Implement clipping in ProcessFormXObject()

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

@@ -2,7 +2,7 @@
 	<PropertyGroup>
 		<TargetFrameworks>netstandard2.0;net462;net471;net6.0;net8.0</TargetFrameworks>
 		<LangVersion>12</LangVersion>
-		<Version>0.1.10</Version>
+		<Version>0.1.11-alpha001</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;net462;net471;net6.0;net8.0</TargetFrameworks>
 		<LangVersion>12</LangVersion>
-		<Version>0.1.10</Version>
+		<Version>0.1.11-alpha001</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;net462;net471;net6.0;net8.0</TargetFrameworks>
 		<LangVersion>12</LangVersion>
-		<Version>0.1.10</Version>
+		<Version>0.1.11-alpha001</Version>
 		<IsTestProject>False</IsTestProject>
 		<GenerateDocumentationFile>true</GenerateDocumentationFile>
 		<SignAssembly>true</SignAssembly>

src/UglyToad.PdfPig.Tests/Images/Jpeg2000HelperTests.cs

@@ -2,12 +2,13 @@
 {
     using PdfPig.Images;
     using System;
+    using UglyToad.PdfPig.Tests.Integration;
 
     public class Jpeg2000HelperTests
     {
         private static readonly Lazy<string> DocumentFolder = new Lazy<string>(() => Path.GetFullPath(Path.Combine(AppDomain.CurrentDomain.BaseDirectory, "..", "..", "..", "Images", "Files", "Jpx")));
 
-        public static IEnumerable<object[]> GetAllDocuments
+        public static IEnumerable<object[]> GetAllJp2Files
         {
             get
             {
@@ -15,24 +16,40 @@
             }
         }
 
+        [Theory]
+        [MemberData(nameof(GetAllJp2Files))]
+        public void GetJp2BitsPerComponent_ReturnsCorrectBitsPerComponent_WhenValidInput(string path)
+        {
+            byte[] image = File.ReadAllBytes(Path.Combine(DocumentFolder.Value, path));
+            Assert.Equal(8, Jpeg2000Helper.GetBitsPerComponent(image));
+        }
+
         [Fact]
         public void GetJp2BitsPerComponent_ThrowsException_WhenInputIsTooShort()
         {
-            Assert.Throws<InvalidOperationException>(() => Jpeg2000Helper.GetJp2BitsPerComponent(new byte[11]));
+            Assert.Throws<InvalidOperationException>(() => Jpeg2000Helper.GetBitsPerComponent(new byte[11]));
         }
         
         [Fact]
         public void GetJp2BitsPerComponent_ThrowsException_WhenSignatureBoxIsInvalid()
         {
-            Assert.Throws<InvalidOperationException>(() => Jpeg2000Helper.GetJp2BitsPerComponent(new byte[12]));
+            Assert.Throws<InvalidOperationException>(() => Jpeg2000Helper.GetBitsPerComponent(new byte[12]));
         }
 
-        [Theory]
-        [MemberData(nameof(GetAllDocuments))]
-        public void GetJp2BitsPerComponent_ReturnsCorrectBitsPerComponent_WhenValidInput(string path)
+        [Fact]
+        public void GetJp2BitsPerComponentJ2K()
         {
-            byte[] image = File.ReadAllBytes(Path.Combine(DocumentFolder.Value, path));
-            Assert.Equal(8, Jpeg2000Helper.GetJp2BitsPerComponent(image));
+            string path = IntegrationHelpers.GetSpecificTestDocumentPath("GHOSTSCRIPT-688999-2.pdf");
+
+            using (var document = PdfDocument.Open(path))
+            {
+                var page1 = document.GetPage(1);
+                var jpxImage = page1.GetImages().Single();
+
+                var bpc = Jpeg2000Helper.GetBitsPerComponent(jpxImage.RawBytes);
+
+                Assert.Equal(8, bpc);
+            }
         }
     }
 }

src/UglyToad.PdfPig.Tests/Integration/StreamProcessorTests.cs

@@ -231,6 +231,11 @@
                 // No op
             }
 
+            protected override void ClipToRectangle(PdfRectangle rectangle, FillingRule clippingRule)
+            {
+                // No op
+            }
+
             public override void PaintShading(NameToken shadingName)
             {
                 // No op

src/UglyToad.PdfPig.Tests/Integration/VisualVerification/SkiaHelpers/SkiaGlyphStreamProcessor.cs

