WIP: Implement compressed tar detection in TarFactory (seeking clarification)

Enhanced TarFactory.IsArchive() to detect compressed tar formats (tar.bz2, tar.lz, etc.) by:
- Checking if stream contains tar header directly (for uncompressed)
- Testing each compression format to see if it contains a tar file

Enhanced TarFactory.Open() methods to decompress compressed tar files to MemoryStream for seekable access.

Current blockers:
- Some compression stream constructors (LZipStream) don't support leaveOpen parameter
- Stream disposal during detection phase causes issues with factory detection flow

Awaiting clarification from @adamhathcock on preferred architecture approach.

Co-authored-by: adamhathcock <527620+adamhathcock@users.noreply.github.com>

src/SharpCompress/Archives/ArchiveFactory.cs

@@ -3,6 +3,10 @@ using System.Collections.Generic;
 using System.IO;
 using System.Linq;
 using SharpCompress.Common;
+using SharpCompress.Common.Tar.Headers;
+using SharpCompress.Compressors;
+using SharpCompress.Compressors.BZip2;
+using SharpCompress.Compressors.LZMA;
 using SharpCompress.Factories;
 using SharpCompress.IO;
 using SharpCompress.Readers;
@@ -131,10 +135,10 @@ public static class ArchiveFactory
     {
         finfo.NotNull(nameof(finfo));
         using Stream stream = finfo.OpenRead();
-        return FindFactory<T>(stream);
+        return FindFactory<T>(stream, finfo.Name);
     }
 
-    private static T FindFactory<T>(Stream stream)
+    private static T FindFactory<T>(Stream stream, string? fileName = null)
         where T : IFactory
     {
         stream.NotNull(nameof(stream));
@@ -159,6 +163,16 @@ public static class ArchiveFactory
             }
         }
 
+        stream.Seek(startPosition, SeekOrigin.Begin);
+
+        // Check if this is a compressed tar file (tar.bz2, tar.lz, etc.)
+        // These formats are supported by ReaderFactory but not by ArchiveFactory
+        var compressedTarMessage = TryGetCompressedTarMessage(stream, fileName);
+        if (compressedTarMessage != null)
+        {
+            throw new InvalidOperationException(compressedTarMessage);
+        }
+
         var extensions = string.Join(", ", factories.Select(item => item.Name));
 
         throw new InvalidOperationException(
@@ -248,4 +262,111 @@ public static class ArchiveFactory
     }
 
