {
                    put(new PdfName("Length" + (k + 1)), new PdfNumber(lengths[k]));
                }
                flateCompress(compressionLevel);
            }
            catch (Exception e) {
                throw new DocumentException(e);
            }
        }

        /**
         * Generates the PDF stream for a font.
         * @param contents the content of a stream
         * @param subType the subtype of the font.
                 * @param compressionLevel      the compression level of the Stream
         * @throws DocumentException error in the stream compression
         * @since       2.1.3 (replaces the constructor without param compressionLevel)
         */
        public StreamFont(byte contents[], String subType, int compressionLevel) throws DocumentException {
            try {
                bytes = contents;
                put(PdfName.LENGTH, new PdfNumber(bytes.length));
                if (subType != null)
                    put(PdfName.SUBTYPE, new PdfName(subType));
                flateCompress(compressionLevel);
            }
            catch (Exception e) {
                throw new DocumentException(e);
            }
        }
    }

    /**
     *Creates new BaseFont
     */
    protected BaseFont() {
    }

    /**
     * Creates a new font. This will always be the default Helvetica font (not embedded).
     * This method is introduced because Helvetica is used in many examples.
     * @return  a BaseFont object (Helvetica, Winansi, not embedded)
     * @throws  IOException                     This shouldn't occur ever
     * @throws  DocumentException       This shouldn't occur ever
     * @since   2.1.1
     */
    public static BaseFont createFont() throws DocumentException, IOException {
        return createFont(BaseFont.HELVETICA, BaseFont.WINANSI, BaseFont.NOT_EMBEDDED);
    }

    /**
     * Creates a new font. This font can be one of the 14 built in types,
     * a Type1 font referred to by an AFM or PFM file, a TrueType font (simple or collection) or a CJK font from the
     * Adobe Asian Font Pack. TrueType fonts and CJK fonts can have an optional style modifier
     * appended to the name. These modifiers are: Bold, Italic and BoldItalic. An
     * example would be "STSong-Light,Bold". Note that this modifiers do not work if
     * the font is embedded. Fonts in TrueType collections are addressed by index such as "msgothic.ttc,1".
     * This would get the second font (indexes start at 0), in this case "MS PGothic".
     * <P>
     * The fonts are cached and if they already exist they are extracted from the cache,
     * not parsed again.
     * <P>
     * Besides the common encodings described by name, custom encodings
     * can also be made. These encodings will only work for the single byte fonts
     * Type1 and TrueType. The encoding string starts with a '#'
     * followed by "simple" or "full". If "simple" there is a decimal for the first character position and then a list
     * of hex values representing the Unicode codes that compose that encoding.<br>
     * The "simple" encoding is recommended for TrueType fonts
     * as the "full" encoding risks not matching the character with the right glyph
     * if not done with care.<br>
     * The "full" encoding is specially aimed at Type1 fonts where the glyphs have to be
     * described by non standard names like the Tex math fonts. Each group of three elements
     * compose a code position: the one byte code order in decimal or as 'x' (x cannot be the space), the name and the Unicode character
     * used to access the glyph. The space must be assigned to character position 32 otherwise
     * text justification will not work.
     * <P>
     * Example for a "simple" encoding that includes the Unicode
     * character space, A, B and ecyrillic:
     * <PRE>
     * "# simple 32 0020 0041 0042 0454"
     * </PRE>
     * <P>
     * Example for a "full" encoding for a Type1 Tex font:
     * <PRE>
     * "# full 'A' nottriangeqlleft 0041 'B' dividemultiply 0042 32 space 0020"
     * </PRE>
     * <P>
     * This method calls:<br>
     * <PRE>
     * createFont(name, encoding, embedded, true, null, null);
     * </PRE>
     * @param name the name of the font or its location on file
     * @param encoding the encoding to be applied to this font
     * @param embedded true if the font is to be embedded in the PDF
     * @return returns a new font. This font may come from the cache
     * @throws DocumentException the font is invalid
     * @throws IOException the font file could not be read
     */
    public static BaseFont createFont(String name, String encoding, boolean embedded) throws DocumentException, IOException {
        return createFont(name, encoding, embedded, true, null, null, false);
    }

    /**
     * Creates a new font. This font can be one of the 14 built in types,
     * a Type1 font referred to by an AFM or PFM file, a TrueType font (simple or collection) or a CJK font from the
     * Adobe Asian Font Pack. TrueType fonts and CJK fonts can have an optional style modifier
     * appended to the name. These modifiers are: Bold, Italic and BoldItalic. An
     * example would be "STSong-Light,Bold". Note that this modifiers do not work if
     * the font is embedded. Fonts in TrueType collections are addressed by index such as "msgothic.ttc,1".
     * This would get the second font (indexes start at 0), in this case "MS PGothic".
     * <P>
     * The fonts are cached and if they already exist they are extracted from the cache,
     * not parsed again.
     * <P>
     * Besides the common encodings described by name, custom encodings
     * can also be made. These encodings will only work for the single byte fonts
     * Type1 and TrueType. The encoding string starts with a '#'
     * followed by "simple" or "full". If "simple" there is a decimal for the first character position and then a list
     * of hex values representing the Unicode codes that compose that encoding.<br>
     * The "simple" encoding is recommended for TrueType fonts
     * as the "full" encoding risks not matching the character with the right glyph
     * if not done with care.<br>
     * The "full" encoding is specially aimed at Type1 fonts where the glyphs have to be
     * described by non standard names like the Tex math fonts. Each group of three elements
     * compose a code position: the one byte code order in decimal or as 'x' (x cannot be the space), the name and the Unicode character
     * used to access the glyph. The space must be assigned to character position 32 otherwise
     * text justification will not work.
     * <P>
     * Example for a "simple" encoding that includes the Unicode
     * character space, A, B and ecyrillic:
     * <PRE>
     * "# simple 32 0020 0041 0042 0454"
     * </PRE>
     * <P>
     * Example for a "full" encoding for a Type1 Tex font:
     * <PRE>
     * "# full 'A' nottriangeqlleft 0041 'B' dividemultiply 0042 32 space 0020"
     * </PRE>
     * <P>
     * This method calls:<br>
     * <PRE>
     * createFont(name, encoding, embedded, true, null, null);
     * </PRE>
     * @param name the name of the font or its location on file
     * @param encoding the encoding to be applied to this font
     * @param embedded true if the font is to be embedded in the PDF
     * @param   forceRead       in some cases (TrueTypeFont, Type1Font), the full font file will be read and kept in memory if forceRead is true
     * @return returns a new font. This font may come from the cache
     * @throws DocumentException the font is invalid
     * @throws IOException the font file could not be read
     * @since   2.1.5
     */
    public static BaseFont createFont(String name, String encoding, boolean embedded, boolean forceRead) throws DocumentException, IOException {
        return createFont(name, encoding, embedded, true, null, null, forceRead);
    }

