Hopefully fixed an issue that only appears in CD tests, where the Jap…

…anese Sokuon character is treated as a small Tsu.
zedseven · Feb 6, 2021 · b2db252 · b2db252
1 parent 6d2ef79
commit b2db252
Show file tree

Hide file tree

Showing 2 changed files with 27 additions and 22 deletions.
diff --git a/Romanization/Languages/Japanese/ModifiedHepburn.cs b/Romanization/Languages/Japanese/ModifiedHepburn.cs
@@ -1,6 +1,8 @@
 using Romanization.LanguageAgnostic;
+using System;
 using System.Collections.Generic;
 using System.Diagnostics.Contracts;
+using System.Globalization;
 
 // ReSharper disable CheckNamespace
 // ReSharper disable CommentTypo
@@ -32,11 +34,14 @@ public sealed class ModifiedHepburn : IRomanizationSystem
 			private readonly CharSub LongOSub = new CharSub($"o{Choonpu}", Constants.MacronO, false);
 			private readonly CharSub LongUSub = new CharSub($"u{Choonpu}", Constants.MacronU, false);
 
-			private readonly CharSub SyllabicNVowelsSub = new CharSub($"[{SyllabicNHiragana}{SyllabicNKatakana}]([{Constants.LatinVowels}])", "n'${1}");
-			private readonly CharSub SyllabicNConsonantsSub = new CharSub($"[{SyllabicNHiragana}{SyllabicNKatakana}]([{Constants.LatinConsonants}])", "n${1}");
+			private readonly CharSub SyllabicNVowelsSub =
+				new CharSub($"[{SyllabicNHiragana}{SyllabicNKatakana}]([{Constants.LatinVowels}])", "n'${1}");
+			private readonly CharSub SyllabicNConsonantsSub =
+				new CharSub($"[{SyllabicNHiragana}{SyllabicNKatakana}]([{Constants.LatinConsonants}])", "n${1}");
 
-			private readonly CharSub SokuonGeneralCaseSub = new CharSub($"[{SokuonHiragana}{SokuonKatakana}]([{Constants.LatinConsonants}])", "${1}${1}");
-			private readonly CharSub SokuonChCaseSub = new CharSub($"[{SokuonHiragana}{SokuonKatakana}]ch", "tch");
+			private readonly CharSub SokuonGeneralCaseSub =
+				new CharSub($"[{SokuonHiragana}{SokuonKatakana}]([{Constants.LatinConsonants}])", "${1}${1}", false);
+			private readonly CharSub SokuonChCaseSub = new CharSub($"[{SokuonHiragana}{SokuonKatakana}]ch", "tch", false);
 
 			/// <summary>
 			/// Instantiates a copy of the system to process romanizations.
@@ -291,23 +296,23 @@ public ModifiedHepburn()
 			/// <returns>A romanized version of the text, leaving unrecognized characters untouched. Note that all romanized text will be lowercase.</returns>
 			[Pure]
 			public string Process(string text)
-				=> text
+				=> Utilities.RunWithCulture(CultureInfo.GetCultureInfo("ja-JP"), () => text
 					// Replace common alternate characters
 					.ReplaceCommonAlternates()
 					// Insert spaces at boundaries between Latin characters and Japanese ones (ie. ニンテンドーDSiブラウザー)
 					.SeparateLanguageBoundaries()
 					// Do multi-char combinations first (Yōon)
-					.ReplaceFromChart(YoonChart)
+					.ReplaceFromChart(YoonChart, StringComparison.CurrentCulture)
 					// Then single-char replacements (Gojūon)
-					.ReplaceFromChart(GojuonChart)
+					.ReplaceFromChart(GojuonChart, StringComparison.CurrentCulture)
 					// Do special subsitutions
 					.ReplaceMany(
 						// Convert chōonpu usage in original text into macrons to mark long vowels in a romanized manner
 						LongASub, LongESub, LongISub, LongOSub, LongUSub,
 						// Render syllabic n as either "n'" or "n" based on whether or not it preceeds a vowel or consonant, respectively
 						SyllabicNVowelsSub, SyllabicNConsonantsSub,
 						// Take sokuon usage into account (repeating the following consonant to mark long consonants)
-						SokuonChCaseSub, SokuonGeneralCaseSub);
+						SokuonChCaseSub, SokuonGeneralCaseSub));
 		}
 	}
 }
diff --git a/RomanizationTests/JapaneseTests/ModifiedHepburnTests.cs b/RomanizationTests/JapaneseTests/ModifiedHepburnTests.cs
@@ -23,24 +23,24 @@ public void LongVowelTest()
 		[TestMethod]
 		public void SyllabicNTest()
 		{
-			Assert.AreEqual("annai",                _system.Process("あんない"));
-			Assert.AreEqual("gunma",                _system.Process("ぐんま"));
-			Assert.AreEqual("kan'i",                _system.Process("かんい"));
-			Assert.AreEqual("shin'you",             _system.Process("しんよう"));
+			Assert.AreEqual("annai",    _system.Process("あんない"));
+			Assert.AreEqual("gunma",    _system.Process("ぐんま"));
+			Assert.AreEqual("kan'i",    _system.Process("かんい"));
+			Assert.AreEqual("shin'you", _system.Process("しんよう"));
 		}
 
 		[TestMethod]
 		public void LongConsonantTest()
 		{
-			Assert.AreEqual("kekka",                _system.Process("けっか"));
-			Assert.AreEqual("sassato",              _system.Process("さっさと"));
-			Assert.AreEqual("zutto",                _system.Process("ずっと"));
-			Assert.AreEqual("kippu",                _system.Process("きっぷ"));
-			Assert.AreEqual("zasshi",               _system.Process("ざっし"));
-			Assert.AreEqual("issho",                _system.Process("いっしょ"));
-			Assert.AreEqual("kotchi",               _system.Process("こっち"));
-			Assert.AreEqual("matcha",               _system.Process("まっちゃ"));
-			Assert.AreEqual("mittsu",               _system.Process("みっつ"));
+			Assert.AreEqual("kekka",   _system.Process("けっか"));
+			Assert.AreEqual("sassato", _system.Process("さっさと"));
+			Assert.AreEqual("zutto",   _system.Process("ずっと"));
+			Assert.AreEqual("kippu",   _system.Process("きっぷ"));
+			Assert.AreEqual("zasshi",  _system.Process("ざっし"));
+			Assert.AreEqual("issho",   _system.Process("いっしょ"));
+			Assert.AreEqual("kotchi",  _system.Process("こっち"));
+			Assert.AreEqual("matcha",  _system.Process("まっちゃ"));
+			Assert.AreEqual("mittsu",  _system.Process("みっつ"));
 		}
 	}
-}
+}