     public static IArchiveFactory AutoFactory { get; } = new AutoArchiveFactory();
+
+    /// <summary>
+    /// Checks if the stream is a compressed tar file (tar.bz2, tar.lz, etc.) that should use ReaderFactory instead.
+    /// Returns an error message if detected, null otherwise.
+    /// </summary>
+    private static string? TryGetCompressedTarMessage(Stream stream, string? fileName)
+    {
+        var startPosition = stream.Position;
+        try
+        {
+            // Check if it's a BZip2 file
+            if (BZip2Stream.IsBZip2(stream))
+            {
+                stream.Seek(startPosition, SeekOrigin.Begin);
+
+                // Try to decompress and check if it contains a tar archive
+                using var decompressed = new BZip2Stream(stream, CompressionMode.Decompress, true);
+                if (IsTarStream(decompressed))
+                {
+                    return "This appears to be a tar.bz2 archive. The Archive API requires seekable streams, but decompression streams are not seekable. "
+                        + "Please use ReaderFactory.Open() instead for forward-only extraction, "
+                        + "or decompress the file first and then open the resulting tar file with ArchiveFactory.Open().";
+                }
+                return null;
+            }
+
+            stream.Seek(startPosition, SeekOrigin.Begin);
+
+            // Check if it's an LZip file
+            if (LZipStream.IsLZipFile(stream))
+            {
+                stream.Seek(startPosition, SeekOrigin.Begin);
+
+                // Try to decompress and check if it contains a tar archive
+                using var decompressed = new LZipStream(stream, CompressionMode.Decompress);
+                if (IsTarStream(decompressed))
+                {
+                    return "This appears to be a tar.lz archive. The Archive API requires seekable streams, but decompression streams are not seekable. "
+                        + "Please use ReaderFactory.Open() instead for forward-only extraction, "
+                        + "or decompress the file first and then open the resulting tar file with ArchiveFactory.Open().";
+                }
+                return null;
+            }
+
+            // Check file extension as a fallback for other compressed tar formats
+            if (fileName != null)
+            {
+                var lowerFileName = fileName.ToLowerInvariant();
+                if (
+                    lowerFileName.EndsWith(".tar.bz2")
+                    || lowerFileName.EndsWith(".tbz")
+                    || lowerFileName.EndsWith(".tbz2")
+                    || lowerFileName.EndsWith(".tb2")
+                    || lowerFileName.EndsWith(".tz2")
+                    || lowerFileName.EndsWith(".tar.lz")
+                    || lowerFileName.EndsWith(".tar.xz")
+                    || lowerFileName.EndsWith(".txz")
+                    || lowerFileName.EndsWith(".tar.zst")
+                    || lowerFileName.EndsWith(".tar.zstd")
+                    || lowerFileName.EndsWith(".tzst")
+                    || lowerFileName.EndsWith(".tzstd")
+                    || lowerFileName.EndsWith(".tar.z")
+                    || lowerFileName.EndsWith(".tz")
+                    || lowerFileName.EndsWith(".taz")
+                )
+                {
+                    return $"The file '{fileName}' appears to be a compressed tar archive. The Archive API requires seekable streams, but decompression streams are not seekable. "
+                        + "Please use ReaderFactory.Open() instead for forward-only extraction, "
+                        + "or decompress the file first and then open the resulting tar file with ArchiveFactory.Open().";
+                }
+            }
+
+            return null;
+        }
+        catch
+        {
+            // If we can't determine, just return null and let the normal error handling proceed
+            return null;
+        }
+        finally
+        {
+            try
+            {
+                stream.Seek(startPosition, SeekOrigin.Begin);
+            }
+            catch
+            {
+                // Ignore seek failures
+            }
+        }
+    }
+
+    /// <summary>
+    /// Checks if a stream contains a tar archive by trying to read a tar header.
+    /// </summary>
+    private static bool IsTarStream(Stream stream)
+    {
+        try
+        {
+            var tarHeader = new TarHeader(new ArchiveEncoding());
+            return tarHeader.Read(new BinaryReader(stream));
+        }
+        catch
+        {
+            return false;
+        }
+    }
 }

src/SharpCompress/Factories/TarFactory.cs

@@ -57,19 +57,238 @@ public class TarFactory
         Stream stream,
         string? password = null,
         int bufferSize = ReaderOptions.DefaultBufferSize
-    ) => TarArchive.IsTarFile(stream);
+    )
+    {
+        if (!stream.CanSeek)
+        {
+            return TarArchive.IsTarFile(stream); // For non-seekable streams, just check if it's a tar file
+        }
+
+        var startPosition = stream.Position;
+
+        // First check if it's a regular tar file
+        if (TarArchive.IsTarFile(stream))
+        {
+            stream.Seek(startPosition, SeekOrigin.Begin); // Seek back for consistency
+            return true;
+        }
+
+        // Seek back after the tar file check
+        stream.Seek(startPosition, SeekOrigin.Begin);
+
+        if (compressionOptions == null)
+        {
+            return false;
+        }
+
+        try
+        {
+            // Try each compression option to see if it contains a tar file
+            foreach (var testOption in compressionOptions)
+            {
+                if (testOption.Type == CompressionType.None)
+                {
+                    continue; // Skip uncompressed
+                }
+
+                stream.Seek(startPosition, SeekOrigin.Begin);
+
+                try
+                {
+                    if (testOption.CanHandle(stream))
+                    {
+                        stream.Seek(startPosition, SeekOrigin.Begin);
+
+                        // Try to decompress and check if it contains a tar archive
+                        // For compression formats that don't support leaveOpen, we need to save/restore position
+                        var positionBeforeDecompress = stream.Position;
+                        Stream? decompressedStream = null;
+                        bool streamWasClosed = false;
+                        
+                        try
+                        {
+                            decompressedStream = testOption.Type switch
+                            {
+                                CompressionType.BZip2 => new BZip2Stream(stream, CompressionMode.Decompress, true),
+                                _ => testOption.CreateStream(stream) // For other types, may close the stream
+                            };
+
+                            if (TarArchive.IsTarFile(decompressedStream))
+                            {
+                                return true;
+                            }
+                        }
+                        catch (ObjectDisposedException)
+                        {
+                            streamWasClosed = true;
+                            throw; // Stream was closed, can't continue
+                        }
+                        finally
+                        {
+                            decompressedStream?.Dispose();
+                            
+                            if (!streamWasClosed && stream.CanSeek)
+                            {
+                                try
+                                {
+                                    stream.Seek(positionBeforeDecompress, SeekOrigin.Begin);
+                                }
+                                catch
+                                {
+                                    // If seek fails, the stream might have been closed
+                                }
+                            }
+                        }
+
+                        // Seek back to start after decompression attempt
+                        stream.Seek(startPosition, SeekOrigin.Begin);
+                    }
+                }
+                catch
+                {
+                    // If decompression fails, it's not this format - continue to next option
+                    try
+                    {
+                        stream.Seek(startPosition, SeekOrigin.Begin);
+                    }
+                    catch
+                    {
+                        // Ignore seek failures
+                    }
+                }
+            }
+
+            return false;
+        }
+        finally
+        {
+            try
+            {
+                stream.Seek(startPosition, SeekOrigin.Begin);
+            }
+            catch
+            {
+                // Ignore seek failures
+            }
+        }
+    }
 