    /** Creates a new font. This font can be one of the 14 built in types,
     * a Type1 font referred to by an AFM or PFM file, a TrueType font (simple or collection) or a CJK font from the
     * Adobe Asian Font Pack. TrueType fonts and CJK fonts can have an optional style modifier
     * appended to the name. These modifiers are: Bold, Italic and BoldItalic. An
     * example would be "STSong-Light,Bold". Note that this modifiers do not work if
     * the font is embedded. Fonts in TrueType collections are addressed by index such as "msgothic.ttc,1".
     * This would get the second font (indexes start at 0), in this case "MS PGothic".
     * <P>
     * The fonts may or may not be cached depending on the flag <CODE>cached</CODE>.
     * If the <CODE>byte</CODE> arrays are present the font will be
     * read from them instead of the name. A name is still required to identify
     * the font type.
     * <P>
     * Besides the common encodings described by name, custom encodings
     * can also be made. These encodings will only work for the single byte fonts
     * Type1 and TrueType. The encoding string starts with a '#'
     * followed by "simple" or "full". If "simple" there is a decimal for the first character position and then a list
     * of hex values representing the Unicode codes that compose that encoding.<br>
     * The "simple" encoding is recommended for TrueType fonts
     * as the "full" encoding risks not matching the character with the right glyph
     * if not done with care.<br>
     * The "full" encoding is specially aimed at Type1 fonts where the glyphs have to be
     * described by non standard names like the Tex math fonts. Each group of three elements
     * compose a code position: the one byte code order in decimal or as 'x' (x cannot be the space), the name and the Unicode character
     * used to access the glyph. The space must be assigned to character position 32 otherwise
     * text justification will not work.
     * <P>
     * Example for a "simple" encoding that includes the Unicode
     * character space, A, B and ecyrillic:
     * <PRE>
     * "# simple 32 0020 0041 0042 0454"
     * </PRE>
     * <P>
     * Example for a "full" encoding for a Type1 Tex font:
     * <PRE>
     * "# full 'A' nottriangeqlleft 0041 'B' dividemultiply 0042 32 space 0020"
     * </PRE>
     * @param name the name of the font or its location on file
     * @param encoding the encoding to be applied to this font
     * @param embedded true if the font is to be embedded in the PDF
     * @param cached true if the font comes from the cache or is added to
     * the cache if new, false if the font is always created new
     * @param ttfAfm the true type font or the afm in a byte array
     * @param pfb the pfb in a byte array
     * @return returns a new font. This font may come from the cache but only if cached
     * is true, otherwise it will always be created new
     * @throws DocumentException the font is invalid
     * @throws IOException the font file could not be read
     * @since   iText 0.80
     */
    public static BaseFont createFont(String name, String encoding, boolean embedded, boolean cached, byte ttfAfm[], byte pfb[]) throws DocumentException, IOException {
        return createFont(name, encoding, embedded, cached, ttfAfm, pfb, false);
    }

    /** Creates a new font. This font can be one of the 14 built in types,
     * a Type1 font referred to by an AFM or PFM file, a TrueType font (simple or collection) or a CJK font from the
     * Adobe Asian Font Pack. TrueType fonts and CJK fonts can have an optional style modifier
     * appended to the name. These modifiers are: Bold, Italic and BoldItalic. An
     * example would be "STSong-Light,Bold". Note that this modifiers do not work if
     * the font is embedded. Fonts in TrueType collections are addressed by index such as "msgothic.ttc,1".
     * This would get the second font (indexes start at 0), in this case "MS PGothic".
     * <P>
     * The fonts may or may not be cached depending on the flag <CODE>cached</CODE>.
     * If the <CODE>byte</CODE> arrays are present the font will be
     * read from them instead of the name. A name is still required to identify
     * the font type.
     * <P>
     * Besides the common encodings described by name, custom encodings
     * can also be made. These encodings will only work for the single byte fonts
     * Type1 and TrueType. The encoding string starts with a '#'
     * followed by "simple" or "full". If "simple" there is a decimal for the first character position and then a list
     * of hex values representing the Unicode codes that compose that encoding.<br>
     * The "simple" encoding is recommended for TrueType fonts
     * as the "full" encoding risks not matching the character with the right glyph
     * if not done with care.<br>
     * The "full" encoding is specially aimed at Type1 fonts where the glyphs have to be
     * described by non standard names like the Tex math fonts. Each group of three elements
     * compose a code position: the one byte code order in decimal or as 'x' (x cannot be the space), the name and the Unicode character
     * used to access the glyph. The space must be assigned to character position 32 otherwise
     * text justification will not work.
     * <P>
     * Example for a "simple" encoding that includes the Unicode
     * character space, A, B and ecyrillic:
     * <PRE>
     * "# simple 32 0020 0041 0042 0454"
     * </PRE>
     * <P>
     * Example for a "full" encoding for a Type1 Tex font:
     * <PRE>
     * "# full 'A' nottriangeqlleft 0041 'B' dividemultiply 0042 32 space 0020"
     * </PRE>
     * @param name the name of the font or its location on file
     * @param encoding the encoding to be applied to this font
     * @param embedded true if the font is to be embedded in the PDF
     * @param cached true if the font comes from the cache or is added to
     * the cache if new, false if the font is always created new
     * @param ttfAfm the true type font or the afm in a byte array
     * @param pfb the pfb in a byte array
     * @param noThrow if true will not throw an exception if the font is not recognized and will return null, if false will throw
     * an exception if the font is not recognized. Note that even if true an exception may be thrown in some circumstances.
     * This parameter is useful for FontFactory that may have to check many invalid font names before finding the right one
     * @return returns a new font. This font may come from the cache but only if cached
     * is true, otherwise it will always be created new
     * @throws DocumentException the font is invalid
     * @throws IOException the font file could not be read
     * @since   2.0.3
     */
    public static BaseFont createFont(String name, String encoding, boolean embedded, boolean cached, byte ttfAfm[], byte pfb[], boolean noThrow) throws DocumentException, IOException {
        return createFont(name, encoding, embedded, cached, ttfAfm, pfb, noThrow, false);
    }