@@ -206,6 +206,11 @@
             // No op
         }
 
+        protected override void ClipToRectangle(PdfRectangle rectangle, FillingRule clippingRule)
+        {
+            // No op
+        }
+
         public override void PaintShading(NameToken shadingName)
         {
             // No op

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

@@ -2,7 +2,7 @@
 	<PropertyGroup>
 		<TargetFrameworks>netstandard2.0;net462;net471;net6.0;net8.0</TargetFrameworks>
 		<LangVersion>12</LangVersion>
-		<Version>0.1.10</Version>
+		<Version>0.1.11-alpha001</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;net462;net471;net6.0;net8.0</TargetFrameworks>
 		<LangVersion>12</LangVersion>
-		<Version>0.1.10</Version>
+		<Version>0.1.11-alpha001</Version>
 		<IsTestProject>False</IsTestProject>
 		<GenerateDocumentationFile>true</GenerateDocumentationFile>
 		<SignAssembly>true</SignAssembly>

src/UglyToad.PdfPig/Geometry/GeometryExtensions.cs

@@ -381,6 +381,19 @@
         #endregion
 
         #region PdfRectangle
+        /// <summary>
+        /// Converts a <see cref="PdfRectangle"/> into its <see cref="PdfPath"/> representation.
+        /// </summary>
+        public static PdfPath ToPdfPath(this PdfRectangle rectangle)
+        {
+            var clippingSubpath = new PdfSubpath();
+            clippingSubpath.Rectangle(rectangle.BottomLeft.X,
+                rectangle.BottomLeft.Y,
+                rectangle.Width,
+                rectangle.Height);
+            return new PdfPath() { clippingSubpath };
+        }
+
         /// <summary>
         /// Whether the point is located inside the rectangle.
         /// </summary>

src/UglyToad.PdfPig/Graphics/BaseStreamProcessor.cs

@@ -145,15 +145,8 @@
         /// </summary>
         protected static PdfPath GetInitialClipping(CropBox cropBox)
         {
-            var cropBoxBounds = cropBox.Bounds;
-
-            // initiate CurrentClippingPath to cropBox
-            var clippingSubpath = new PdfSubpath();
-            clippingSubpath.Rectangle(cropBoxBounds.BottomLeft.X,
-                cropBoxBounds.BottomLeft.Y,
-                cropBoxBounds.Width,
-                cropBoxBounds.Height);
-            var clippingPath = new PdfPath() { clippingSubpath };
+            // Initiate CurrentClippingPath to cropBox
+            var clippingPath = cropBox.Bounds.ToPdfPath();
             clippingPath.SetClipping(FillingRule.EvenOdd);
             return clippingPath;
         }
@@ -581,7 +574,14 @@
                 new MemoryInputBytes(contentStream),
                 ParsingOptions.Logger);
 
-            // 3. We don't respect clipping currently.
+            // 3. Clip according to the form dictionary's BBox entry.
+            if (formStream.StreamDictionary.TryGet<ArrayToken>(NameToken.Bbox, PdfScanner, out var bboxToken))
+            {
+                var points = bboxToken.Data.OfType<NumericToken>().Select(x => x.Double).ToArray();
+                PdfRectangle bbox = new PdfRectangle(points[0], points[1], points[2], points[3]);
+                PdfRectangle transformedBox = startState.CurrentTransformationMatrix.Transform(bbox);
+                ClipToRectangle(transformedBox, FillingRule.EvenOdd); // TODO - Check that Even Odd is valid
+            }
 
             // 4. Paint the objects.
             bool hasCircularReference = HasFormXObjectCircularReference(formStream, xObjectName, operations);
@@ -668,6 +668,9 @@
         /// <inheritdoc/>
         public abstract void ModifyClippingIntersect(FillingRule clippingRule);
 