     #endregion
 
     #region IArchiveFactory
 
     /// <inheritdoc/>
-    public IArchive Open(Stream stream, ReaderOptions? readerOptions = null) =>
-        TarArchive.Open(stream, readerOptions);
+    public IArchive Open(Stream stream, ReaderOptions? readerOptions = null)
+    {
+        readerOptions ??= new ReaderOptions();
+
+        // Try to detect and handle compressed tar formats
+        if (stream.CanSeek)
+        {
+            var startPosition = stream.Position;
+
+            // Try each compression option to see if we can decompress it
+            foreach (var testOption in compressionOptions)
+            {
+                if (testOption.Type == CompressionType.None)
+                {
+                    continue; // Skip uncompressed
+                }
+
+                stream.Seek(startPosition, SeekOrigin.Begin);
+
+                if (testOption.CanHandle(stream))
+                {
+                    stream.Seek(startPosition, SeekOrigin.Begin);
+
+                    // Decompress the entire stream into a seekable MemoryStream
+                    using var decompressedStream = testOption.CreateStream(stream);
+                    var memoryStream = new MemoryStream();
+                    decompressedStream.CopyTo(memoryStream);
+                    memoryStream.Position = 0;
+
+                    // Verify it's actually a tar file
+                    if (TarArchive.IsTarFile(memoryStream))
+                    {
+                        memoryStream.Position = 0;
+                        // Return a TarArchive from the decompressed memory stream
+                        // The TarArchive will own the MemoryStream and dispose it when disposed
+                        var options = new ReaderOptions
+                        {
+                            LeaveStreamOpen = false, // Ensure the MemoryStream is disposed with the archive
+                            ArchiveEncoding = readerOptions?.ArchiveEncoding ?? new ArchiveEncoding()
+                        };
+                        return TarArchive.Open(memoryStream, options);
+                    }
+
+                    memoryStream.Dispose();
+                }
+            }
+
+            stream.Seek(startPosition, SeekOrigin.Begin);
+        }
+
+        // Fall back to normal tar archive opening
+        return TarArchive.Open(stream, readerOptions);
+    }
 
