allow indirect references in xref streams as an attempt to fix #457

Fix #458 - GrahamScan exception

src/UglyToad.PdfPig.Core/PdfPoint.cs

@@ -67,7 +67,7 @@
         {
             return new PdfPoint(X + dx, Y);
         }
-        
+
         /// <summary>
         /// Creates a new <see cref="PdfPoint"/> which is the current point moved in the y direction relative to its current position by a value.
         /// </summary>

src/UglyToad.PdfPig.Core/UglyToad.PdfPig.Core.csproj

@@ -2,7 +2,7 @@
   <PropertyGroup>
     <TargetFrameworks>netstandard2.0;net45;net451;net452;net46;net461;net462;net47;net6.0</TargetFrameworks>
     <LangVersion>latest</LangVersion>
-    <Version>0.1.6</Version>
+    <Version>0.1.7-alpha-001</Version>
     <IsTestProject>False</IsTestProject>
     <GenerateDocumentationFile>true</GenerateDocumentationFile>
     <SignAssembly>true</SignAssembly>

src/UglyToad.PdfPig.DocumentLayoutAnalysis/UglyToad.PdfPig.DocumentLayoutAnalysis.csproj

@@ -2,7 +2,7 @@
   <PropertyGroup>
     <TargetFrameworks>netstandard2.0;net45;net451;net452;net46;net461;net462;net47;net6.0</TargetFrameworks>
     <LangVersion>latest</LangVersion>
-    <Version>0.1.6</Version>
+    <Version>0.1.7-alpha-001</Version>
     <IsTestProject>False</IsTestProject>
     <GenerateDocumentationFile>true</GenerateDocumentationFile>
     <SignAssembly>true</SignAssembly>

src/UglyToad.PdfPig.Fonts/UglyToad.PdfPig.Fonts.csproj

@@ -2,7 +2,7 @@
   <PropertyGroup>
     <TargetFrameworks>netstandard2.0;net45;net451;net452;net46;net461;net462;net47;net6.0</TargetFrameworks>
     <LangVersion>latest</LangVersion>
-    <Version>0.1.6</Version>
+    <Version>0.1.7-alpha-001</Version>
     <IsTestProject>False</IsTestProject>
     <GenerateDocumentationFile>true</GenerateDocumentationFile>
     <SignAssembly>true</SignAssembly>

src/UglyToad.PdfPig.Tests/Geometry/PdfPointTests.cs

@@ -528,6 +528,52 @@
                 }
             }
         };
+
+        public static IEnumerable<object[]> Issue458Data => new[]
+        {
+            new object[]
+            {
+                new PdfPoint[]
+                {
+                    new PdfPoint(134.74199999999985, 1611.657),
+                    new PdfPoint(277.58043749999985, 1611.657),
+                    new PdfPoint(314.74199999999985, 1611.657),
+                    new PdfPoint(507.0248828124999, 1611.657),
+                    new PdfPoint(545.1419999999998, 1611.657),
+                    new PdfPoint(632.9658281249997, 1611.657),
+                    new PdfPoint(668.5709999999998, 1611.657),
+                    new PdfPoint(831.3395078125002, 1611.657),
+                    new PdfPoint(868.1139999999998, 1611.657),
+                    new PdfPoint(892.8907187499999, 1611.657),
+                    new PdfPoint(1010.0569999999999, 1611.6569999999997),
+                    new PdfPoint(1046.2174003906248, 1611.7654812011715),
+                    new PdfPoint(1010.0569999999999, 1611.657),
+                    new PdfPoint(1046.2174003906248, 1611.7654812011717),
+                    new PdfPoint(1085.145, 1611.8822639999996),
+                    new PdfPoint(1255.484941406251, 1612.3932838242179),
+                    new PdfPoint(1301.144, 1612.5302609999994),
+                    new PdfPoint(1359.0006406250002, 1612.7038309218747),
+                    new PdfPoint(1301.144, 1612.5302609999997),
+                    new PdfPoint(1359.0006406250002, 1612.703830921875),
+                    new PdfPoint(1400.915, 1612.8295739999996),
+                    new PdfPoint(1505.379062499999, 1613.1429661874993),
+                    new PdfPoint(1400.915, 1612.8295739999999),
+                    new PdfPoint(1505.379062499999, 1613.1429661874995),
+                    new PdfPoint(1543.886, 1613.2584869999996),
+                    new PdfPoint(1600.9401015625003, 1613.4296493046872),
+                    new PdfPoint(1641.597, 1613.5516199999997),
+                    new PdfPoint(1764.5577421874998, 1613.9205022265612),
+                    new PdfPoint(1641.597, 1613.55162),
+                    new PdfPoint(1764.5577421874998, 1613.9205022265614)
+                },
+                new PdfPoint[]
+                {
+                    new PdfPoint(134.74199999999985, 1611.657),
+                    new PdfPoint(1010.0569999999999, 1611.6569999999997),
+                    new PdfPoint(1764.5577421874998, 1613.9205022265614),
+                }
+            }
+        };
         #endregion
 