+        /// <inheritdoc/>
+        protected abstract void ClipToRectangle(PdfRectangle rectangle, FillingRule clippingRule);
+
         /// <inheritdoc/>
         public virtual void SetNamedGraphicsState(NameToken stateName)
         {

src/UglyToad.PdfPig/Graphics/ContentStreamProcessor.cs

@@ -4,7 +4,6 @@ namespace UglyToad.PdfPig.Graphics
 {
     using System;
     using System.Collections.Generic;
-    using Colors;
     using Content;
     using Filters;
     using Geometry;
@@ -212,7 +211,7 @@ namespace UglyToad.PdfPig.Graphics
             return point;
         }
 
-        public void AddCurrentSubpath() // Not an override
+        private void AddCurrentSubpath()
         {
             if (CurrentSubpath is null)
             {
@@ -385,16 +384,12 @@ namespace UglyToad.PdfPig.Graphics
             var currentState = GetCurrentState();
             if (CurrentPath.IsStroked)
             {
-                CurrentPath.LineDashPattern = currentState.LineDashPattern;
-                CurrentPath.StrokeColor = currentState.CurrentStrokingColor;
-                CurrentPath.LineWidth = currentState.LineWidth;
-                CurrentPath.LineCapStyle = currentState.CapStyle;
-                CurrentPath.LineJoinStyle = currentState.JoinStyle;
+                CurrentPath.SetStrokeDetails(currentState);
             }
 
             if (CurrentPath.IsFilled)
             {
-                CurrentPath.FillColor = currentState.CurrentNonStrokingColor;
+                CurrentPath.SetFillDetails(currentState);
             }
 
             if (ParsingOptions.ClipPaths)
@@ -427,7 +422,8 @@ namespace UglyToad.PdfPig.Graphics
 
             if (ParsingOptions.ClipPaths)
             {
-                var currentClipping = GetCurrentState().CurrentClippingPath!;
+                var graphicsState = GetCurrentState();
+                var currentClipping = graphicsState.CurrentClippingPath!;
                 currentClipping.SetClipping(clippingRule);
 
                 var newClippings = CurrentPath.Clip(currentClipping, ParsingOptions.Logger);
@@ -437,11 +433,32 @@ namespace UglyToad.PdfPig.Graphics
                 }
                 else
                 {
-                    GetCurrentState().CurrentClippingPath = newClippings;
+                    graphicsState.CurrentClippingPath = newClippings;
                 }
             }
         }
 
+        protected override void ClipToRectangle(PdfRectangle rectangle, FillingRule clippingRule)
+        {
+            // https://github.com/apache/pdfbox/blob/f4bfe47de37f6fe69e8f98b164c3546facfd5e91/pdfbox/src/main/java/org/apache/pdfbox/contentstream/PDFStreamEngine.java#L611
+            var graphicsState = GetCurrentState();
+            var clip = rectangle.ToPdfPath();
+            clip.SetClipping(clippingRule);
+
+            var currentClipping = graphicsState.CurrentClippingPath!;
+            currentClipping.SetClipping(clippingRule);
+            var newClippings = clip.Clip(currentClipping, ParsingOptions.Logger);
+
+            if (newClippings is null)
+            {
+                ParsingOptions.Logger.Warn("Empty clipping path found. Clipping path not updated.");
+            }
+            else
+            {
+                graphicsState.CurrentClippingPath = newClippings;
+            }
+        }
+
         protected override void RenderInlineImage(InlineImage inlineImage)
         {
             images.Add(Union<XObjectContentRecord, InlineImage>.Two(inlineImage));

src/UglyToad.PdfPig/Graphics/PdfPath.cs

@@ -6,7 +6,9 @@
     using UglyToad.PdfPig.Graphics.Core;
 
     /// <summary>
-    /// A path is made up of one or more disconnected subpaths, each comprising a sequence of connected segments. The topology of the path is unrestricted: it may be concave or convex, may contain multiple subpaths representing disjoint areas, and may intersect itself in arbitrary ways.
+    /// A path is made up of one or more disconnected subpaths, each comprising a sequence of connected segments.
+    /// The topology of the path is unrestricted: it may be concave or convex, may contain multiple subpaths representing
+    /// disjoint areas, and may intersect itself in arbitrary ways.
     /// <para>A path shall be composed of straight and curved line segments, which may connect to one another or may be disconnected.</para>
     /// </summary>
     public class PdfPath : List<PdfSubpath>
@@ -29,7 +31,7 @@
         /// <summary>
         /// The fill color.
         /// </summary>
-        public IColor? FillColor { get; internal set; }
+        public IColor? FillColor { get; private set; }
 
