Fix loading bloom filter binary (#1379)

Author: reakaleek

src/Elastic.Documentation.LegacyDocs/BloomFilter.cs

@@ -141,14 +141,16 @@ public void Save(string filePath)
 	}
 
 	/// <summary>
-	/// Loads a Bloom Filter from a binary file.
+	/// Loads a Bloom Filter from a stream.
+	/// The stream is expected to contain data in the same format as Save() produces.
 	/// </summary>
-	/// <param name="filePath">The path to the file containing the filter data.</param>
+	/// <param name="stream">The stream containing the filter data.</param>
 	/// <returns>A new BloomFilter instance.</returns>
-	public static BloomFilter Load(string filePath)
+	public static BloomFilter Load(Stream stream)
 	{
-		using var stream = File.OpenRead(filePath);
-		using var reader = new BinaryReader(stream);
+		// Use a BinaryReader, but leave the stream open as it's managed externally.
+		using var reader = new BinaryReader(stream, Encoding.UTF8, leaveOpen: true);
+
 		// 1. Read metadata (Size and HashCount)
 		var size = reader.ReadInt32();
 		var hashCount = reader.ReadInt32();

src/Elastic.Documentation.LegacyDocs/LegacyPageChecker.cs

@@ -2,15 +2,17 @@
 // Elasticsearch B.V licenses this file to you under the Apache 2.0 License.
 // See the LICENSE file in the project root for more information
 
-using System.IO.Abstractions;
 using Elastic.Documentation.Configuration;
 
 namespace Elastic.Documentation.LegacyDocs;
 
-public class LegacyPageChecker(IFileSystem fs)
+public class LegacyPageChecker
 {
 	private BloomFilter? _bloomFilter;
-	private readonly string _bloomFilterBinaryPath = Path.Combine(Paths.WorkingDirectoryRoot.FullName, "src", "Elastic.Documentation.LegacyDocs", "legacy-pages.bloom.bin");
+	private const string RootNamespace = "Elastic.Documentation.LegacyDocs";
+	private const string FileName = "legacy-pages.bloom.bin";
+	private const string ResourceName = $"{RootNamespace}.{FileName}";
+	private readonly string _bloomFilterBinaryPath = Path.Combine(Paths.WorkingDirectoryRoot.FullName, "src", RootNamespace, FileName);
 
 
 	public bool PathExists(string path)
@@ -19,11 +21,13 @@ public bool PathExists(string path)
 		return _bloomFilter.Check(path);
 	}
 
-	private BloomFilter LoadBloomFilter()
+	private static BloomFilter LoadBloomFilter()
 	{
-		var bloomFilterBinaryInfo = fs.FileInfo.New(_bloomFilterBinaryPath);
-		_bloomFilter ??= BloomFilter.Load(bloomFilterBinaryInfo.FullName);
-		return _bloomFilter;
+		var assembly = typeof(LegacyPageChecker).Assembly;
+		using var stream = assembly.GetManifestResourceStream(ResourceName) ?? throw new FileNotFoundException(
+			$"Embedded resource '{ResourceName}' not found in assembly '{assembly.FullName}'. " +
+			"Ensure the Build Action for 'legacy-pages.bloom.bin' is 'Embedded Resource' and the path/name is correct.");
+		return BloomFilter.Load(stream);
 	}
 
 	public void GenerateBloomFilterBinary(IPagesProvider pagesProvider)

src/tooling/docs-assembler/Cli/LegacyDocsCommands.cs

@@ -33,8 +33,8 @@ public async Task<int> CreateBloomBin(string builtDocsDir, Cancel ctx = default)
 			NoHints = true
 		}.StartAsync(ctx);
 		var pagesProvider = new LocalPagesProvider(builtDocsDir);
-		var legacyPageLookup = new LegacyPageChecker(new FileSystem());
-		legacyPageLookup.GenerateBloomFilterBinary(pagesProvider);
+		var legacyPageChecker = new LegacyPageChecker();
+		legacyPageChecker.GenerateBloomFilterBinary(pagesProvider);
 		await collector.StopAsync(ctx);
 		return collector.Errors;
 	}
@@ -48,8 +48,8 @@ public async Task<int> PageExists(string path, Cancel ctx = default)
 		{
 			NoHints = true
 		}.StartAsync(ctx);
-		var legacyPageLookup = new LegacyPageChecker(new FileSystem());
-		var result = legacyPageLookup.PathExists(path);
+		var legacyPageChecker = new LegacyPageChecker();
+		var result = legacyPageChecker.PathExists(path);
 		Console.WriteLine(result ? "exists" : "does not exist");
 		await collector.StopAsync(ctx);
 		return collector.Errors;

src/tooling/docs-assembler/Cli/RepositoryCommands.cs

@@ -128,7 +128,7 @@ public async Task<int> BuildAll(
 
 		var pathProvider = new GlobalNavigationPathProvider(navigationFile, assembleSources, assembleContext);
 		var htmlWriter = new GlobalNavigationHtmlWriter(navigationFile, assembleContext, navigation, assembleSources);
-		var legacyPageChecker = new LegacyPageChecker(assembleContext.ReadFileSystem);
+		var legacyPageChecker = new LegacyPageChecker();
 		var historyMapper = new PageLegacyUrlMapper(legacyPageChecker, assembleSources.HistoryMappings);
 
 		var builder = new AssemblerBuilder(logger, assembleContext, navigation, htmlWriter, pathProvider, historyMapper);

tests/Elastic.Documentation.LegacyDocs.Tests/LegacyPageLookupTests.cs

@@ -12,7 +12,7 @@ public class LegacyPageCheckerTests
 	[Fact]
 	public void TestVersions()
 	{
-		var legacyPageChecker = new LegacyPageChecker(new FileSystem());
+		var legacyPageChecker = new LegacyPageChecker();
 		var expected = new Dictionary<string, bool>
 		{
 			["8.0"] = false,