         [Fact]
@@ -570,7 +616,6 @@
             }
         }
 
-
         [Theory]
         [MemberData(nameof(MinimumAreaRectangleData))]
         public void MinimumAreaRectangle(PdfPoint[] points, PdfPoint[] expected)
@@ -585,5 +630,51 @@
                 Assert.Equal(expected[i], mar[i], PointComparer);
             }
         }
+
+        [Theory]
+        [MemberData(nameof(Issue458Data))]
+        public void Issue458(PdfPoint[] points, PdfPoint[] expected)
+        {
+            /*
+             * https://github.com/UglyToad/PdfPig/issues/458
+             * An unhandled exception of type 'System.ArgumentOutOfRangeException' occurred in System.Linq.dll: 'Specified argument was out of the range of valid values.'
+             *  at System.Linq.ThrowHelper.ThrowArgumentOutOfRangeException(ExceptionArgument argument)
+             *  at System.Linq.Enumerable.ElementAt[TSource](IEnumerable`1 source, Int32 index)
+             */
+
+            var result = GeometryExtensions.GrahamScan(points);
+
+            // Data is noisy so we just check it does not throw an exception
+            // and that key points are present (other points might be present,
+            // e.g. 'not enough equal dupplicates' or 'not collinear enough' points)
+            foreach (var point in expected)
+            {
+                Assert.Contains(point, result);
+            }
+        }
+
+        [Theory]
+        [MemberData(nameof(Issue458Data))]
+        public void Issue458_Inv(PdfPoint[] points, PdfPoint[] expected)
+        {
+            /*
+             * https://github.com/UglyToad/PdfPig/issues/458
+             * An unhandled exception of type 'System.ArgumentOutOfRangeException' occurred in System.Linq.dll: 'Specified argument was out of the range of valid values.'
+             *  at System.Linq.ThrowHelper.ThrowArgumentOutOfRangeException(ExceptionArgument argument)
+             *  at System.Linq.Enumerable.ElementAt[TSource](IEnumerable`1 source, Int32 index)
+             */
+            var pointsInv = points.Select(p => new PdfPoint(p.Y, p.X)).ToArray();
+            var expectedInv = expected.Select(p => new PdfPoint(p.Y, p.X)).ToArray();
+
+            var result = GeometryExtensions.GrahamScan(pointsInv);
+
+            // Data is noisy so we just check it does not throw an exception
+            // and that key points are present (other points might be present,
+            // e.g. 'not enough equal dupplicates' or 'not collinear enough' points)
+            foreach (var point in expectedInv)
+            {
+                Assert.Contains(point, result);
+            }
+        }
     }
 }

src/UglyToad.PdfPig.Tokenization/UglyToad.PdfPig.Tokenization.csproj

@@ -2,7 +2,7 @@
   <PropertyGroup>
     <TargetFrameworks>netstandard2.0;net45;net451;net452;net46;net461;net462;net47;net6.0</TargetFrameworks>
     <LangVersion>latest</LangVersion>
-    <Version>0.1.6</Version>
+    <Version>0.1.7-alpha-001</Version>
     <IsTestProject>False</IsTestProject>
     <GenerateDocumentationFile>true</GenerateDocumentationFile>
     <SignAssembly>true</SignAssembly>