         /// <summary>
         /// Returns true if the path is stroked.
@@ -39,31 +41,31 @@
         /// <summary>
         /// The stroke color.
         /// </summary>
-        public IColor? StrokeColor { get; internal set; }
+        public IColor? StrokeColor { get; private set; }
 
         /// <summary>
         /// Thickness in user space units of path to be stroked.
         /// </summary>
-        public double LineWidth { get; internal set; }
+        public double LineWidth { get; private set; }
 
         /// <summary>
         /// The pattern to be used for stroked lines.
         /// </summary>
-        public LineDashPattern? LineDashPattern { get; internal set; }
+        public LineDashPattern? LineDashPattern { get; private set; }
 
         /// <summary>
         /// The cap style to be used for stroked lines.
         /// </summary>
-        public LineCapStyle LineCapStyle { get; internal set; }
+        public LineCapStyle LineCapStyle { get; private set; }
 
         /// <summary>
         /// The join style to be used for stroked lines.
         /// </summary>
-        public LineJoinStyle LineJoinStyle { get; internal set; }
+        public LineJoinStyle LineJoinStyle { get; private set; }
 
         /// <summary>
-        /// Set the clipping mode for this path and IsClipping to true.
-        /// <para>IsFilled and IsStroked flags will be set to false.</para>
+        /// Set the clipping mode for this path and <c>IsClipping</c> to <c>true</c>.
+        /// <para><c>IsFilled</c> and <c>IsStroked</c> flags will be set to <c>false</c>.</para>
         /// </summary>
         public void SetClipping(FillingRule fillingRule)
         {
@@ -74,7 +76,7 @@
         }
 
         /// <summary>
-        /// Set the filling rule for this path and IsFilled to true.
+        /// Set the filling rule for this path and <c>IsFilled</c> to <c>true</c>.
         /// </summary>
         public void SetFilled(FillingRule fillingRule)
         {
@@ -83,13 +85,35 @@
         }
 
         /// <summary>
-        /// Set IsStroked to true.
+        /// Set <c>IsStroked</c> to <c>true</c>.
         /// </summary>
         public void SetStroked()
         {
             IsStroked = true;
         }
 
+        /// <summary>
+        /// Set the path stroke details, i.e. <c>LineDashPattern</c>, <c>StrokeColor</c>, <c>LineWidth</c>, <c>LineCapStyle</c> and <c>LineJoinStyle</c>.
+        /// </summary>
+        /// <param name="graphicsState">The current graphics state.</param>
+        public void SetStrokeDetails(CurrentGraphicsState graphicsState)
+        {
+            LineDashPattern = graphicsState.LineDashPattern;
+            StrokeColor = graphicsState.CurrentStrokingColor;
+            LineWidth = graphicsState.LineWidth;
+            LineCapStyle = graphicsState.CapStyle;
+            LineJoinStyle = graphicsState.JoinStyle;
+        }
+
+        /// <summary>
+        /// Set the path fill details, i.e. <c>FillColor</c>.
+        /// </summary>
+        /// <param name="graphicsState">The current graphics state.</param>
+        public void SetFillDetails(CurrentGraphicsState graphicsState)
+        {
+            FillColor = graphicsState.CurrentNonStrokingColor;
+        }
+
         /// <summary>
         /// Create a clone with no Subpaths.
         /// </summary>

src/UglyToad.PdfPig/Images/Jpeg2000Helper.cs

@@ -8,7 +8,7 @@
         /// <summary>
         /// Get bits per component values for Jp2 (Jpx) encoded images (first component).
         /// </summary>
-        public static byte GetJp2BitsPerComponent(ReadOnlySpan<byte> jp2Bytes)
+        public static byte GetBitsPerComponent(ReadOnlySpan<byte> jp2Bytes)
         {
             // Ensure the input has at least 12 bytes for the signature box
             if (jp2Bytes.Length < 12)
@@ -18,16 +18,21 @@
 