    /** Creates a new font. This font can be one of the 14 built in types,
     * a Type1 font referred to by an AFM or PFM file, a TrueType font (simple or collection) or a CJK font from the
     * Adobe Asian Font Pack. TrueType fonts and CJK fonts can have an optional style modifier
     * appended to the name. These modifiers are: Bold, Italic and BoldItalic. An
     * example would be "STSong-Light,Bold". Note that this modifiers do not work if
     * the font is embedded. Fonts in TrueType collections are addressed by index such as "msgothic.ttc,1".
     * This would get the second font (indexes start at 0), in this case "MS PGothic".
     * <P>
     * The fonts may or may not be cached depending on the flag <CODE>cached</CODE>.
     * If the <CODE>byte</CODE> arrays are present the font will be
     * read from them instead of the name. A name is still required to identify
     * the font type.
     * <P>
     * Besides the common encodings described by name, custom encodings
     * can also be made. These encodings will only work for the single byte fonts
     * Type1 and TrueType. The encoding string starts with a '#'
     * followed by "simple" or "full". If "simple" there is a decimal for the first character position and then a list
     * of hex values representing the Unicode codes that compose that encoding.<br>
     * The "simple" encoding is recommended for TrueType fonts
     * as the "full" encoding risks not matching the character with the right glyph
     * if not done with care.<br>
     * The "full" encoding is specially aimed at Type1 fonts where the glyphs have to be
     * described by non standard names like the Tex math fonts. Each group of three elements
     * compose a code position: the one byte code order in decimal or as 'x' (x cannot be the space), the name and the Unicode character
     * used to access the glyph. The space must be assigned to character position 32 otherwise
     * text justification will not work.
     * <P>
     * Example for a "simple" encoding that includes the Unicode
     * character space, A, B and ecyrillic:
     * <PRE>
     * "# simple 32 0020 0041 0042 0454"
     * </PRE>
     * <P>
     * Example for a "full" encoding for a Type1 Tex font:
     * <PRE>
     * "# full 'A' nottriangeqlleft 0041 'B' dividemultiply 0042 32 space 0020"
     * </PRE>
     * @param name the name of the font or its location on file
     * @param encoding the encoding to be applied to this font
     * @param embedded true if the font is to be embedded in the PDF
     * @param cached true if the font comes from the cache or is added to
     * the cache if new, false if the font is always created new
     * @param ttfAfm the true type font or the afm in a byte array
     * @param pfb the pfb in a byte array
     * @param noThrow if true will not throw an exception if the font is not recognized and will return null, if false will throw
     * an exception if the font is not recognized. Note that even if true an exception may be thrown in some circumstances.
     * This parameter is useful for FontFactory that may have to check many invalid font names before finding the right one
     * @param   forceRead       in some cases (TrueTypeFont, Type1Font), the full font file will be read and kept in memory if forceRead is true
     * @return returns a new font. This font may come from the cache but only if cached
     * is true, otherwise it will always be created new
     * @throws DocumentException the font is invalid
     * @throws IOException the font file could not be read
     * @since   2.1.5
     */
    public static BaseFont createFont(String name, String encoding, boolean embedded, boolean cached, byte ttfAfm[], byte pfb[], boolean noThrow, boolean forceRead) throws DocumentException, IOException {
        String nameBase = getBaseName(name);
        encoding = normalizeEncoding(encoding);
        boolean isBuiltinFonts14 = BuiltinFonts14.containsKey(name);
        boolean isCJKFont = isBuiltinFonts14 ? false : CJKFont.isCJKFont(nameBase, encoding);
        if (isBuiltinFonts14 || isCJKFont)
            embedded = false;
        else if (encoding.equals(IDENTITY_H) || encoding.equals(IDENTITY_V))
            embedded = true;
        BaseFont fontFound = null;
        BaseFont fontBuilt = null;
        String key = name + "\n" + encoding + "\n" + embedded;
        if (cached) {
            synchronized (fontCache) {
                fontFound = fontCache.get(key);
            }
            if (fontFound != null)
                return fontFound;
        }
        if (isBuiltinFonts14 || name.toLowerCase().endsWith(".afm") || name.toLowerCase().endsWith(".pfm")) {
            fontBuilt = new Type1Font(name, encoding, embedded, ttfAfm, pfb, forceRead);
            fontBuilt.fastWinansi = encoding.equals(CP1252);
        }
        else if (nameBase.toLowerCase().endsWith(".ttf") || nameBase.toLowerCase().endsWith(".otf") || nameBase.toLowerCase().indexOf(".ttc,") > 0) {
            if (encoding.equals(IDENTITY_H) || encoding.equals(IDENTITY_V))
                fontBuilt = new TrueTypeFontUnicode(name, encoding, embedded, ttfAfm, forceRead);
            else {
                fontBuilt = new TrueTypeFont(name, encoding, embedded, ttfAfm, false, forceRead);
                fontBuilt.fastWinansi = encoding.equals(CP1252);
            }
        }
        else if (isCJKFont)
            fontBuilt = new CJKFont(name, encoding, embedded);
        else if (noThrow)
            return null;
        else
            throw new DocumentException(MessageLocalization.getComposedMessage("font.1.with.2.is.not.recognized", name, encoding));
        if (cached) {
            synchronized (fontCache) {
                fontFound = fontCache.get(key);
                if (fontFound != null)
                    return fontFound;
                fontCache.put(key, fontBuilt);
            }
        }
        return fontBuilt;
    }