src/UglyToad.PdfPig.Tokens/UglyToad.PdfPig.Tokens.csproj

@@ -2,7 +2,7 @@
   <PropertyGroup>
     <TargetFrameworks>netstandard2.0;net45;net451;net452;net46;net461;net462;net47;net6.0</TargetFrameworks>
     <LangVersion>latest</LangVersion>
-    <Version>0.1.6</Version>
+    <Version>0.1.7-alpha-001</Version>
     <IsTestProject>False</IsTestProject>
     <GenerateDocumentationFile>true</GenerateDocumentationFile>
     <SignAssembly>true</SignAssembly>

src/UglyToad.PdfPig/Encryption/AesEncryptionHelper.cs

@@ -1,7 +1,6 @@
 namespace UglyToad.PdfPig.Encryption
 {
     using System;
-    using System.Diagnostics;
     using System.IO;
     using System.Security.Cryptography;
 
@@ -29,11 +28,6 @@
 
                 var buffer = new byte[256];
 
-                if (data.Length > 256)
-                {
-                    Debugger.Break();
-                }
-
                 using (var decryptor = rijndael.CreateDecryptor(rijndael.Key, rijndael.IV))
                 using (var input = new MemoryStream(data))
                 using (var output = new MemoryStream())

src/UglyToad.PdfPig/Filters/DefaultFilterProvider.cs

@@ -1,11 +1,11 @@
 namespace UglyToad.PdfPig.Filters
 {
+    using Core;
     using System;
     using System.Collections.Generic;
     using System.Linq;
-    using Core;
     using Tokens;
-    using UglyToad.PdfPig.Util;
+    using Util;
 
     /// <inheritdoc />
     /// <summary>

src/UglyToad.PdfPig/Geometry/GeometryExtensions.cs

@@ -269,11 +269,12 @@
 
             double polarAngle(PdfPoint point1, PdfPoint point2)
             {
+                // This is used for grouping, we could use Math.Round()
                 return Math.Atan2(point2.Y - point1.Y, point2.X - point1.X) % Math.PI;
             }
 
-            Stack<PdfPoint> stack = new Stack<PdfPoint>();
-            var sortedPoints = points.OrderBy(p => p.Y).ThenBy(p => p.X).ToList();
+            var stack = new Stack<PdfPoint>();
+            var sortedPoints = points.OrderBy(p => p.X).ThenBy(p => p.Y).ToList();
             var P0 = sortedPoints[0];
             var groups = sortedPoints.Skip(1).GroupBy(p => polarAngle(P0, p)).OrderBy(g => g.Key);
 
@@ -309,7 +310,7 @@
             for (int i = 2; i < sortedPoints.Count; i++)
             {
                 var point = sortedPoints[i];
-                while (!ccw(stack.ElementAt(1), stack.Peek(), point))
+                while (stack.Count > 1 && !ccw(stack.ElementAt(1), stack.Peek(), point))
                 {
                     stack.Pop();
                 }

src/UglyToad.PdfPig/Parser/FileStructure/CrossReferenceParser.cs

@@ -269,7 +269,7 @@
                 return false;
             }
             
-            xrefTablePart = crossReferenceStreamParser.Parse(objByteOffset, fromTableAtOffset, objectStream);
+            xrefTablePart = crossReferenceStreamParser.Parse(objByteOffset, fromTableAtOffset, objectStream, pdfScanner);
 
             return true;
         }

src/UglyToad.PdfPig/Parser/Parts/CrossReference/CrossReferenceStreamParser.cs

@@ -4,6 +4,7 @@
     using Core;
     using Filters;
     using PdfPig.CrossReference;
+    using Tokenization.Scanner;
     using Tokens;
     using Util;
 
@@ -19,9 +20,9 @@
         /// <summary>
         /// Parses through the unfiltered stream and populates the xrefTable HashMap.
         /// </summary>
-        public CrossReferenceTablePart Parse(long streamOffset, long? fromTableAtOffset, StreamToken stream)
+        public CrossReferenceTablePart Parse(long streamOffset, long? fromTableAtOffset, StreamToken stream, IPdfTokenScanner pdfTokenScanner)
         {
-            var decoded = stream.Decode(filterProvider);
+            var decoded = stream.DecodeWithLookup(filterProvider, pdfTokenScanner);
 
             var fieldSizes = new CrossReferenceStreamFieldSize(stream.StreamDictionary);
 