             // Verify the JP2 signature box
             uint length = BinaryPrimitives.ReadUInt32BigEndian(jp2Bytes.Slice(0, 4));
-            uint type = BinaryPrimitives.ReadUInt32BigEndian(jp2Bytes.Slice(4, 4));
-            uint magic = BinaryPrimitives.ReadUInt32BigEndian(jp2Bytes.Slice(8, 4));
-
-            if (length != 0x0000000C || type != 0x6A502020 || magic != 0x0D0A870A)
+            if (length == 0xFF4FFF51)
             {
-                throw new InvalidOperationException("Invalid JP2 signature box.");
+                // J2K format detected (SOC marker) (See GHOSTSCRIPT-688999-2.pdf)
+                return ParseCodestream(jp2Bytes);
             }
 
-            // Proceed to parse JP2 boxes
-            return ParseBoxes(jp2Bytes.Slice(12));
+            uint type = BinaryPrimitives.ReadUInt32BigEndian(jp2Bytes.Slice(4, 4));
+            uint magic = BinaryPrimitives.ReadUInt32BigEndian(jp2Bytes.Slice(8, 4));
+            if (length == 0x0000000C && type == 0x6A502020 && magic == 0x0D0A870A)
+            {
+                // JP2 format detected
+                return ParseBoxes(jp2Bytes.Slice(12));
+            }
+
+            throw new InvalidOperationException("Invalid JP2 or J2K signature.");
         }
 
         private static byte ParseBoxes(ReadOnlySpan<byte> jp2Bytes)
@@ -37,7 +42,7 @@
             {
                 if (offset + 8 > jp2Bytes.Length)
                 {
-                    throw new InvalidOperationException("Invalid JP2 box structure.");
+                    throw new InvalidOperationException("Invalid JP2 or J2K box structure.");
                 }
 
                 // Read box length and type
@@ -55,7 +60,7 @@
                 offset += (int)(boxLength > 0 ? boxLength : 8); // Box length of 0 means the rest of the file
             }
 
-            throw new InvalidOperationException("Codestream box not found in JP2 file.");
+            throw new InvalidOperationException("Codestream box not found in JP2 or J2K file.");
         }
 
         private static byte ParseCodestream(ReadOnlySpan<byte> codestream)

src/UglyToad.PdfPig/UglyToad.PdfPig.csproj

@@ -2,7 +2,7 @@
 	<PropertyGroup>
 		<TargetFrameworks>netstandard2.0;net462;net471;net6.0;net8.0</TargetFrameworks>
 		<LangVersion>12</LangVersion>
-		<Version>0.1.10</Version>
+		<Version>0.1.11-alpha001</Version>
 		<IsTestProject>False</IsTestProject>
 		<GenerateDocumentationFile>true</GenerateDocumentationFile>
 		<SignAssembly>true</SignAssembly>

src/UglyToad.PdfPig/XObjects/XObjectFactory.cs

@@ -60,11 +60,11 @@
                     if (dictionary.TryGet(NameToken.BitsPerComponent, out NumericToken? bitsPerComponentToken))
                     {
                         bitsPerComponent = bitsPerComponentToken.Int;
-                        System.Diagnostics.Debug.Assert(bitsPerComponent == Jpeg2000Helper.GetJp2BitsPerComponent(xObject.Stream.Data.Span));
+                        System.Diagnostics.Debug.Assert(bitsPerComponent == Jpeg2000Helper.GetBitsPerComponent(xObject.Stream.Data.Span));
                     }
                     else
                     {
-                        bitsPerComponent = Jpeg2000Helper.GetJp2BitsPerComponent(xObject.Stream.Data.Span);
+                        bitsPerComponent = Jpeg2000Helper.GetBitsPerComponent(xObject.Stream.Data.Span);
                         System.Diagnostics.Debug.Assert(new int[] { 1, 2, 4, 8, 16 }.Contains(bitsPerComponent));
                     }
                 }

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>0.1.10</Version>
+    <Version>0.1.11-alpha001</Version>
     <AssemblyVersion>0.1.10.0</AssemblyVersion>
     <FileVersion>0.1.10.0</FileVersion>
     <PackageIconUrl>https://raw.githubusercontent.com/UglyToad/PdfPig/master/documentation/pdfpig.png</PackageIconUrl>