Fix CID / CFF font glyphs and add tests

src/UglyToad.PdfPig.Fonts/CompactFontFormat/CharStrings/Type2CharStrings.cs

@@ -11,7 +11,7 @@
     /// as the local (per font) and global (per font set) subroutines.
     /// The CharStrings are lazily evaluated.
     /// </summary>
-    internal class Type2CharStrings
+    internal sealed class Type2CharStrings
     {
         private readonly object locker = new object();
         private readonly Dictionary<string, Type2Glyph> glyphs = new Dictionary<string, Type2Glyph>();
@@ -45,7 +45,7 @@ public Type2Glyph Generate(string name, double defaultWidthX, double nominalWidt
 
                 if (!CharStrings.TryGetValue(name, out var sequence))
                 {
-                    if (!CharStrings.TryGetValue(".notdef", out sequence))
+                    if (!CharStrings.TryGetValue(GlyphList.NotDefined, out sequence))
                     {
                         throw new InvalidOperationException($"No charstring sequence with the name /{name} in this font.");
                     }
@@ -149,15 +149,14 @@ private static void SetWidthFromArgumentsIfPresent(Type2BuildCharContext context
             }
         }
 
-        public class CommandSequence
+        public sealed class CommandSequence
         {
             public IReadOnlyList<float> Values { get; }
             public IReadOnlyList<CommandIdentifier> CommandIdentifiers { get; }
 
             /// <summary>
             /// The ordered list of numbers and commands for a Type 2 charstring or subroutine.
             /// </summary>
-
             public CommandSequence(IReadOnlyList<float> values, IReadOnlyList<CommandIdentifier> commandIdentifiers)
             {
                 Values = values;
@@ -218,7 +217,7 @@ public CommandIdentifier(int commandIndex, bool isMultiByteCommand, byte command
     /// Since Type 2 CharStrings may define their width as the first argument (as a delta from the font's nominal width X)
     /// we can retrieve both details for the Type 2 glyph.
     /// </summary>
-    internal class Type2Glyph
+    internal sealed class Type2Glyph
     {
         /// <summary>
         /// The path of the glyph.

src/UglyToad.PdfPig.Fonts/CompactFontFormat/Charsets/CompactFontFormatCharset.cs

@@ -37,7 +37,7 @@ public virtual string GetNameByGlyphId(int glyphId)
 
         public virtual string GetNameByStringId(int stringId)
         {
-            return GlyphIdToStringIdAndName.Single(x => x.Value.stringId == stringId).Value.name;
+            return GlyphIdToStringIdAndName.SingleOrDefault(x => x.Value.stringId == stringId).Value.name;
         }
 
         public virtual int GetStringIdByGlyphId(int glyphId)

src/UglyToad.PdfPig.Fonts/CompactFontFormat/CompactFontFormatBaseEncoding.cs

@@ -16,7 +16,7 @@ public override string GetName(int code)
         {
             if (!codeToNameMap.TryGetValue(code, out var name))
             {
-                return ".notdef";
+                return NotDefined;
             }
 
             return name;

src/UglyToad.PdfPig.Fonts/CompactFontFormat/CompactFontFormatFont.cs

@@ -27,7 +27,7 @@ public class CompactFontFormatFont
         /// <summary>
         /// The font matrix for this font.
         /// </summary>
-        public TransformationMatrix FontMatrix => TopDictionary.FontMatrix;
+        public TransformationMatrix FontMatrix => TopDictionary.FontMatrix.HasValue ? TopDictionary.FontMatrix.Value : TransformationMatrix.FromValues(0.001, 0, 0, 0.001, 0, 0);
 
         /// <summary>
         /// The value of Weight from the top dictionary or <see langword="null"/>.
@@ -50,6 +50,32 @@ internal CompactFontFormatFont(CompactFontFormatTopLevelDictionary topDictionary
             Encoding = fontEncoding;
         }
 
+        /// <summary>
+        /// Get the character name. Returns <c>null</c> if cannot be processed.
+        /// </summary>
+        public string GetCharacterName(int characterCode, bool isCid)
+        {
+            if (Encoding != null)
+            {
+                return Encoding.GetName(characterCode);
+            }
+
+            if (Charset.IsCidCharset || isCid)
+            {
+                return Charset?.GetNameByStringId(characterCode);
+            }
+
+            string characterName = GlyphList.AdobeGlyphList.UnicodeCodePointToName(characterCode);
+
+            if (characterName.Equals(GlyphList.NotDefined, StringComparison.OrdinalIgnoreCase))
+            {
+                // BobLD: Not tested
+                return Charset?.GetNameByStringId(characterCode);
+            }
+
+            return characterName;
+        }
+
         /// <summary>
         /// Get the bounding box for the character with the given name.
         /// </summary>
@@ -143,9 +169,17 @@ protected virtual decimal GetNominalWidthX(string characterName)
         {
             return PrivateDictionary.NominalWidthX;
         }
+
+        /// <summary>
+        /// Get the Font Matrix for the corresponding character name, if available. Return <c>null</c> if not.
+        /// </summary>
+        public virtual TransformationMatrix? GetFontMatrix(string characterName)
+        {
+            return TopDictionary.FontMatrix;
+        }
     }
 