src/UglyToad.PdfPig/PdfExtensions.cs

@@ -45,14 +45,18 @@
             }
 
             return typedToken;
-        }
-
+        }
+
         /// <summary>
         /// Get the decoded data from this stream.
         /// </summary>
         public static IReadOnlyList<byte> Decode(this StreamToken stream, IFilterProvider filterProvider)
+            => DecodeWithLookup(stream, filterProvider, null);
+        internal static IReadOnlyList<byte> DecodeWithLookup(this StreamToken stream, IFilterProvider filterProvider, IPdfTokenScanner pdfTokenScanner)
         {
-            var filters = filterProvider.GetFilters(stream.StreamDictionary);
+            var filters = pdfTokenScanner != null
+                ? new FilterProviderWithLookup(filterProvider).GetFilters(stream.StreamDictionary, pdfTokenScanner)
+                : filterProvider.GetFilters(stream.StreamDictionary);
 
             var transform = stream.Data;
             for (var i = 0; i < filters.Count; i++)

src/UglyToad.PdfPig/PdfFonts/Parser/Handlers/TrueTypeFontHandler.cs

@@ -107,18 +107,25 @@
             CMap toUnicodeCMap = null;
             if (dictionary.TryGet(NameToken.ToUnicode, out var toUnicodeObj))
             {
-                var toUnicode = DirectObjectFinder.Get<StreamToken>(toUnicodeObj, pdfScanner);
-
-                var decodedUnicodeCMap = toUnicode.Decode(filterProvider, pdfScanner);
-
-                if (decodedUnicodeCMap != null)
+                try
                 {
-                    toUnicodeCMap = CMapCache.Parse(new ByteArrayInputBytes(decodedUnicodeCMap));
+                    var toUnicode = DirectObjectFinder.Get<StreamToken>(toUnicodeObj, pdfScanner);
+
+                    var decodedUnicodeCMap = toUnicode.Decode(filterProvider, pdfScanner);
+
+                    if (decodedUnicodeCMap != null)
+                    {
+                        toUnicodeCMap = CMapCache.Parse(new ByteArrayInputBytes(decodedUnicodeCMap));
+                    }
+                }
+                catch (Exception ex)
+                {
+                    log.Error("Failed to decode ToUnicode CMap for a TrueType font in file due to an exception.", ex);
                 }
             }
 
             Encoding encoding = encodingReader.Read(dictionary, descriptor);
-            
+
             if (encoding == null && font?.TableRegister?.CMapTable != null
                                  && font.TableRegister.PostScriptTable?.GlyphNames != null)
             {
@@ -199,7 +206,7 @@
                             $"Expected a TrueType font in the TrueType font descriptor, instead it was {descriptor.FontFile.FileType}.");
                     }
                 }
-                
+
                 var font = TrueTypeFontParser.Parse(new TrueTypeDataBytes(new ByteArrayInputBytes(fontFile)));
 
                 return font;

src/UglyToad.PdfPig/UglyToad.PdfPig.csproj

@@ -2,7 +2,7 @@
   <PropertyGroup>
     <TargetFrameworks>netstandard2.0;net45;net451;net452;net46;net461;net462;net47;net6.0</TargetFrameworks>
     <LangVersion>latest</LangVersion>
-    <Version>0.1.6</Version>
+    <Version>0.1.7-alpha-001</Version>
     <IsTestProject>False</IsTestProject>
     <GenerateDocumentationFile>true</GenerateDocumentationFile>
     <SignAssembly>true</SignAssembly>

tools/UglyToad.PdfPig.Package/UglyToad.PdfPig.Package.csproj

@@ -11,7 +11,7 @@
     <PackageTags>PDF;Reader;Document;Adobe;PDFBox;PdfPig;pdf-extract;pdf-to-text;pdf;file;text;C#;dotnet;.NET</PackageTags>
     <RepositoryUrl>https://github.com/UglyToad/PdfPig</RepositoryUrl>
     <GenerateDocumentationFile>true</GenerateDocumentationFile>
-    <Version>0.1.6</Version>
+    <Version>0.1.7-alpha-001</Version>
     <AssemblyVersion>0.1.5.0</AssemblyVersion>
     <FileVersion>0.1.5.0</FileVersion>
     <PackageIconUrl>https://raw.githubusercontent.com/UglyToad/PdfPig/master/documentation/pdfpig.png</PackageIconUrl>