    /**
     * Creates a font based on an existing document font. The created font font may not
     * behave as expected, depending on the encoding or subset.
     * @param fontRef the reference to the document font
     * @return the font
     */
    public static BaseFont createFont(PRIndirectReference fontRef) {
        return new DocumentFont(fontRef);
    }

    /**
     * Gets the name without the modifiers Bold, Italic or BoldItalic.
     * @param name the full name of the font
     * @return the name without the modifiers Bold, Italic or BoldItalic
     */
    protected static String getBaseName(String name) {
        if (name.endsWith(",Bold"))
            return name.substring(0, name.length() - 5);
        else if (name.endsWith(",Italic"))
            return name.substring(0, name.length() - 7);
        else if (name.endsWith(",BoldItalic"))
            return name.substring(0, name.length() - 11);
        else
            return name;
    }

    /**
     * Normalize the encoding names. "winansi" is changed to "Cp1252" and
     * "macroman" is changed to "MacRoman".
     * @param enc the encoding to be normalized
     * @return the normalized encoding
     */
    protected static String normalizeEncoding(String enc) {
        if (enc.equals("winansi") || enc.equals(""))
            return CP1252;
        else if (enc.equals("macroman"))
            return MACROMAN;
        else
            return enc;
    }

    /**
     * Creates the <CODE>widths</CODE> and the <CODE>differences</CODE> arrays
     */
    protected void createEncoding() {
        if (encoding.startsWith("#")) {
            specialMap = new IntHashtable();
            StringTokenizer tok = new StringTokenizer(encoding.substring(1), " ,\t\n\r\f");
            if (tok.nextToken().equals("full")) {
                while (tok.hasMoreTokens()) {
                    String order = tok.nextToken();
                    String name = tok.nextToken();
                    char uni = (char)Integer.parseInt(tok.nextToken(), 16);
                    int orderK;
                    if (order.startsWith("'"))
                        orderK = order.charAt(1);
                    else
                        orderK = Integer.parseInt(order);
                    orderK %= 256;
                    specialMap.put(uni, orderK);
                    differences[orderK] = name;
                    unicodeDifferences[orderK] = uni;
                    widths[orderK] = getRawWidth(uni, name);
                    charBBoxes[orderK] = getRawCharBBox(uni, name);
                }
            }
            else {
                int k = 0;
                if (tok.hasMoreTokens())
                    k = Integer.parseInt(tok.nextToken());
                while (tok.hasMoreTokens() && k < 256) {
                    String hex = tok.nextToken();
                    int uni = Integer.parseInt(hex, 16) % 0x10000;
                    String name = GlyphList.unicodeToName(uni);
                    if (name != null) {
                        specialMap.put(uni, k);
                        differences[k] = name;
                        unicodeDifferences[k] = (char)uni;
                        widths[k] = getRawWidth(uni, name);
                        charBBoxes[k] = getRawCharBBox(uni, name);
                        ++k;
                    }
                }
            }
            for (int k = 0; k < 256; ++k) {
                if (differences[k] == null) {
                    differences[k] = notdef;
                }
            }
        }
        else if (fontSpecific) {
            for (int k = 0; k < 256; ++k) {
                widths[k] = getRawWidth(k, null);
                charBBoxes[k] = getRawCharBBox(k, null);
            }
        }
        else {
            String s;
            String name;
            char c;
            byte b[] = new byte[1];
            for (int k = 0; k < 256; ++k) {
                b[0] = (byte)k;
                s = PdfEncodings.convertToString(b, encoding);
                if (s.length() > 0) {
                    c = s.charAt(0);
                }
                else {
                    c = '?';
                }
                name = GlyphList.unicodeToName(c);
                if (name == null)
                    name = notdef;
                differences[k] = name;
                unicodeDifferences[k] = c;
                widths[k] = getRawWidth(c, name);
                charBBoxes[k] = getRawCharBBox(c, name);
            }
        }
    }

    /**
     * Gets the width from the font according to the Unicode char <CODE>c</CODE>
     * or the <CODE>name</CODE>. If the <CODE>name</CODE> is null it's a symbolic font.
     * @param c the unicode char
     * @param name the glyph name
     * @return the width of the char
     */
    abstract int getRawWidth(int c, String name);

    /**
     * Gets the kerning between two Unicode chars.
     * @param char1 the first char
     * @param char2 the second char
     * @return the kerning to be applied in normalized 1000 units
     */
    public abstract int getKerning(int char1, int char2);

    /**
     * Sets the kerning between two Unicode chars.
     * @param char1 the first char
     * @param char2 the second char
     * @param kern the kerning to apply in normalized 1000 units
     * @return <code>true</code> if the kerning was applied, <code>false</code> otherwise
     */
    public abstract boolean setKerning(int char1, int char2, int kern);

    /**
     * Gets the width of a <CODE>char</CODE> in normalized 1000 units.
     * @param char1 the unicode <CODE>char</CODE> to get the width of
     * @return the width in normalized 1000 units
     */
    public int getWidth(int char1) {
        if (fastWinansi) {
            if (char1 < 128 || char1 >= 160 && char1 <= 255)
                return widths[char1];
            else
                return widths[PdfEncodings.winansi.get(char1)];
        }
        else {
            int total = 0;
            byte mbytes[] = convertToBytes((char)char1);
            for (int k = 0; k < mbytes.length; ++k)
                total += widths[0xff & mbytes[k]];
            return total;
        }
    }

    /**
     * Gets the width of a <CODE>String</CODE> in normalized 1000 units.
     * @param text the <CODE>String</CODE> to get the width of
     * @return the width in normalized 1000 units
     */
    public int getWidth(String text) {
        int total = 0;
        if (fastWinansi) {
            int len = text.length();
            for (int k = 0; k < len; ++k) {
                char char1 = text.charAt(k);
                if (char1 < 128 || char1 >= 160 && char1 <= 255)
                    total += widths[char1];
                else
                    total += widths[PdfEncodings.winansi.get(char1)];
            }
            return total;
        }
        else {
            byte mbytes[] = convertToBytes(text);
            for (int k = 0; k < mbytes.length; ++k)
                total += widths[0xff & mbytes[k]];
        }
        return total;
    }

