Allow comparison between detectors

Allow --compare flag to be supported for ./build report command
to seed detectors with the same subset of languages supported by all

.gitignore

@@ -5,6 +5,7 @@ riderModule.iml
 /_ReSharper.Caches/
 
 .idea
+.vs
 *.sln.DotSettings.User
 
 nuget

build/Build/Program.cs

@@ -11,18 +11,23 @@
 
 var language = new Option<string[]>(["--language"], "languages to generate an accuracy report for")
 {
-	Arity = ArgumentArity.ZeroOrMore
+	Arity = ArgumentArity.ZeroOrMore,
 };
 
 var detector = new Option<string[]>(["--implementation"], "implementations to generate an accuracy report for")
 {
 	Arity = ArgumentArity.ZeroOrMore,
 };
+detector.FromAmong("Lingua", "NTextCat", "LanguageDetection");
+
+var compare = new Option<bool>("--compare",
+	"whether implementations should use the same subset of supported languages when generating an accuracy report");
 
 var cmd = new RootCommand
 {
 	language,
 	detector,
+	compare,
 	new Argument<string[]>("targets")
 	{
 		Description =
@@ -90,7 +95,7 @@
 	{
 		var filter = new StringBuilder();
 		var languages = cmdLine.GetValueForOption(language);
-		if (languages is not null)
+		if (languages?.Length > 0)
 		{
 			foreach (var l in languages)
 			{
@@ -104,23 +109,34 @@
 		}
 
 		var detectors = cmdLine.GetValueForOption(detector);
-		if (detectors is not null)
+		if (detectors?.Length > 0)
 		{
-			foreach (var d in detectors)
+			if (filter.Length > 0)
+				filter.Append(" & ");
+
+			filter.Append('(');
+
+			for (var index = 0; index < detectors.Length; index++)
 			{
-				if (filter.Length > 0)
-					filter.Append(" & ");
+				if (index > 0)
+					filter.Append(" | ");
 
+				var d = detectors[index];
 				filter.Append("(FullyQualifiedName~.");
 				filter.Append(d);
 				filter.Append(')');
 			}
+
+			filter.Append(')');
 		}
 
-		Run("dotnet",
-			filter.Length > 0
-				? $"test tests/Lingua.AccuracyReport.Tests -c Release --no-build --filter \"{filter}\""
-				: "test tests/Lingua.AccuracyReport.Tests -c Release --no-build");
+		var additionalArgs = new StringBuilder();
+		if (cmdLine.GetValueForOption(compare))
+			additionalArgs.Append(" --environment TEST_COMPARE=\"true\"");
+		if (filter.Length > 0)
+			additionalArgs.Append($" --filter \"{filter}\" --environment TEST_FILTER=\"{filter}\"");
+
+		Run("dotnet", $"test tests/Lingua.AccuracyReport.Tests -c Release{additionalArgs}");
 
 		CombinedAccuracyReport.Create();
 	});

tests/Lingua.AccuracyReport.Tests/Comparison/ComparisonTests.cs

@@ -1,7 +1,6 @@
 using System.Runtime.InteropServices;
 using FluentAssertions;
-using Xunit;
-using static Lingua.Language;
+
 namespace Lingua.AccuracyReport.Tests.Comparison;
 
 public class ComparisonTests

tests/Lingua.AccuracyReport.Tests/GlobalUsings.cs

@@ -0,0 +1,2 @@
+global using Xunit;
+global using static Lingua.Language;

tests/Lingua.AccuracyReport.Tests/LanguageDetection/AfrikaansDetectionAccuracyReport.cs

@@ -1,23 +1,20 @@
-using Xunit;
-using static Lingua.Language;
-
 namespace Lingua.AccuracyReport.Tests.LanguageDetection;
 
 public class AfrikaansDetectionAccuracyReport(LanguageDetectionStatistics<LanguageDetectionLanguageDetectorFactory> statistics)
 	: LanguageDetectionDetectionAccuracyReport(Afrikaans, statistics),
 		IClassFixture<LanguageDetectionStatistics<LanguageDetectionLanguageDetectorFactory>>
 {
-	[Theory(DisplayName = "single word detection")]
+	[SingleWordReportTheory(Implementation.LanguageDetection, Afrikaans)]
 	[SingleWordData(Afrikaans)]
 	public override void SingleWordsAreIdentifiedCorrectly(string singleWord) =>
 		ComputeSingleWordStatistics(singleWord);
 
-	[Theory(DisplayName = "word pair detection")]
+	[WordPairsReportTheory(Implementation.LanguageDetection, Afrikaans)]
 	[WordPairsData(Afrikaans)]
 	public override void WordPairsAreIdentifiedCorrectly(string wordPair) =>
 		ComputeWordPairStatistics(wordPair);
 
-	[Theory(DisplayName = "sentence detection")]
+	[SentenceReportTheory(Implementation.LanguageDetection, Afrikaans)]
 	[SentenceData(Afrikaans)]
 	public override void EntireSentencesAreIdentifiedCorrectly(string sentence) =>
 		ComputeSentenceStatistics(sentence);

tests/Lingua.AccuracyReport.Tests/LanguageDetection/AlbanianDetectionAccuracyReport.cs

@@ -1,23 +1,20 @@
-using Xunit;
-using static Lingua.Language;
-
 namespace Lingua.AccuracyReport.Tests.LanguageDetection;
 
 public class AlbanianDetectionAccuracyReport(LanguageDetectionStatistics<LanguageDetectionLanguageDetectorFactory> statistics)
 	: LanguageDetectionDetectionAccuracyReport(Albanian, statistics),
 		IClassFixture<LanguageDetectionStatistics<LanguageDetectionLanguageDetectorFactory>>
 {
-	[Theory(DisplayName = "single word detection")]
+	[SingleWordReportTheory(Implementation.LanguageDetection, Albanian)]
 	[SingleWordData(Albanian)]
 	public override void SingleWordsAreIdentifiedCorrectly(string singleWord) =>
 		ComputeSingleWordStatistics(singleWord);
 
