Handle TIFF image with incorrect IPTC data type (long instead of byte)

src/ImageSharp/Formats/Tiff/TiffDecoderMetadataCreator.cs

@@ -1,8 +1,11 @@
 // Copyright (c) Six Labors.
 // Licensed under the Apache License, Version 2.0.
 
+using System;
+using System.Buffers.Binary;
 using System.Collections.Generic;
-
+using System.Linq;
+using System.Runtime.InteropServices;
 using SixLabors.ImageSharp.Metadata;
 using SixLabors.ImageSharp.Metadata.Profiles.Exif;
 using SixLabors.ImageSharp.Metadata.Profiles.Icc;
@@ -56,10 +59,9 @@ namespace SixLabors.ImageSharp.Formats.Experimental.Tiff
 
                     if (coreMetadata.IptcProfile == null)
                     {
-                        IExifValue<byte[]> val = frame.ExifProfile.GetValue(ExifTag.IPTC);
+                        if (TryGetIptc(frame.ExifProfile.Values, out var iptcBytes))
-                        if (val != null)
                         {
-                            coreMetadata.IptcProfile = new IptcProfile(val.Value);
+                            coreMetadata.IptcProfile = new IptcProfile(iptcBytes);
                         }
                     }
 
@@ -77,6 +79,53 @@ namespace SixLabors.ImageSharp.Formats.Experimental.Tiff
             return coreMetadata;
         }
 
+        private static bool TryGetIptc(IReadOnlyList<IExifValue> exifValues, out byte[] iptcBytes)
+        {
+            iptcBytes = null;
+            IExifValue iptc = exifValues.FirstOrDefault(f => f.Tag == ExifTag.IPTC);
+
+            if (iptc != null)
+            {
+                if (iptc.DataType == ExifDataType.Byte)
+                {
+                    iptcBytes = (byte[])iptc.GetValue();
+                    return true;
+                }
+                else if (iptc.DataType == ExifDataType.Long)
+                {
+                    var iptcValues = (uint[])iptc.GetValue();
+                    iptcBytes = new byte[iptcValues.Length * 4];
+                    Buffer.BlockCopy(iptcValues, 0, iptcBytes, 0, iptcValues.Length * 4);
+                    if (iptcBytes[0] == 0x1c)
+                    {
+                        return true;
+                    }
+                    else if (iptcBytes[3] != 0x1c)
+                    {
+                        return false;
+                    }
+
+                    // Probably wrong endianess, swap byte order.
+                    Span<byte> iptcBytesSpan = iptcBytes.AsSpan();
+                    Span<byte> buffer = stackalloc byte[4];
+                    for (int i = 0; i < iptcBytes.Length; i += 4)
+                    {
+                        iptcBytesSpan.Slice(i, 4).CopyTo(buffer);
+                        iptcBytes[i] = buffer[3];
+                        iptcBytes[i + 1] = buffer[2];
+                        iptcBytes[i + 2] = buffer[1];
+                        iptcBytes[i + 3] = buffer[0];
+                    }
+
+                    return true;
+                }
+
+                return false;
+            }
+
+            return false;
+        }
+
         private static TiffBitsPerPixel GetBitsPerPixel(TiffFrameMetadata firstFrameMetaData)
             => (TiffBitsPerPixel)firstFrameMetaData.BitsPerPixel;
     }

tests/ImageSharp.Tests/Formats/Tiff/TiffMetadataTests.cs

@@ -2,6 +2,7 @@
 // Licensed under the Apache License, Version 2.0.
 
 using System;
+using System.Globalization;
 using System.IO;
 using System.Linq;
 
@@ -24,6 +25,8 @@ namespace SixLabors.ImageSharp.Tests.Formats.Tiff
     {
         private readonly Configuration configuration;
 
+        private static TiffDecoder TiffDecoder => new TiffDecoder();
+
         public TiffMetadataTests()
         {
             this.configuration = new Configuration();
@@ -130,12 +133,25 @@ namespace SixLabors.ImageSharp.Tests.Formats.Tiff
             }
         }
 
