[refactor] Separate noise filtering and centralize noise file detection

docs/specs/SPEC-noise-file-detection.md

@@ -0,0 +1,33 @@
+# Noise File Detection Specification
+
+## Purpose
+
+This document lists the known noise file names filtered by the client inventory pipeline and their platform origin.
+
+The runtime source of truth is the embedded JSON resource:
+`src/ByteSync.Client/Services/Inventories/noise-files.json`.
+
+## Known noise file names
+
+| File name | Origin platform | Typical purpose |
+| --- | --- | --- |
+| `desktop.ini` | Windows | Folder customization metadata |
+| `thumbs.db` | Windows | Thumbnail cache |
+| `ehthumbs.db` | Windows | Media Center thumbnail cache |
+| `ehthumbs_vista.db` | Windows | Vista Media Center thumbnail cache |
+| `.desktop.ini` | Windows/Linux legacy compatibility | Legacy hidden variant |
+| `.thumbs.db` | Windows/Linux legacy compatibility | Legacy hidden variant |
+| `.DS_Store` | macOS | Finder metadata |
+| `.AppleDouble` | macOS | Resource fork metadata |
+| `.LSOverride` | macOS | Launch Services overrides |
+| `.Spotlight-V100` | macOS | Spotlight indexing data |
+| `.Trashes` | macOS | Trash metadata or folder marker |
+| `.fseventsd` | macOS | File system event metadata |
+| `.TemporaryItems` | macOS | Temporary items marker |
+| `.VolumeIcon.icns` | macOS | Custom volume icon |
+| `.directory` | Linux (KDE) | Directory display metadata |
+
+## Matching behavior
+
+- On Linux, matching is case-sensitive.
+- On non-Linux platforms, matching is case-insensitive.

src/ByteSync.Client/ByteSync.Client.csproj

@@ -111,6 +111,7 @@
         </EmbeddedResource>
         <None Remove="local.settings.json"/>
         <EmbeddedResource Include="local.settings.json"/>
+        <EmbeddedResource Include="Services\Inventories\noise-files.json"/>
     </ItemGroup>
     <ItemGroup>
         <UpToDateCheckInput Remove="Assets\Resources\Resources.en.resx"/>
@@ -258,4 +259,4 @@
             <SubType>Code</SubType>
         </Compile>
     </ItemGroup>
-</Project>
+</Project>

src/ByteSync.Client/Services/Inventories/FileSystemInspector.cs

@@ -63,13 +63,7 @@ public bool IsSystemAttribute(FileInfo fileInfo)
 
     public bool IsNoiseFileName(FileInfo fileInfo, OSPlatforms os)
     {
-        var comparison = os == OSPlatforms.Linux ? StringComparison.Ordinal : StringComparison.OrdinalIgnoreCase;
-
-        return fileInfo.Name.Equals("desktop.ini", comparison)
-               || fileInfo.Name.Equals("thumbs.db", comparison)
-               || fileInfo.Name.Equals(".desktop.ini", comparison)
-               || fileInfo.Name.Equals(".thumbs.db", comparison)
-               || fileInfo.Name.Equals(".DS_Store", comparison);
+        return NoiseFileDetector.IsNoiseFileName(fileInfo.Name, os);
     }
 
     public bool IsReparsePoint(FileSystemInfo fsi)
@@ -120,4 +114,4 @@ private bool SafeIsReparsePoint(FileSystemInfo fsi)
             return false;
         }
     }
-}
+}