-    internal class CompactFontFormatCidFont : CompactFontFormatFont
+    internal sealed class CompactFontFormatCidFont : CompactFontFormatFont
     {
         public IReadOnlyList<CompactFontFormatTopLevelDictionary> FontDictionaries { get; }
         public IReadOnlyList<CompactFontFormatPrivateDictionary> PrivateDictionaries { get; }
@@ -183,6 +217,26 @@ protected override decimal GetNominalWidthX(string characterName)
             return dictionary.NominalWidthX;
         }
 
+        public override TransformationMatrix? GetFontMatrix(string characterName)
+        {
+            // BobLd: It seems PdfBox just returns TopDictionary.FontMatrix
+            // But see https://bugs.ghostscript.com/show_bug.cgi?id=690724
+            // and https://github.com/veraPDF/veraPDF-library/issues/1010
+            // TODO - We might need to multiply both matrices together
+
+            if (TopDictionary.FontMatrix is not null)
+            {
+                return TopDictionary.FontMatrix;
+            }
+
+            if (!TryGetFontDictionaryForCharacter(characterName, out var dictionary))
+            {
+                return null;
+            }
+
+            return dictionary.FontMatrix;
+        }
+
         private bool TryGetPrivateDictionaryForCharacter(string characterName, out CompactFontFormatPrivateDictionary dictionary)
         {
             dictionary = null;
@@ -199,5 +253,22 @@ private bool TryGetPrivateDictionaryForCharacter(string characterName, out Compa
 
             return true;
         }
+
+        private bool TryGetFontDictionaryForCharacter(string characterName, out CompactFontFormatTopLevelDictionary dictionary)
+        {
+            dictionary = null;
+
+            var glyphId = Charset.GetGlyphIdByName(characterName);
+
+            var fd = FdSelect.GetFontDictionaryIndex(glyphId);
+            if (fd == -1)
+            {
+                return false;
+            }
+
+            dictionary = FontDictionaries[fd];
+
+            return true;
+        }
     }
 }

src/UglyToad.PdfPig.Fonts/CompactFontFormat/CompactFontFormatFontCollection.cs

@@ -8,7 +8,7 @@
     /// A Compact Font Format (CFF) font program as described in The Compact Font Format specification (Adobe Technical Note #5176).
     /// A CFF font may contain multiple fonts and achieves compression by sharing details between fonts in the set.
     /// </summary>
-    public class CompactFontFormatFontCollection
+    public sealed class CompactFontFormatFontCollection
     {
         /// <summary>
         /// The decoded header table for this font.
@@ -65,16 +65,11 @@ public TransformationMatrix GetFirstTransformationMatrix()
         /// <summary>
         /// Get the name for the character with the given character code from the font.
         /// </summary>
-        public string GetCharacterName(int characterCode)
+        public string GetCharacterName(int characterCode, bool isCid)
         {
-            var font = FirstFont;
+            var name = FirstFont.GetCharacterName(characterCode, isCid);
 
-            if (font.Encoding != null)
-            {
-                return font.Encoding.GetName(characterCode);
-            }
-
-            return ".notdef";
+            return name ?? GlyphList.NotDefined;
         }
     }
 }

src/UglyToad.PdfPig.Fonts/CompactFontFormat/CompactFontFormatFormat0Encoding.cs

@@ -2,12 +2,12 @@
 {
     using System.Collections.Generic;
 
-    internal class CompactFontFormatFormat0Encoding : CompactFontFormatBuiltInEncoding
+    internal sealed class CompactFontFormatFormat0Encoding : CompactFontFormatBuiltInEncoding
     {
         public CompactFontFormatFormat0Encoding(IReadOnlyList<(int code, int sid, string str)> values,
             IReadOnlyList<Supplement> supplements) : base(supplements)
         {
-            Add(0, 0, ".notdef");
+            Add(0, 0, NotDefined);
 
             foreach (var value in values)
             {

src/UglyToad.PdfPig.Fonts/CompactFontFormat/CompactFontFormatFormat1Encoding.cs

@@ -2,15 +2,15 @@
 {
     using System.Collections.Generic;
 
-    internal class CompactFontFormatFormat1Encoding : CompactFontFormatBuiltInEncoding
+    internal sealed class CompactFontFormatFormat1Encoding : CompactFontFormatBuiltInEncoding
     {
         public int NumberOfRanges { get; set; }
 
         public CompactFontFormatFormat1Encoding(int numberOfRanges, IReadOnlyList<(int code, int sid, string str)> values, IReadOnlyList<Supplement> supplements) : base(supplements)
         {
             NumberOfRanges = numberOfRanges;
 
-            Add(0, 0, ".notdef");
+            Add(0, 0, NotDefined);
 
             foreach (var value in values)
             {

src/UglyToad.PdfPig.Fonts/CompactFontFormat/Dictionaries/CompactFontFormatTopLevelDictionary.cs