+        [Theory]
+        [WithFile(InvalidIptcData, PixelTypes.Rgba32)]
+        public void CanDecodeImage_WithIptcDataAsLong<TPixel>(TestImageProvider<TPixel> provider)
+            where TPixel : unmanaged, IPixel<TPixel>
+        {
+            using Image<TPixel> image = provider.GetImage(TiffDecoder);
+
+            Assert.NotNull(image.Metadata.IptcProfile);
+            IptcValue byline = image.Metadata.IptcProfile.Values.FirstOrDefault(data => data.Tag == IptcTag.Byline);
+            Assert.NotNull(byline);
+            Assert.Equal("Studio Mantyniemi", byline.Value);
+        }
+
         [Theory]
         [WithFile(SampleMetadata, PixelTypes.Rgba32)]
         public void BaselineTags<TPixel>(TestImageProvider<TPixel> provider)
             where TPixel : unmanaged, IPixel<TPixel>
         {
-            using (Image<TPixel> image = provider.GetImage(new TiffDecoder()))
+            using (Image<TPixel> image = provider.GetImage(TiffDecoder))
             {
                 TiffMetadata meta = image.Metadata.GetTiffMetadata();
 
@@ -188,7 +204,7 @@ namespace SixLabors.ImageSharp.Tests.Formats.Tiff
         public void SubfileType<TPixel>(TestImageProvider<TPixel> provider)
             where TPixel : unmanaged, IPixel<TPixel>
         {
-            using (Image<TPixel> image = provider.GetImage(new TiffDecoder()))
+            using (Image<TPixel> image = provider.GetImage(TiffDecoder))
             {
                 TiffMetadata meta = image.Metadata.GetTiffMetadata();
                 Assert.NotNull(meta);
@@ -311,7 +327,7 @@ namespace SixLabors.ImageSharp.Tests.Formats.Tiff
             coreMeta.HorizontalResolution = 4500;
             coreMeta.VerticalResolution = 5400;
 
-            var datetime = DateTime.Now.ToString();
+            var datetime = DateTime.Now.ToString(CultureInfo.InvariantCulture);
             frameMeta.ImageDescription = "test ImageDescription";
             frameMeta.DateTime = datetime;
 

tests/ImageSharp.Tests/Metadata/Profiles/IPTC/IptcProfileTests.cs

@@ -5,6 +5,7 @@ using System;
 using System.Collections.Generic;
 using System.IO;
 using System.Linq;
+using SixLabors.ImageSharp.Formats.Experimental.Tiff;
 using SixLabors.ImageSharp.Formats.Jpeg;
 using SixLabors.ImageSharp.Metadata.Profiles.Iptc;
 using SixLabors.ImageSharp.PixelFormats;
@@ -16,6 +17,8 @@ namespace SixLabors.ImageSharp.Tests.Metadata.Profiles.IPTC
     {
         private static JpegDecoder JpegDecoder => new JpegDecoder() { IgnoreMetadata = false };
 
+        private static TiffDecoder TiffDecoder => new TiffDecoder() { IgnoreMetadata = false };
+
         public static IEnumerable<object[]> AllIptcTags()
         {
             foreach (object tag in Enum.GetValues(typeof(IptcTag)))
@@ -100,34 +103,52 @@ namespace SixLabors.ImageSharp.Tests.Metadata.Profiles.IPTC
 
         [Theory]
         [WithFile(TestImages.Jpeg.Baseline.Iptc, PixelTypes.Rgba32)]
-        public void ReadIptcMetadata_Works<TPixel>(TestImageProvider<TPixel> provider)
+        public void ReadIptcMetadata_FromJpg_Works<TPixel>(TestImageProvider<TPixel> provider)
             where TPixel : unmanaged, IPixel<TPixel>
         {
             using (Image<TPixel> image = provider.GetImage(JpegDecoder))
             {
                 Assert.NotNull(image.Metadata.IptcProfile);
                 var iptcValues = image.Metadata.IptcProfile.Values.ToList();
-                ContainsIptcValue(iptcValues, IptcTag.Caption, "description");
+                IptcProfileContainsExpectedValues(iptcValues);
-                ContainsIptcValue(iptcValues, IptcTag.CaptionWriter, "description writer");
-                ContainsIptcValue(iptcValues, IptcTag.Headline, "headline");
-                ContainsIptcValue(iptcValues, IptcTag.SpecialInstructions, "special instructions");
-                ContainsIptcValue(iptcValues, IptcTag.Byline, "author");
-                ContainsIptcValue(iptcValues, IptcTag.BylineTitle, "author title");
-                ContainsIptcValue(iptcValues, IptcTag.Credit, "credits");
-                ContainsIptcValue(iptcValues, IptcTag.Source, "source");
-                ContainsIptcValue(iptcValues, IptcTag.Name, "title");
-                ContainsIptcValue(iptcValues, IptcTag.CreatedDate, "20200414");
-                ContainsIptcValue(iptcValues, IptcTag.City, "city");
-                ContainsIptcValue(iptcValues, IptcTag.SubLocation, "sublocation");
-                ContainsIptcValue(iptcValues, IptcTag.ProvinceState, "province-state");
-                ContainsIptcValue(iptcValues, IptcTag.Country, "country");
-                ContainsIptcValue(iptcValues, IptcTag.Category, "category");
-                ContainsIptcValue(iptcValues, IptcTag.Urgency, "1");
-                ContainsIptcValue(iptcValues, IptcTag.Keywords, "keywords");
-                ContainsIptcValue(iptcValues, IptcTag.CopyrightNotice, "copyright");
             }
         }
 