src/ByteSync.Client/Services/Inventories/NoiseFileDetector.cs

@@ -0,0 +1,44 @@
+using System.Reflection;
+using System.Text.Json;
+using ByteSync.Common.Business.Misc;
+
+namespace ByteSync.Services.Inventories;
+
+public static class NoiseFileDetector
+{
+    private const string NoiseFileResourceSuffix = ".Services.Inventories.noise-files.json";
+    private static readonly string[] KnownNoiseFileNames = LoadNoiseFileNames();
+
+    private static readonly HashSet<string> CaseSensitiveNoiseFileNames = new(KnownNoiseFileNames, StringComparer.Ordinal);
+    private static readonly HashSet<string> CaseInsensitiveNoiseFileNames = new(KnownNoiseFileNames, StringComparer.OrdinalIgnoreCase);
+
+    public static bool IsNoiseFileName(string? fileName, OSPlatforms os)
+    {
+        if (string.IsNullOrWhiteSpace(fileName))
+        {
+            return false;
+        }
+
+        return os == OSPlatforms.Linux
+            ? CaseSensitiveNoiseFileNames.Contains(fileName)
+            : CaseInsensitiveNoiseFileNames.Contains(fileName);
+    }
+
+    private static string[] LoadNoiseFileNames()
+    {
+        var assembly = typeof(NoiseFileDetector).Assembly;
+        var resourceName = assembly.GetManifestResourceNames()
+            .Single(rn => rn.EndsWith(NoiseFileResourceSuffix, StringComparison.Ordinal));
+
+        using var stream = assembly.GetManifestResourceStream(resourceName);
+        ArgumentNullException.ThrowIfNull(stream);
+
+        var parsed = JsonSerializer.Deserialize<string[]>(stream);
+        ArgumentNullException.ThrowIfNull(parsed);
+
+        return parsed
+            .Where(s => !string.IsNullOrWhiteSpace(s))
+            .Distinct(StringComparer.Ordinal)
+            .ToArray();
+    }
+}

src/ByteSync.Client/Services/Inventories/noise-files.json

@@ -0,0 +1,17 @@
+[
+  "desktop.ini",
+  "thumbs.db",
+  "ehthumbs.db",
+  "ehthumbs_vista.db",
+  ".desktop.ini",
+  ".thumbs.db",
+  ".DS_Store",
+  ".AppleDouble",
+  ".LSOverride",
+  ".Spotlight-V100",
+  ".Trashes",
+  ".fseventsd",
+  ".TemporaryItems",
+  ".VolumeIcon.icns",
+  ".directory"
+]

tests/ByteSync.Client.UnitTests/Services/Inventories/FileSystemInspectorTests.cs

@@ -1,4 +1,5 @@
 using ByteSync.Business.Inventories;
+using ByteSync.Common.Business.Misc;
 using ByteSync.Interfaces.Controls.Inventories;
 using ByteSync.Services.Inventories;
 using FluentAssertions;
@@ -131,4 +132,46 @@ public void ClassifyEntry_FallsBackToRegularFile_WhenPosixClassifierThrows()
             Directory.Delete(tempDirectory.FullName, true);
         }
     }
-}
+
+    [Test]
+    public void IsNoiseFileName_ShouldReturnTrue_ForKnownNoiseFile()
+    {
+        var inspector = new FileSystemInspector();
+        var tempDirectory = Directory.CreateDirectory(Path.Combine(Path.GetTempPath(), Guid.NewGuid().ToString("N")));
+        var filePath = Path.Combine(tempDirectory.FullName, "thumbs.db");
+        File.WriteAllText(filePath, "x");
+        var fileInfo = new FileInfo(filePath);
+
+        try
+        {
+            var result = inspector.IsNoiseFileName(fileInfo, OSPlatforms.Windows);
+
+            result.Should().BeTrue();
+        }
+        finally
+        {
+            Directory.Delete(tempDirectory.FullName, true);
+        }
+    }
+
+    [Test]
+    public void IsNoiseFileName_ShouldReturnFalse_ForUnknownFile()
+    {
+        var inspector = new FileSystemInspector();
+        var tempDirectory = Directory.CreateDirectory(Path.Combine(Path.GetTempPath(), Guid.NewGuid().ToString("N")));
+        var filePath = Path.Combine(tempDirectory.FullName, "regular.txt");
+        File.WriteAllText(filePath, "x");
+        var fileInfo = new FileInfo(filePath);
+
+        try
+        {
+            var result = inspector.IsNoiseFileName(fileInfo, OSPlatforms.Windows);
+
+            result.Should().BeFalse();
+        }
+        finally
+        {
+            Directory.Delete(tempDirectory.FullName, true);
+        }
+    }
+}