     /// <inheritdoc/>
-    public IArchive Open(FileInfo fileInfo, ReaderOptions? readerOptions = null) =>
-        TarArchive.Open(fileInfo, readerOptions);
+    public IArchive Open(FileInfo fileInfo, ReaderOptions? readerOptions = null)
+    {
+        readerOptions ??= new ReaderOptions();
+
+        // Try to detect and handle compressed tar formats by file extension and content
+        using var fileStream = fileInfo.OpenRead();
+
+        // Try each compression option
+        foreach (var testOption in compressionOptions)
+        {
+            if (testOption.Type == CompressionType.None)
+            {
+                continue; // Skip uncompressed
+            }
+
+            // Check if file extension matches
+            var fileName = fileInfo.Name.ToLowerInvariant();
+            if (testOption.KnownExtensions.Any(ext => fileName.EndsWith(ext)))
+            {
+                fileStream.Position = 0;
+
+                // Verify it's the right compression format
+                if (testOption.CanHandle(fileStream))
+                {
+                    fileStream.Position = 0;
+
+                    // Decompress the entire file into a seekable MemoryStream
+                    using var decompressedStream = testOption.CreateStream(fileStream);
+                    var memoryStream = new MemoryStream();
+                    decompressedStream.CopyTo(memoryStream);
+                    memoryStream.Position = 0;
+
+                    // Verify it's actually a tar file
+                    if (TarArchive.IsTarFile(memoryStream))
+                    {
+                        memoryStream.Position = 0;
+                        // Return a TarArchive from the decompressed memory stream
+                        // The TarArchive will own the MemoryStream and dispose it when disposed
+                        var options = new ReaderOptions
+                        {
+                            LeaveStreamOpen = false, // Ensure the MemoryStream is disposed with the archive
+                            ArchiveEncoding = readerOptions?.ArchiveEncoding ?? new ArchiveEncoding()
+                        };
+                        return TarArchive.Open(memoryStream, options);
+                    }
+
+                    memoryStream.Dispose();
+                }
+            }
+        }
+
+        // fileStream will be closed by the using statement
+
+        // Fall back to normal tar archive opening
+        return TarArchive.Open(fileInfo, readerOptions);
+    }
 
     #endregion
 

tests/SharpCompress.Test/ArchiveFactoryCompressedTarTests.cs

@@ -0,0 +1,63 @@
+using System;
+using System.IO;
+using SharpCompress.Archives;
+using Xunit;
+
+namespace SharpCompress.Test;
+
+public class ArchiveFactoryCompressedTarTests : TestBase
+{
+    [Fact]
+    public void ArchiveFactory_Open_TarBz2_ThrowsHelpfulException()
+    {
+        var testFile = Path.Combine(TEST_ARCHIVES_PATH, "Tar.tar.bz2");
+        var exception = Assert.Throws<InvalidOperationException>(() =>
+        {
+            using var archive = ArchiveFactory.Open(testFile);
+        });
+
+        Assert.Contains("tar.bz2", exception.Message);
+        Assert.Contains("ReaderFactory", exception.Message);
+    }
+
+    [Fact]
+    public void ArchiveFactory_Open_TarLz_ThrowsHelpfulException()
+    {
+        var testFile = Path.Combine(TEST_ARCHIVES_PATH, "Tar.tar.lz");
+        var exception = Assert.Throws<InvalidOperationException>(() =>
+        {
+            using var archive = ArchiveFactory.Open(testFile);
+        });
+
+        Assert.Contains("tar.lz", exception.Message);
+        Assert.Contains("ReaderFactory", exception.Message);
+    }
+
+    [Fact]
+    public void ArchiveFactory_Open_TarBz2Stream_ThrowsHelpfulException()
+    {
+        var testFile = Path.Combine(TEST_ARCHIVES_PATH, "Tar.tar.bz2");
+        using var stream = File.OpenRead(testFile);
+        var exception = Assert.Throws<InvalidOperationException>(() =>
+        {
+            using var archive = ArchiveFactory.Open(stream);
+        });
+
+        Assert.Contains("tar.bz2", exception.Message);
+        Assert.Contains("ReaderFactory", exception.Message);
+    }
+
+    [Fact]
+    public void ArchiveFactory_Open_TarLzStream_ThrowsHelpfulException()
+    {
+        var testFile = Path.Combine(TEST_ARCHIVES_PATH, "Tar.tar.lz");
+        using var stream = File.OpenRead(testFile);
+        var exception = Assert.Throws<InvalidOperationException>(() =>
+        {
+            using var archive = ArchiveFactory.Open(stream);
+        });
+
+        Assert.Contains("tar.lz", exception.Message);
+        Assert.Contains("ReaderFactory", exception.Message);
+    }
+}