+        [Theory]
+        [WithFile(TestImages.Tiff.IptcData, PixelTypes.Rgba32)]
+        public void ReadIptcMetadata_FromTiff_Works<TPixel>(TestImageProvider<TPixel> provider)
+            where TPixel : unmanaged, IPixel<TPixel>
+        {
+            using (Image<TPixel> image = provider.GetImage(TiffDecoder))
+            {
+                Assert.NotNull(image.Metadata.IptcProfile);
+                var iptcValues = image.Metadata.IptcProfile.Values.ToList();
+                IptcProfileContainsExpectedValues(iptcValues);
+            }
+        }
+
+        private static void IptcProfileContainsExpectedValues(List<IptcValue> iptcValues)
+        {
+            ContainsIptcValue(iptcValues, IptcTag.Caption, "description");
+            ContainsIptcValue(iptcValues, IptcTag.CaptionWriter, "description writer");
+            ContainsIptcValue(iptcValues, IptcTag.Headline, "headline");
+            ContainsIptcValue(iptcValues, IptcTag.SpecialInstructions, "special instructions");
+            ContainsIptcValue(iptcValues, IptcTag.Byline, "author");
+            ContainsIptcValue(iptcValues, IptcTag.BylineTitle, "author title");
+            ContainsIptcValue(iptcValues, IptcTag.Credit, "credits");
+            ContainsIptcValue(iptcValues, IptcTag.Source, "source");
+            ContainsIptcValue(iptcValues, IptcTag.Name, "title");
+            ContainsIptcValue(iptcValues, IptcTag.CreatedDate, "20200414");
+            ContainsIptcValue(iptcValues, IptcTag.City, "city");
+            ContainsIptcValue(iptcValues, IptcTag.SubLocation, "sublocation");
+            ContainsIptcValue(iptcValues, IptcTag.ProvinceState, "province-state");
+            ContainsIptcValue(iptcValues, IptcTag.Country, "country");
+            ContainsIptcValue(iptcValues, IptcTag.Category, "category");
+            ContainsIptcValue(iptcValues, IptcTag.Urgency, "1");
+            ContainsIptcValue(iptcValues, IptcTag.Keywords, "keywords");
+            ContainsIptcValue(iptcValues, IptcTag.CopyrightNotice, "copyright");
+        }
+
         [Theory]
         [WithFile(TestImages.Jpeg.Baseline.App13WithEmptyIptc, PixelTypes.Rgba32)]
         public void ReadApp13_WithEmptyIptc_Works<TPixel>(TestImageProvider<TPixel> provider)

tests/ImageSharp.Tests/TestImages.cs

@@ -574,6 +574,10 @@ namespace SixLabors.ImageSharp.Tests
 
             public const string SampleMetadata = "Tiff/metadata_sample.tiff";
 
+            // Iptc data as long[] instead of byte[]
+            public const string InvalidIptcData = "Tiff/7324fcaff3aad96f27899da51c1bb5d9.tiff";
+            public const string IptcData = "Tiff/iptc.tiff";
+
             public static readonly string[] Multiframes = { MultiframeDeflateWithPreview, MultiframeLzwPredictor /*, MultiFrameDifferentSize, MultiframeDifferentSizeTiled, MultiFrameDifferentVariants,*/ };
 
             public static readonly string[] Metadata = { SampleMetadata };

tests/Images/Input/Tiff/7324fcaff3aad96f27899da51c1bb5d9.tiff

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:622d69dba0a8a67aa3b87e384a2b9ea8d29689eaa5cb5d0eee857f98ed660517
+size 15154924

tests/Images/Input/Tiff/iptc.tiff

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d7a8d89df97b35ab3be9745a345687defd427bf4ca519ad3c5a52208d98f7694
+size 15170