/**
 * Gets the descent of a <CODE>String</CODE> in normalized 1000 units. The descent will always be
 * less than or equal to zero even if all the characters have an higher descent.
 * @param text the <CODE>String</CODE> to get the descent of
 * @return the descent in normalized 1000 units
 */
    public int getDescent(String text) {
        int min = 0;
        char chars[] = text.toCharArray();
        for (int k = 0; k < chars.length; ++k) {
            int bbox[] = getCharBBox(chars[k]);
            if (bbox != null && bbox[1] < min)
                min = bbox[1];
        }
        return min;
    }

/**
 * Gets the ascent of a <CODE>String</CODE> in normalized 1000 units. The ascent will always be
 * greater than or equal to zero even if all the characters have a lower ascent.
 * @param text the <CODE>String</CODE> to get the ascent of
 * @return the ascent in normalized 1000 units
 */
    public int getAscent(String text) {
        int max = 0;
        char chars[] = text.toCharArray();
        for (int k = 0; k < chars.length; ++k) {
            int bbox[] = getCharBBox(chars[k]);
            if (bbox != null && bbox[3] > max)
                max = bbox[3];
        }
        return max;
    }

/**
 * Gets the descent of a <CODE>String</CODE> in points. The descent will always be
 * less than or equal to zero even if all the characters have an higher descent.
 * @param text the <CODE>String</CODE> to get the descent of
 * @param fontSize the size of the font
 * @return the descent in points
 */
    public float getDescentPoint(String text, float fontSize)
    {
        return getDescent(text) * 0.001f * fontSize;
    }

/**
 * Gets the ascent of a <CODE>String</CODE> in points. The ascent will always be
 * greater than or equal to zero even if all the characters have a lower ascent.
 * @param text the <CODE>String</CODE> to get the ascent of
 * @param fontSize the size of the font
 * @return the ascent in points
 */
    public float getAscentPoint(String text, float fontSize)
    {
        return getAscent(text) * 0.001f * fontSize;
    }
// ia>

    /**
     * Gets the width of a <CODE>String</CODE> in points taking kerning
     * into account.
     * @param text the <CODE>String</CODE> to get the width of
     * @param fontSize the font size
     * @return the width in points
     */
    public float getWidthPointKerned(String text, float fontSize) {
        float size = getWidth(text) * 0.001f * fontSize;
        if (!hasKernPairs())
            return size;
        int len = text.length() - 1;
        int kern = 0;
        char c[] = text.toCharArray();
        for (int k = 0; k < len; ++k) {
            kern += getKerning(c[k], c[k + 1]);
        }
        return size + kern * 0.001f * fontSize;
    }

    /**
     * Gets the width of a <CODE>String</CODE> in points.
     * @param text the <CODE>String</CODE> to get the width of
     * @param fontSize the font size
     * @return the width in points
     */
    public float getWidthPoint(String text, float fontSize) {
        return getWidth(text) * 0.001f * fontSize;
    }

    /**
     * Gets the width of a <CODE>char</CODE> in points.
     * @param char1 the <CODE>char</CODE> to get the width of
     * @param fontSize the font size
     * @return the width in points
     */
    public float getWidthPoint(int char1, float fontSize) {
        return getWidth(char1) * 0.001f * fontSize;
    }

    /**
     * Converts a <CODE>String</CODE> to a </CODE>byte</CODE> array according
     * to the font's encoding.
     * @param text the <CODE>String</CODE> to be converted
     * @return an array of <CODE>byte</CODE> representing the conversion according to the font's encoding
     */
    byte[] convertToBytes(String text) {
        if (directTextToByte)
            return PdfEncodings.convertToBytes(text, null);
        if (specialMap != null) {
            byte[] b = new byte[text.length()];
            int ptr = 0;
            int length = text.length();
            for (int k = 0; k < length; ++k) {
                char c = text.charAt(k);
                if (specialMap.containsKey(c))
                    b[ptr++] = (byte)specialMap.get(c);
            }
            if (ptr < length) {
                byte[] b2 = new byte[ptr];
                System.arraycopy(b, 0, b2, 0, ptr);
                return b2;
            }
            else
                return b;
        }
        return PdfEncodings.convertToBytes(text, encoding);
    }

    /**
     * Converts a <CODE>char</CODE> to a </CODE>byte</CODE> array according
     * to the font's encoding.
     * @param char1 the <CODE>char</CODE> to be converted
     * @return an array of <CODE>byte</CODE> representing the conversion according to the font's encoding
     */
    byte[] convertToBytes(int char1) {
        if (directTextToByte)
            return PdfEncodings.convertToBytes((char)char1, null);
        if (specialMap != null) {
            if (specialMap.containsKey(char1))
                return new byte[]{(byte)specialMap.get(char1)};
            else
                return new byte[0];
        }
        return PdfEncodings.convertToBytes((char)char1, encoding);
    }

    /** Outputs to the writer the font dictionaries and streams.
     * @param writer the writer for this document
     * @param ref the font indirect reference
     * @param params several parameters that depend on the font type
     * @throws IOException on error
     * @throws DocumentException error in generating the object
     */
    abstract void writeFont(PdfWriter writer, PdfIndirectReference ref, Object params[]) throws DocumentException, IOException;

    /**
     * Returns a PdfStream object with the full font program (if possible).
     * This method will return null for some types of fonts (CJKFont, Type3Font)
     * or if there is no font program available (standard Type 1 fonts).
     * @return  a PdfStream with the font program
     * @since   2.1.3
     */
    abstract PdfStream getFullFontStream() throws IOException, DocumentException;

    /** Gets the encoding used to convert <CODE>String</CODE> into <CODE>byte[]</CODE>.
     * @return the encoding name
     */
    public String getEncoding() {
        return encoding;
    }