tests/ByteSync.Client.UnitTests/Services/Inventories/NoiseFileDetectorTests.cs

@@ -0,0 +1,111 @@
+using System.Text.Json;
+using ByteSync.Common.Business.Misc;
+using ByteSync.Services.Inventories;
+using FluentAssertions;
+using NUnit.Framework;
+
+namespace ByteSync.Client.UnitTests.Services.Inventories;
+
+public class NoiseFileDetectorTests
+{
+    private static readonly string[] KnownNoiseFileNames = LoadNoiseFileNamesFromEmbeddedResource();
+
+    [TestCaseSource(nameof(KnownNoiseFileNames))]
+    public void IsNoiseFileName_ShouldReturnTrue_ForKnownNoiseFiles_OnWindows(string fileName)
+    {
+        var result = NoiseFileDetector.IsNoiseFileName(fileName, OSPlatforms.Windows);
+
+        result.Should().BeTrue();
+    }
+
+    [TestCaseSource(nameof(KnownNoiseFileNames))]
+    public void IsNoiseFileName_ShouldReturnTrue_ForKnownNoiseFiles_OnLinux(string fileName)
+    {
+        var result = NoiseFileDetector.IsNoiseFileName(fileName, OSPlatforms.Linux);
+
+        result.Should().BeTrue();
+    }
+
+    [TestCase("DESKTOP.INI")]
+    [TestCase("THUMBS.DB")]
+    [TestCase("EHTHUMBS.DB")]
+    [TestCase("EHTHUMBS_VISTA.DB")]
+    [TestCase(".ds_store")]
+    [TestCase(".appledouble")]
+    [TestCase(".lsoverride")]
+    [TestCase(".spotlight-v100")]
+    [TestCase(".trashes")]
+    [TestCase(".FSEVENTSD")]
+    [TestCase(".temporaryitems")]
+    [TestCase(".volumeicon.icns")]
+    [TestCase(".DIRECTORY")]
+    public void IsNoiseFileName_ShouldBeCaseInsensitive_OnNonLinuxPlatforms(string fileName)
+    {
+        var windowsResult = NoiseFileDetector.IsNoiseFileName(fileName, OSPlatforms.Windows);
+        var macResult = NoiseFileDetector.IsNoiseFileName(fileName, OSPlatforms.MacOs);
+
+        windowsResult.Should().BeTrue();
+        macResult.Should().BeTrue();
+    }
+
+    [TestCase("DESKTOP.INI")]
+    [TestCase("THUMBS.DB")]
+    [TestCase("EHTHUMBS.DB")]
+    [TestCase("EHTHUMBS_VISTA.DB")]
+    [TestCase(".ds_store")]
+    [TestCase(".appledouble")]
+    [TestCase(".lsoverride")]
+    [TestCase(".spotlight-v100")]
+    [TestCase(".trashes")]
+    [TestCase(".FSEVENTSD")]
+    [TestCase(".temporaryitems")]
+    [TestCase(".volumeicon.icns")]
+    [TestCase(".DIRECTORY")]
+    public void IsNoiseFileName_ShouldBeCaseSensitive_OnLinux(string fileName)
+    {
+        var result = NoiseFileDetector.IsNoiseFileName(fileName, OSPlatforms.Linux);
+
+        result.Should().BeFalse();
+    }
+
+    [TestCase("readme.md")]
+    [TestCase("normal.txt")]
+    [TestCase(".gitignore")]
+    public void IsNoiseFileName_ShouldReturnFalse_ForUnknownFileNames(string fileName)
+    {
+        var windowsResult = NoiseFileDetector.IsNoiseFileName(fileName, OSPlatforms.Windows);
+        var linuxResult = NoiseFileDetector.IsNoiseFileName(fileName, OSPlatforms.Linux);
+
+        windowsResult.Should().BeFalse();
+        linuxResult.Should().BeFalse();
+    }
+
+    [TestCase(null)]
+    [TestCase("")]
+    [TestCase("   ")]
+    public void IsNoiseFileName_ShouldReturnFalse_ForEmptyValues(string? fileName)
+    {
+        var windowsResult = NoiseFileDetector.IsNoiseFileName(fileName, OSPlatforms.Windows);
+        var linuxResult = NoiseFileDetector.IsNoiseFileName(fileName, OSPlatforms.Linux);
+
+        windowsResult.Should().BeFalse();
+        linuxResult.Should().BeFalse();
+    }
+
+    private static string[] LoadNoiseFileNamesFromEmbeddedResource()
+    {
+        var assembly = typeof(NoiseFileDetector).Assembly;
+        var resourceName = assembly.GetManifestResourceNames()
+            .SingleOrDefault(rn => rn.EndsWith(".Services.Inventories.noise-files.json", StringComparison.Ordinal));
+
+        resourceName.Should().NotBeNull();
+
+        using var stream = assembly.GetManifestResourceStream(resourceName!);
+        stream.Should().NotBeNull();
+
+        var data = JsonSerializer.Deserialize<string[]>(stream!);
+        data.Should().NotBeNull();
+
+        return data!;
+    }
+}