@@ -2,7 +2,7 @@
 {
     using Core;
 
-    internal class CompactFontFormatTopLevelDictionary
+    internal sealed class CompactFontFormatTopLevelDictionary
     {
         public const int UnsetOffset = -1;
 
@@ -30,7 +30,7 @@ internal class CompactFontFormatTopLevelDictionary
 
         public CompactFontFormatCharStringType CharStringType { get; set; } = CompactFontFormatCharStringType.Type2;
 
-        public TransformationMatrix FontMatrix { get; set; } = TransformationMatrix.FromValues(0.001, 0, 0, 0.001, 0, 0);
+        public TransformationMatrix? FontMatrix { get; set; }
 
         public decimal StrokeWidth { get; set; }
 
@@ -79,7 +79,7 @@ public override string ToString()
         }
     }
 
-    internal class CidFontOperators
+    internal sealed class CidFontOperators
     {
         public RegistryOrderingSupplement Ros { get; set; }
 
@@ -100,7 +100,7 @@ internal class CidFontOperators
         public string FontName { get; set; }
     }
 
-    internal class RegistryOrderingSupplement
+    internal sealed class RegistryOrderingSupplement
     {
         public string Registry { get; set; }
 

src/UglyToad.PdfPig.Fonts/Encodings/Encoding.cs

@@ -8,6 +8,11 @@
     /// </summary>
     public abstract class Encoding
     {
+        /// <summary>
+        /// <c>.notdef</c>.
+        /// </summary>
+        protected internal const string NotDefined = ".notdef";
+
         /// <summary>
         /// Mutable code to name map.
         /// </summary>
@@ -24,7 +29,7 @@ public abstract class Encoding
         protected readonly Dictionary<string, int> NameToCode = new Dictionary<string, int>(250);
 
         /// <summary>
-        /// Maps from names to character cocdes.
+        /// Maps from names to character codes.
         /// </summary>
         public IReadOnlyDictionary<string, int> NameToCodeMap => NameToCode;
 
@@ -48,25 +53,24 @@ public bool ContainsCode(int code)
         {
             return CodeToName.ContainsKey(code);
         }
-        
+
         /// <summary>
         /// Get the character name corresponding to the given code.
         /// </summary>
         public virtual string GetName(int code)
         {
             if (!CodeToName.TryGetValue(code, out var name))
             {
-                return ".notdef";
+                return NotDefined;
             }
 
             return name;
         }
 
-
         /// <summary>
-        /// Get the character code from name 
+        /// Get the character code from name
         /// </summary>
-        /// <param name="name">Character name (eg. euro, ampersand, A, space)</param>
+        /// <param name="name">Character name (e.g. euro, ampersand, A, space)</param>
         /// <returns>-1 if not found otherwise the character code</returns>
         public virtual int GetCode(string name)
         {
@@ -90,7 +94,7 @@ protected void Add(int code, string name)
                 NameToCode[name] = code;
             }
         }
-        
+
         /// <summary>
         /// Get a known encoding instance with the given name.
         /// </summary>

src/UglyToad.PdfPig.Fonts/GlyphList.cs

@@ -9,9 +9,12 @@
     /// <summary>
     /// A list which maps PostScript glyph names to unicode values.
     /// </summary>
-    public class GlyphList
+    public sealed class GlyphList
     {
-        private const string NotDefined = ".notdef";
+        /// <summary>
+        /// <c>.notdef</c>.
+        /// </summary>
+        public const string NotDefined = ".notdef";
 
         private readonly IReadOnlyDictionary<string, string> nameToUnicode;
         private readonly IReadOnlyDictionary<string, string> unicodeToName;

src/UglyToad.PdfPig.Fonts/TrueType/TrueTypeFont.cs

@@ -137,7 +137,7 @@ public bool TryGetBoundingBox(int characterCode, Func<int, int?> characterCodeTo
         /// </summary>
         public bool TryGetPath(int characterCode, Func<int, int?> characterCodeToGlyphId, out IReadOnlyList<PdfSubpath> path)
         {
-            path = EmptyArray<PdfSubpath>.Instance;
+            path = null;
 
             if (!TryGetGlyphIndex(characterCode, characterCodeToGlyphId, out var index)
                 || TableRegister.GlyphTable == null)

src/UglyToad.PdfPig.Fonts/Type1/CharStrings/Type1CharstringDecryptedBytes.cs

@@ -4,7 +4,7 @@
     using System.Collections.Generic;
     using System.Globalization;
 
-    internal class Type1CharstringDecryptedBytes
+    internal sealed class Type1CharstringDecryptedBytes
     {
         public IReadOnlyList<byte> Bytes { get; }
 
@@ -18,7 +18,7 @@ public Type1CharstringDecryptedBytes(IReadOnlyList<byte> bytes, int index)
         {
             Bytes = bytes ?? throw new ArgumentNullException(nameof(bytes));
             Index = index;
-            Name = ".notdef";
+            Name = GlyphList.NotDefined;
             Source = SourceType.Subroutine;
         }