Feature: Automatically detect zip encoding (#17045)

Signed-off-by: oxygen-dioxide <[email protected]>
Co-authored-by: Yair <[email protected]>

Author: oxygen-dioxide

Directory.Packages.props

@@ -38,6 +38,7 @@
     <PackageVersion Include="Microsoft.Graphics.Win2D" Version="1.3.2" />
     <PackageVersion Include="TagLibSharp" Version="2.3.0" />
     <PackageVersion Include="Tulpep.ActiveDirectoryObjectPicker" Version="3.0.11" />
+    <PackageVersion Include="UTF.Unknown" Version="2.5.1" />
     <PackageVersion Include="WinUIEx" Version="2.5.1" />
     <PackageVersion Include="Vanara.Windows.Extensions" Version="4.0.1" />
     <PackageVersion Include="Vanara.Windows.Shell" Version="4.0.1" />

src/Files.App/Actions/Content/Archives/Decompress/DecompressArchive.cs

@@ -43,6 +43,11 @@ public override async Task ExecuteAsync(object? parameter = null)
 
 			var isArchiveEncrypted = await FilesystemTasks.Wrap(() => StorageArchiveService.IsEncryptedAsync(archive.Path));
 			var isArchiveEncodingUndetermined = await FilesystemTasks.Wrap(() => StorageArchiveService.IsEncodingUndeterminedAsync(archive.Path));
+			Encoding? detectedEncoding = null;
+			if (isArchiveEncodingUndetermined)
+			{
+				detectedEncoding = await FilesystemTasks.Wrap(() => StorageArchiveService.DetectEncodingAsync(archive.Path));
+			}
 			var password = string.Empty;
 			Encoding? encoding = null;
 
@@ -51,7 +56,8 @@ public override async Task ExecuteAsync(object? parameter = null)
 			{
 				IsArchiveEncrypted = isArchiveEncrypted,
 				IsArchiveEncodingUndetermined = isArchiveEncodingUndetermined,
-				ShowPathSelection = true
+				ShowPathSelection = true,
+				DetectedEncoding = detectedEncoding,
 			};
 			decompressArchiveDialog.ViewModel = decompressArchiveViewModel;
 

src/Files.App/Data/Contracts/IStorageArchiveService.cs

@@ -59,10 +59,17 @@ public interface IStorageArchiveService
 		/// <summary>
 		/// Gets the value that indicates whether the archive file's encoding is undetermined.
 		/// </summary>
-		/// <param name="archiveFilePath">The archive file path to check if the item is encrypted.</param>
+		/// <param name="archiveFilePath">The archive file path to check if the encoding is undetermined.</param>
 		/// <returns>True if the archive file's encoding is undetermined; otherwise, false.</returns>
 		Task<bool> IsEncodingUndeterminedAsync(string archiveFilePath);
 
+		/// <summary>
+		/// Detect encoding for a zip file whose encoding is undetermined.
+		/// </summary>
+		/// <param name="archiveFilePath">The archive file path to detect encoding</param>
+		/// <returns>Null if the archive file doesn't need to detect encoding or its encoding can't be detected; otherwise, the encoding detected.</returns>
+		Task<Encoding?> DetectEncodingAsync(string archiveFilePath);
+
 		/// <summary>
 		/// Gets the <see cref="SevenZipExtractor"/> instance from the archive file path.
 		/// </summary>

src/Files.App/Data/Items/EncodingItem.cs

@@ -22,7 +22,7 @@ public sealed class EncodingItem
         /// Initializes a new instance of the <see cref="EncodingItem"/> class.
         /// </summary>
         /// <param name="code">The code of the language.</param>
-        public EncodingItem(string code)
+        public EncodingItem(string? code)
         {
             if (string.IsNullOrEmpty(code))
             {
@@ -36,6 +36,45 @@ public EncodingItem(string code)
             }
         }
 
-        public override string ToString() => Name;
+		public EncodingItem(Encoding encoding, string name)
+		{
+			Encoding = encoding;
+			Name = name;
+		}
+
+		public static EncodingItem[] Defaults = new string?[] {
+			null,//System Default
+            "UTF-8",
+			
+            //All possible Windows system encodings
+            //reference: https://en.wikipedia.org/wiki/Windows_code_page
+            //East Asian
+            "shift_jis",       //Japanese
+            "gb2312",          //Simplified Chinese
+            "big5",            //Traditional Chinese
+            "ks_c_5601-1987",  //Korean
+            
+            //Southeast Asian
+            "Windows-1258",    //Vietnamese
+            "Windows-874",     //Thai
+            
+            //Middle East
+            "Windows-1256",    //Arabic
+            "Windows-1255",    //Hebrew
+            "Windows-1254",    //Turkish
+            
+            //European
+            "Windows-1252",    //Western European
+            "Windows-1250",    //Central European
+            "Windows-1251",    //Cyrillic
+            "Windows-1253",    //Greek
+            "Windows-1257",    //Baltic
+            
+            "macintosh",
+		}
+			.Select(x => new EncodingItem(x))
+			.ToArray();
+
+		public override string ToString() => Name;
     }
 }