-	[Theory(DisplayName = "word pair detection")]
+	[WordPairsReportTheory(Implementation.LanguageDetection, Albanian)]
 	[WordPairsData(Albanian)]
 	public override void WordPairsAreIdentifiedCorrectly(string wordPair) =>
 		ComputeWordPairStatistics(wordPair);
 
-	[Theory(DisplayName = "sentence detection")]
+	[SentenceReportTheory(Implementation.LanguageDetection, Albanian)]
 	[SentenceData(Albanian)]
 	public override void EntireSentencesAreIdentifiedCorrectly(string sentence) =>
 		ComputeSentenceStatistics(sentence);

tests/Lingua.AccuracyReport.Tests/LanguageDetection/ArabicDetectionAccuracyReport.cs

@@ -1,23 +1,20 @@
-using Xunit;
-using static Lingua.Language;
-
 namespace Lingua.AccuracyReport.Tests.LanguageDetection;
 
 public class ArabicDetectionAccuracyReport(LanguageDetectionStatistics<LanguageDetectionLanguageDetectorFactory> statistics)
 	: LanguageDetectionDetectionAccuracyReport(Arabic, statistics),
 		IClassFixture<LanguageDetectionStatistics<LanguageDetectionLanguageDetectorFactory>>
 {
-	[Theory(DisplayName = "single word detection")]
+	[SingleWordReportTheory(Implementation.LanguageDetection, Arabic)]
 	[SingleWordData(Arabic)]
 	public override void SingleWordsAreIdentifiedCorrectly(string singleWord) =>
 		ComputeSingleWordStatistics(singleWord);
 
-	[Theory(DisplayName = "word pair detection")]
+	[WordPairsReportTheory(Implementation.LanguageDetection, Arabic)]
 	[WordPairsData(Arabic)]
 	public override void WordPairsAreIdentifiedCorrectly(string wordPair) =>
 		ComputeWordPairStatistics(wordPair);
 
-	[Theory(DisplayName = "sentence detection")]
+	[SentenceReportTheory(Implementation.LanguageDetection, Arabic)]
 	[SentenceData(Arabic)]
 	public override void EntireSentencesAreIdentifiedCorrectly(string sentence) =>
 		ComputeSentenceStatistics(sentence);

tests/Lingua.AccuracyReport.Tests/LanguageDetection/BengaliDetectionAccuracyReport.cs

@@ -1,23 +1,20 @@
-using Xunit;
-using static Lingua.Language;
-
 namespace Lingua.AccuracyReport.Tests.LanguageDetection;
 
 public class BengaliDetectionAccuracyReport(LanguageDetectionStatistics<LanguageDetectionLanguageDetectorFactory> statistics)
 	: LanguageDetectionDetectionAccuracyReport(Bengali, statistics),
 		IClassFixture<LanguageDetectionStatistics<LanguageDetectionLanguageDetectorFactory>>
 {
-	[Theory(DisplayName = "single word detection")]
+	[SingleWordReportTheory(Implementation.LanguageDetection, Bengali)]
 	[SingleWordData(Bengali)]
 	public override void SingleWordsAreIdentifiedCorrectly(string singleWord) =>
 		ComputeSingleWordStatistics(singleWord);
 
-	[Theory(DisplayName = "word pair detection")]
+	[WordPairsReportTheory(Implementation.LanguageDetection, Bengali)]
 	[WordPairsData(Bengali)]
 	public override void WordPairsAreIdentifiedCorrectly(string wordPair) =>
 		ComputeWordPairStatistics(wordPair);
 
-	[Theory(DisplayName = "sentence detection")]
+	[SentenceReportTheory(Implementation.LanguageDetection, Bengali)]
 	[SentenceData(Bengali)]
 	public override void EntireSentencesAreIdentifiedCorrectly(string sentence) =>
 		ComputeSentenceStatistics(sentence);

tests/Lingua.AccuracyReport.Tests/LanguageDetection/BokmalDetectionAccuracyReport.cs

@@ -1,23 +1,20 @@
-using Xunit;
-using static Lingua.Language;
-
 namespace Lingua.AccuracyReport.Tests.LanguageDetection;
 
 public class BokmalDetectionAccuracyReport(LanguageDetectionStatistics<LanguageDetectionLanguageDetectorFactory> statistics)
 	: LanguageDetectionDetectionAccuracyReport(Bokmal, statistics),
 		IClassFixture<LanguageDetectionStatistics<LanguageDetectionLanguageDetectorFactory>>
 {
-	[Theory(DisplayName = "single word detection")]
+	[SingleWordReportTheory(Implementation.LanguageDetection, Bokmal)]
 	[SingleWordData(Bokmal)]
 	public override void SingleWordsAreIdentifiedCorrectly(string singleWord) =>
 		ComputeSingleWordStatistics(singleWord);
 
-	[Theory(DisplayName = "word pair detection")]
+	[WordPairsReportTheory(Implementation.LanguageDetection, Bokmal)]
 	[WordPairsData(Bokmal)]
 	public override void WordPairsAreIdentifiedCorrectly(string wordPair) =>
 		ComputeWordPairStatistics(wordPair);
 
-	[Theory(DisplayName = "sentence detection")]
+	[SentenceReportTheory(Implementation.LanguageDetection, Bokmal)]
 	[SentenceData(Bokmal)]
 	public override void EntireSentencesAreIdentifiedCorrectly(string sentence) =>
 		ComputeSentenceStatistics(sentence);