    /** Gets the font parameter identified by <CODE>key</CODE>. Valid values
     * for <CODE>key</CODE> are <CODE>ASCENT</CODE>, <CODE>AWT_ASCENT</CODE>, <CODE>CAPHEIGHT</CODE>,
     * <CODE>DESCENT</CODE>, <CODE>AWT_DESCENT</CODE>,
     * <CODE>ITALICANGLE</CODE>, <CODE>BBOXLLX</CODE>, <CODE>BBOXLLY</CODE>, <CODE>BBOXURX</CODE>
     * and <CODE>BBOXURY</CODE>.
     * @param key the parameter to be extracted
     * @param fontSize the font size in points
     * @return the parameter in points
     */
    public abstract float getFontDescriptor(int key, float fontSize);

    /** Gets the font type. The font types can be: FONT_TYPE_T1,
     * FONT_TYPE_TT, FONT_TYPE_CJK and FONT_TYPE_TTUNI.
     * @return the font type
     */
    public int getFontType() {
        return fontType;
    }

    /** Gets the embedded flag.
     * @return <CODE>true</CODE> if the font is embedded.
     */
    public boolean isEmbedded() {
        return embedded;
    }

    /** Gets the symbolic flag of the font.
     * @return <CODE>true</CODE> if the font is symbolic
     */
    public boolean isFontSpecific() {
        return fontSpecific;
    }

    /** Creates a unique subset prefix to be added to the font name when the font is embedded and subset.
     * @return the subset prefix
     */
    public static String createSubsetPrefix() {
        StringBuilder s = new StringBuilder("");
        for (int k = 0; k < 6; ++k)
            s.append((char)(Math.random() * 26 + 'A'));
        return s + "+";
    }

    /** Gets the Unicode character corresponding to the byte output to the pdf stream.
     * @param index the byte index
     * @return the Unicode character
     */
    char getUnicodeDifferences(int index) {
        return unicodeDifferences[index];
    }

    /** Gets the postscript font name.
     * @return the postscript font name
     */
    public abstract String getPostscriptFontName();

    /**
     * Sets the font name that will appear in the pdf font dictionary.
     * Use with care as it can easily make a font unreadable if not embedded.
     * @param name the new font name
     */
    public abstract void setPostscriptFontName(String name);

    /** Gets the full name of the font. If it is a True Type font
     * each array element will have {Platform ID, Platform Encoding ID,
     * Language ID, font name}. The interpretation of this values can be
     * found in the Open Type specification, chapter 2, in the 'name' table.<br>
     * For the other fonts the array has a single element with {"", "", "",
     * font name}.
     * @return the full name of the font
     */
    public abstract String[][] getFullFontName();

    /** Gets all the entries of the names-table. If it is a True Type font
     * each array element will have {Name ID, Platform ID, Platform Encoding ID,
     * Language ID, font name}. The interpretation of this values can be
     * found in the Open Type specification, chapter 2, in the 'name' table.<br>
     * For the other fonts the array has a single element with {"4", "", "", "",
     * font name}.
     * @return the full name of the font
     * @since 2.0.8
     */
    public abstract String[][] getAllNameEntries();

    /** Gets the full name of the font. If it is a True Type font
     * each array element will have {Platform ID, Platform Encoding ID,
     * Language ID, font name}. The interpretation of this values can be
     * found in the Open Type specification, chapter 2, in the 'name' table.<br>
     * For the other fonts the array has a single element with {"", "", "",
     * font name}.
     * @param name the name of the font
     * @param encoding the encoding of the font
     * @param ttfAfm the true type font or the afm in a byte array
     * @throws DocumentException on error
     * @throws IOException on error
     * @return the full name of the font
     */
    public static String[][] getFullFontName(String name, String encoding, byte ttfAfm[]) throws DocumentException, IOException {
        String nameBase = getBaseName(name);
        BaseFont fontBuilt = null;
        if (nameBase.toLowerCase().endsWith(".ttf") || nameBase.toLowerCase().endsWith(".otf") || nameBase.toLowerCase().indexOf(".ttc,") > 0)
            fontBuilt = new TrueTypeFont(name, CP1252, false, ttfAfm, true, false);
        else
            fontBuilt = createFont(name, encoding, false, false, ttfAfm, null);
        return fontBuilt.getFullFontName();
    }

    /** Gets all the names from the font. Only the required tables are read.
     * @param name the name of the font
     * @param encoding the encoding of the font
     * @param ttfAfm the true type font or the afm in a byte array
     * @throws DocumentException on error
     * @throws IOException on error
     * @return an array of Object[] built with {getPostscriptFontName(), getFamilyFontName(), getFullFontName()}
     */
    public static Object[] getAllFontNames(String name, String encoding, byte ttfAfm[]) throws DocumentException, IOException {
        String nameBase = getBaseName(name);
        BaseFont fontBuilt = null;
        if (nameBase.toLowerCase().endsWith(".ttf") || nameBase.toLowerCase().endsWith(".otf") || nameBase.toLowerCase().indexOf(".ttc,") > 0)
            fontBuilt = new TrueTypeFont(name, CP1252, false, ttfAfm, true, false);
        else
            fontBuilt = createFont(name, encoding, false, false, ttfAfm, null);
        return new Object[]{fontBuilt.getPostscriptFontName(), fontBuilt.getFamilyFontName(), fontBuilt.getFullFontName()};
    }

    /** Gets all the entries of the namestable from the font. Only the required tables are read.
     * @param name the name of the font
     * @param encoding the encoding of the font
     * @param ttfAfm the true type font or the afm in a byte array
     * @throws DocumentException on error
     * @throws IOException on error
     * @return an array of Object[] built with {getPostscriptFontName(), getFamilyFontName(), getFullFontName()}
     * @since 2.0.8
     */
    public static String[][] getAllNameEntries(String name, String encoding, byte ttfAfm[]) throws DocumentException, IOException {
        String nameBase = getBaseName(name);
        BaseFont fontBuilt = null;
        if (nameBase.toLowerCase().endsWith(".ttf") || nameBase.toLowerCase().endsWith(".otf") || nameBase.toLowerCase().indexOf(".ttc,") > 0)
            fontBuilt = new TrueTypeFont(name, CP1252, false, ttfAfm, true, false);
        else
            fontBuilt = createFont(name, encoding, false, false, ttfAfm, null);
        return fontBuilt.getAllNameEntries();
    }