src/Files.App/Files.App.csproj

@@ -95,6 +95,7 @@
         <PackageReference Include="Microsoft.Graphics.Win2D" />
         <PackageReference Include="TagLibSharp" />
         <PackageReference Include="Tulpep.ActiveDirectoryObjectPicker" />
+        <PackageReference Include="UTF.Unknown" />
         <PackageReference Include="WinUIEx" />
         <PackageReference Include="Vanara.Windows.Extensions" />
         <PackageReference Include="Vanara.Windows.Shell" />

src/Files.App/Services/Storage/StorageArchiveService.cs

@@ -6,8 +6,8 @@
 using ICSharpCode.SharpZipLib.Zip;
 using SevenZip;
 using System.IO;
-using System.Linq;
 using System.Text;
+using UtfUnknown;
 using Windows.Storage;
 using Windows.Win32;
 
@@ -90,7 +90,8 @@ public async Task<bool> CompressAsync(ICompressArchiveModel compressionModel)
 		/// <inheritdoc/>
 		public Task<bool> DecompressAsync(string archiveFilePath, string destinationFolderPath, string password = "", Encoding? encoding = null)
 		{
-			if(encoding == null){
+			if (encoding == null)
+			{
 				return DecompressAsyncWithSevenZip(archiveFilePath, destinationFolderPath, password);
 			}
 			else
@@ -203,22 +204,22 @@ async Task<bool> DecompressAsyncWithSharpZipLib(string archiveFilePath, string d
 				string.IsNullOrEmpty(destinationFolderPath))
 				return false;
 			using var zipFile = new ZipFile(archiveFilePath, StringCodec.FromEncoding(encoding));
-			if(zipFile is null)
+			if (zipFile is null)
 				return false;
-			
-			if(!string.IsNullOrEmpty(password))
+
+			if (!string.IsNullOrEmpty(password))
 				zipFile.Password = password;
 
 			// Initialize a new in-progress status card
 			var statusCard = StatusCenterHelper.AddCard_Decompress(
 				archiveFilePath.CreateEnumerable(),
 				destinationFolderPath.CreateEnumerable(),
 				ReturnResult.InProgress);
-			
+
 			// Check if the decompress operation canceled
 			if (statusCard.CancellationToken.IsCancellationRequested)
 				return false;
-			
+
 			StatusCenterItemProgressModel fsProgress = new(
 				statusCard.ProgressEventSource,
 				enumerationCompleted: true,
@@ -233,51 +234,52 @@ async Task<bool> DecompressAsyncWithSharpZipLib(string archiveFilePath, string d
 			{
 				long processedBytes = 0;
 				int processedFiles = 0;
-
-				foreach (ZipEntry zipEntry in zipFile)
+				await Task.Run(async () =>
 				{
-					if (statusCard.CancellationToken.IsCancellationRequested)
+					foreach (ZipEntry zipEntry in zipFile)
 					{
-						isSuccess = false;
-						break;
-					}
-
-					if (!zipEntry.IsFile)
-					{
-						continue; // Ignore directories
-					}
+						if (statusCard.CancellationToken.IsCancellationRequested)
+						{
+							isSuccess = false;
+							break;
+						}
 
-					string entryFileName = zipEntry.Name;
-					string fullZipToPath = Path.Combine(destinationFolderPath, entryFileName);
-					string directoryName = Path.GetDirectoryName(fullZipToPath);
+						if (!zipEntry.IsFile)
+						{
+							continue; // Ignore directories
+						}
 
-					if (!Directory.Exists(directoryName))
-					{
-						Directory.CreateDirectory(directoryName);
-					}
+						string entryFileName = zipEntry.Name;
+						string fullZipToPath = Path.Combine(destinationFolderPath, entryFileName);
+						string directoryName = Path.GetDirectoryName(fullZipToPath);
 
-					byte[] buffer = new byte[4096]; // 4K is a good default
-					using (Stream zipStream = zipFile.GetInputStream(zipEntry))
-					using (FileStream streamWriter = File.Create(fullZipToPath))
-					{
-						await ThreadingService.ExecuteOnUiThreadAsync(() =>
+						if (!Directory.Exists(directoryName))
 						{
-							fsProgress.FileName = entryFileName;
-							fsProgress.Report();
-						});
+							Directory.CreateDirectory(directoryName);
+						}
 
-						StreamUtils.Copy(zipStream, streamWriter, buffer);
-					}
-					processedBytes += zipEntry.Size;
-					if (fsProgress.TotalSize > 0)
-					{
-						fsProgress.Report(processedBytes / (double)fsProgress.TotalSize * 100);
+						byte[] buffer = new byte[4096]; // 4K is a good default
+						using (Stream zipStream = zipFile.GetInputStream(zipEntry))
+						using (FileStream streamWriter = File.Create(fullZipToPath))
+						{
+							await ThreadingService.ExecuteOnUiThreadAsync(() =>
+							{
+								fsProgress.FileName = entryFileName;
+								fsProgress.Report();
+							});
+
+							StreamUtils.Copy(zipStream, streamWriter, buffer);
+						}
+						processedBytes += zipEntry.Size;
+						if (fsProgress.TotalSize > 0)
+						{
+							fsProgress.Report(processedBytes / (double)fsProgress.TotalSize * 100);
+						}
+						processedFiles++;
+						fsProgress.AddProcessedItemsCount(1);
+						fsProgress.Report();
 					}
-					processedFiles++;
-					fsProgress.AddProcessedItemsCount(1);
-					fsProgress.Report();
-				}
-
+				});
 				if (!statusCard.CancellationToken.IsCancellationRequested)
 				{
 					isSuccess = true;
@@ -321,7 +323,7 @@ await ThreadingService.ExecuteOnUiThreadAsync(() =>
 			return isSuccess;
 		}
 
-		
+
 		/// <inheritdoc/>
 		public string GenerateArchiveNameFromItems(IReadOnlyList<ListedItem> items)
 		{
@@ -355,7 +357,7 @@ public async Task<bool> IsEncodingUndeterminedAsync(string archiveFilePath)
 			{
 				using (ZipFile zipFile = new ZipFile(archiveFilePath))
 				{
-					return !zipFile.Cast<ZipEntry>().All(entry=>entry.IsUnicodeText);
+					return !zipFile.Cast<ZipEntry>().All(entry => entry.IsUnicodeText);
 				}
 			}
 			catch (Exception ex)
@@ -365,6 +367,42 @@ public async Task<bool> IsEncodingUndeterminedAsync(string archiveFilePath)
 			}
 		}
 
+		public async Task<Encoding?> DetectEncodingAsync(string archiveFilePath)
+		{
+			//Temporarily using cp437 to decode zip file
+			//because SharpZipLib requires an encoding when decoding
+			//and cp437 contains all bytes as character
+			//which means that we can store any byte array as cp437 string losslessly
+			var cp437 = Encoding.GetEncoding(437);
+			try
+			{
+				using (ZipFile zipFile = new ZipFile(archiveFilePath, StringCodec.FromEncoding(cp437)))
+				{
+					var fileNameBytes = cp437.GetBytes(
+						String.Join("\n",
+							zipFile.Cast<ZipEntry>()
+								.Where(e => !e.IsUnicodeText)
+								.Select(e => e.Name)
+						)
+					);
+					var detectionResult = CharsetDetector.DetectFromBytes(fileNameBytes);
+					if (detectionResult.Detected != null && detectionResult.Detected.Confidence > 0.5)
+					{
+						return detectionResult.Detected.Encoding;
+					}
+					else
+					{
+						return null;
+					}
+				}
+			}
+			catch (Exception ex)
+			{
+				Console.WriteLine($"SharpZipLib error: {ex.Message}");
+				return null;
+			}
+		}
+
 		/// <inheritdoc/>
 		public async Task<SevenZipExtractor?> GetSevenZipExtractorAsync(string archiveFilePath, string password = "")
 		{

src/Files.App/Strings/en-US/Resources.resw

@@ -2099,6 +2099,9 @@
   <data name="Encoding" xml:space="preserve">
     <value>Encoding</value>
   </data>
+  <data name="EncodingDetected" xml:space="preserve">
+    <value>{0} (detected)</value>
+  </data>
   <data name="ExtractToPath" xml:space="preserve">
     <value>Path</value>
   </data>

src/Files.App/ViewModels/Dialogs/DecompressArchiveDialogViewModel.cs

@@ -44,6 +44,16 @@ public bool IsArchiveEncodingUndetermined
 			set => SetProperty(ref isArchiveEncodingUndetermined, value);
 		}
 
+		private Encoding? detectedEncoding;
+		public Encoding? DetectedEncoding
+		{
+			get => detectedEncoding;
+			set { 
+				SetProperty(ref detectedEncoding, value);
+				RefreshEncodingOptions();
+			}
+		}
+
 		private bool showPathSelection;
 		public bool ShowPathSelection
 		{
@@ -53,19 +63,27 @@ public bool ShowPathSelection
 
 		public DisposableArray? Password { get; private set; }
 
-		public EncodingItem[] EncodingOptions { get; set; } = new string?[] {
-			null,//System Default
-			"UTF-8",
-			"shift_jis",
-			"gb2312",
-			"big5",
-			"ks_c_5601-1987",
-			"Windows-1252",
-			"macintosh",
-		}
-			.Select(x=>new EncodingItem(x))
-			.ToArray();
+		public EncodingItem[] EncodingOptions { get; set; } = EncodingItem.Defaults;
 		public EncodingItem SelectedEncoding { get; set; }
+		void RefreshEncodingOptions()
+		{
+			if (detectedEncoding != null)
+			{
+				EncodingOptions = EncodingItem.Defaults
+				.Prepend(new EncodingItem(
+					detectedEncoding, 
+					string.Format(Strings.EncodingDetected.GetLocalizedResource(), detectedEncoding.EncodingName)
+				))
+				.ToArray();
+			}
+			else
+			{
+				EncodingOptions = EncodingItem.Defaults;
+			}
+			SelectedEncoding = EncodingOptions.FirstOrDefault();
+		}
+
+		
 
 		public IRelayCommand PrimaryButtonClickCommand { get; private set; }