    /** Gets the family name of the font. If it is a True Type font
     * each array element will have {Platform ID, Platform Encoding ID,
     * Language ID, font name}. The interpretation of this values can be
     * found in the Open Type specification, chapter 2, in the 'name' table.<br>
     * For the other fonts the array has a single element with {"", "", "",
     * font name}.
     * @return the family name of the font
     */
    public abstract String[][] getFamilyFontName();

    /** Gets the code pages supported by the font. This has only meaning
     * with True Type fonts.
     * @return the code pages supported by the font
     */
    public String[] getCodePagesSupported() {
        return new String[0];
    }

    /** Enumerates the postscript font names present inside a
     * True Type Collection.
     * @param ttcFile the file name of the font
     * @throws DocumentException on error
     * @throws IOException on error
     * @return the postscript font names
     */
    public static String[] enumerateTTCNames(String ttcFile) throws DocumentException, IOException {
        return new EnumerateTTC(ttcFile).getNames();
    }

    /** Enumerates the postscript font names present inside a
     * True Type Collection.
     * @param ttcArray the font as a <CODE>byte</CODE> array
     * @throws DocumentException on error
     * @throws IOException on error
     * @return the postscript font names
     */
    public static String[] enumerateTTCNames(byte ttcArray[]) throws DocumentException, IOException {
        return new EnumerateTTC(ttcArray).getNames();
    }

    /** Gets the font width array.
     * @return the font width array
     */
    public int[] getWidths() {
        return widths;
    }

    /** Gets the array with the names of the characters.
     * @return the array with the names of the characters
     */
    public String[] getDifferences() {
        return differences;
    }

    /** Gets the array with the unicode characters.
     * @return the array with the unicode characters
     */
    public char[] getUnicodeDifferences() {
        return unicodeDifferences;
    }

    /** Gets the state of the property.
     * @return value of property forceWidthsOutput
     */
    public boolean isForceWidthsOutput() {
        return forceWidthsOutput;
    }

    /** Set to <CODE>true</CODE> to force the generation of the
     * widths array.
     * @param forceWidthsOutput <CODE>true</CODE> to force the generation of the
     * widths array
     */
    public void setForceWidthsOutput(boolean forceWidthsOutput) {
        this.forceWidthsOutput = forceWidthsOutput;
    }

    /** Gets the direct conversion of <CODE>char</CODE> to <CODE>byte</CODE>.
     * @return value of property directTextToByte.
     * @see #setDirectTextToByte(boolean directTextToByte)
     */
    public boolean isDirectTextToByte() {
        return directTextToByte;
    }

    /** Sets the conversion of <CODE>char</CODE> directly to <CODE>byte</CODE>
     * by casting. This is a low level feature to put the bytes directly in
     * the content stream without passing through String.getBytes().
     * @param directTextToByte New value of property directTextToByte.
     */
    public void setDirectTextToByte(boolean directTextToByte) {
        this.directTextToByte = directTextToByte;
    }

    /** Indicates if all the glyphs and widths for that particular
     * encoding should be included in the document.
     * @return <CODE>false</CODE> to include all the glyphs and widths.
     */
    public boolean isSubset() {
        return subset;
    }

    /** Indicates if all the glyphs and widths for that particular
     * encoding should be included in the document. When set to <CODE>true</CODE>
     * only the glyphs used will be included in the font. When set to <CODE>false</CODE>
     * and {@link #addSubsetRange(int[])} was not called the full font will be included
     * otherwise just the characters ranges will be included.
     * @param subset new value of property subset
     */
    public void setSubset(boolean subset) {
        this.subset = subset;
    }

    /** Gets the font resources.
     * @param key the full name of the resource
     * @return the <CODE>InputStream</CODE> to get the resource or
     * <CODE>null</CODE> if not found
     */
    public static InputStream getResourceStream(String key) {
        return getResourceStream(key, null);
    }

    /** Gets the font resources.
     * @param key the full name of the resource
     * @param loader the ClassLoader to load the resource or null to try the ones available
     * @return the <CODE>InputStream</CODE> to get the resource or
     * <CODE>null</CODE> if not found
     */
    public static InputStream getResourceStream(String key, ClassLoader loader) {
        if (key.startsWith("/"))
            key = key.substring(1);
        InputStream is = null;
        if (loader != null) {
            is = loader.getResourceAsStream(key);
            if (is != null)
                return is;
        }
        // Try to use Context Class Loader to load the properties file.
        try {
            ClassLoader contextClassLoader = Thread.currentThread().getContextClassLoader();
            if (contextClassLoader != null) {
                is = contextClassLoader.getResourceAsStream(key);
            }
        } catch (Throwable e) {}

        if (is == null) {
            is = BaseFont.class.getResourceAsStream("/" + key);
        }
        if (is == null) {
            is = ClassLoader.getSystemResourceAsStream(key);
        }
        return is;
    }

    /** Gets the Unicode equivalent to a CID.
     * The (inexistent) CID <FF00> is translated as '\n'.
     * It has only meaning with CJK fonts with Identity encoding.
     * @param c the CID code
     * @return the Unicode equivalent
     */
    public int getUnicodeEquivalent(int c) {
        return c;
    }

    /** Gets the CID code given an Unicode.
     * It has only meaning with CJK fonts.
     * @param c the Unicode
     * @return the CID equivalent
     */
    public int getCidCode(int c) {
        return c;
    }

    /** Checks if the font has any kerning pairs.
     * @return <CODE>true</CODE> if the font has any kerning pairs
     */
    public abstract boolean hasKernPairs();

    /**
     * Checks if a character exists in this font.
     * @param c the character to check
     * @return <CODE>true</CODE> if the character has a glyph,
     * <CODE>false</CODE> otherwise
     */
    public boolean charExists(int c) {
        byte b[] = convertToBytes(c);
        return b.length > 0;
    }

    /**
     * Sets the character advance.
     * @param c the character
     * @param advance the character advance normalized to 1000 units
     * @return <CODE>true</CODE> if the advance was set,
     * <CODE>false</CODE> otherwise
     */
    public boolean setCharAdvance(int c, int advance) {
        byte b[] = convertToBytes(c);
        if (b.length == 0)
            return false;
        widths[0xff & b[0]] = advance;
        return true;
    }

    private static void addFont(PRIndirectReference fontRef, IntHashtable hits, ArrayList<Object[]> fonts) {
        PdfObject obj = PdfReader.getPdfObject(fontRef);
        if (obj == null || !obj.isDictionary())
            return;
        PdfDictionary font = (PdfDictionary)obj;
        PdfName subtype = font.getAsName(PdfName.SUBTYPE);
        if (!PdfName.TYPE1.equals(subtype) && !PdfName.TRUETYPE.equals(subtype) && !PdfName.TYPE0.equals(subtype))
            return;
        PdfName name = font.getAsName(PdfName.BASEFONT);
        fonts.add(new Object[]{PdfName.decodeName(name.toString()), fontRef});
        hits.put(fontRef.getNumber(), 1);
    }

    private static void recourseFonts(PdfDictionary page, IntHashtable hits, ArrayList<Object[]> fonts, int level) {
        ++level;
        if (level > 50) // in case we have an endless loop
            return;
        if (page == null)
            return;
        PdfDictionary resources = page.getAsDict(PdfName.RESOURCES);
        if (resources == null)
            return;
        PdfDictionary font = resources.getAsDict(PdfName.FONT);
        if (font != null) {
            for (PdfName key : font.getKeys()) {
                PdfObject ft = font.get(key);
                if (ft == null || !ft.isIndirect())
                    continue;
                int hit = ((PRIndirectReference)ft).getNumber();
                if (hits.containsKey(hit))
                    continue;
                addFont((PRIndirectReference)ft, hits, fonts);
            }
        }
        PdfDictionary xobj = resources.getAsDict(PdfName.XOBJECT);
        if (xobj != null) {
            for (PdfName key : xobj.getKeys()) {
                PdfObject po = xobj.getDirectObject(key);
                if (po instanceof PdfDictionary)
                    recourseFonts((PdfDictionary)po, hits, fonts, level);
            }
        }
    }

    /**
     * Gets a list of all document fonts. Each element of the <CODE>ArrayList</CODE>
     * contains a <CODE>Object[]{String,PRIndirectReference}</CODE> with the font name
     * and the indirect reference to it.
     * @param reader the document where the fonts are to be listed from
     * @return the list of fonts and references
     */
    public static ArrayList<Object[]> getDocumentFonts(PdfReader reader) {
        IntHashtable hits = new IntHashtable();
        ArrayList<Object[]> fonts = new ArrayList<Object[]>();
        int npages = reader.getNumberOfPages();
        for (int k = 1; k <= npages; ++k)
            recourseFonts(reader.getPageN(k), hits, fonts, 1);
        return fonts;
    }

    /**
     * Gets a list of the document fonts in a particular page. Each element of the <CODE>ArrayList</CODE>
     * contains a <CODE>Object[]{String,PRIndirectReference}</CODE> with the font name
     * and the indirect reference to it.
     * @param reader the document where the fonts are to be listed from
     * @param page the page to list the fonts from
     * @return the list of fonts and references
     */
    public static ArrayList<Object[]> getDocumentFonts(PdfReader reader, int page) {
        IntHashtable hits = new IntHashtable();
        ArrayList<Object[]> fonts = new ArrayList<Object[]>();
        recourseFonts(reader.getPageN(page), hits, fonts, 1);
        return fonts;
    }

    /**
     * Gets the smallest box enclosing the character contours. It will return
     * <CODE>null</CODE> if the font has not the information or the character has no
     * contours, as in the case of the space, for example. Characters with no contours may
     * also return [0,0,0,0].
     * @param c the character to get the contour bounding box from
     * @return an array of four floats with the bounding box in the format [llx,lly,urx,ury] or
     * <code>null</code>
     */
    public int[] getCharBBox(int c) {
        byte b[] = convertToBytes(c);
        if (b.length == 0)
            return null;
        else
            return charBBoxes[b[0] & 0xff];
    }

    protected abstract int[] getRawCharBBox(int c, String name);

    /**
     * iText expects Arabic Diactrics (tashkeel) to have zero advance but some fonts,
     * most notably those that come with Windows, like times.ttf, have non-zero
     * advance for those characters. This method makes those character to have zero
     * width advance and work correctly in the iText Arabic shaping and reordering
     * context.
     */
    public void correctArabicAdvance() {
        for (char c = '\u064b'; c <= '\u0658'; ++c)
            setCharAdvance(c, 0);
        setCharAdvance('\u0670', 0);
        for (char c = '\u06d6'; c <= '\u06dc'; ++c)
            setCharAdvance(c, 0);
        for (char c = '\u06df'; c <= '\u06e4'; ++c)
            setCharAdvance(c, 0);
        for (char c = '\u06e7'; c <= '\u06e8'; ++c)
            setCharAdvance(c, 0);
        for (char c = '\u06ea'; c <= '\u06ed'; ++c)
            setCharAdvance(c, 0);
    }

    /**
     * Adds a character range when subsetting. The range is an <CODE>int</CODE> array
     * where the first element is the start range inclusive and the second element is the
     * end range inclusive. Several ranges are allowed in the same array.
     * @param range the character range
     */
    public void addSubsetRange(int[] range) {
        if (subsetRanges == null)
            subsetRanges = new ArrayList<int[]>();
        subsetRanges.add(range);
    }

        /**
         * Returns the compression level used for the font streams.
         * @return the compression level (0 = best speed, 9 = best compression, -1 is default)
         * @since 2.1.3
         */
        public int getCompressionLevel() {
                return compressionLevel;
        }

        /**
         * Sets the compression level to be used for the font streams.
         * @param compressionLevel a value between 0 (best speed) and 9 (best compression)
         * @since 2.1.3
         */
        public void setCompressionLevel(int compressionLevel) {
                if (compressionLevel < PdfStream.NO_COMPRESSION || compressionLevel > PdfStream.BEST_COMPRESSION)
                        this.compressionLevel = PdfStream.DEFAULT_COMPRESSION;
                else
                        this.compressionLevel = compressionLevel;
        }
}