Add support for inlined user dictionary in the Kuromoji plugin (#4138)

Addresses elastic/elasticsearch#45489

(cherry picked from commit c458da9)

Author: Mpdreamz

src/Nest/Analysis/Plugins/Kuromoji/KuromojiTokenizer.cs

@@ -1,4 +1,5 @@
-using System.Runtime.Serialization;
+using System.Collections.Generic;
+using System.Runtime.Serialization;
 using Elasticsearch.Net.Utf8Json;
 
 namespace Nest
@@ -44,6 +45,11 @@ public interface IKuromojiTokenizer : ITokenizer
 		/// </summary>
 		[DataMember(Name ="user_dictionary")]
 		string UserDictionary { get; set; }
+
+		/// <summary> Inline rule version of <see cref="UserDictionary"/> </summary>
+		[DataMember(Name ="user_dictionary_rules")]
+		IEnumerable<string> UserDictionaryRules { get; set; }
+
 	}
 
 	/// <inheritdoc />
@@ -65,6 +71,9 @@ public class KuromojiTokenizer : TokenizerBase, IKuromojiTokenizer
 
 		/// <inheritdoc />
 		public string UserDictionary { get; set; }
+
+		/// <inheritdoc />
+		public IEnumerable<string> UserDictionaryRules { get; set; }
 	}
 
 	/// <inheritdoc />
@@ -78,6 +87,7 @@ public class KuromojiTokenizerDescriptor
 		int? IKuromojiTokenizer.NBestCost { get; set; }
 		string IKuromojiTokenizer.NBestExamples { get; set; }
 		string IKuromojiTokenizer.UserDictionary { get; set; }
+		IEnumerable<string> IKuromojiTokenizer.UserDictionaryRules { get; set; }
 
 		/// <inheritdoc />
 		public KuromojiTokenizerDescriptor Mode(KuromojiTokenizationMode? mode) => Assign(mode, (a, v) => a.Mode = v);
@@ -93,5 +103,11 @@ public class KuromojiTokenizerDescriptor
 
 		/// <inheritdoc />
 		public KuromojiTokenizerDescriptor NBestCost(int? cost) => Assign(cost, (a, v) => a.NBestCost = v);
+
+		/// <inheritdoc />
+		public KuromojiTokenizerDescriptor UserDictionaryRules(IEnumerable<string> rules) => Assign(rules, (a, v) => a.UserDictionaryRules = rules);
+
+		/// <inheritdoc />
+		public KuromojiTokenizerDescriptor UserDictionaryRules(params string[] rules) => Assign(rules, (a, v) => a.UserDictionaryRules = rules);
 	}
 }

src/Tests/Tests/Analysis/Tokenizers/TokenizerTests.cs

@@ -114,23 +114,27 @@ public class IcuTests : TokenizerAssertionBase<IcuTests>
 			public override string Name => "icu";
 		}
 
+		[SkipVersion("<7.4.0", "not all options available before this version")]
 		public class KuromojiTests : TokenizerAssertionBase<KuromojiTests>
 		{
 			private const string Example = "/箱根山-箱根/成田空港-成田/";
+			private const string Inline = "東京スカイツリー,東京 スカイツリー,トウキョウ スカイツリー,カスタム名詞";
 
 			public override FuncTokenizer Fluent => (n, t) => t.Kuromoji(n, e => e
 				.Mode(KuromojiTokenizationMode.Extended)
 				.DiscardPunctuation()
 				.NBestExamples(Example)
 				.NBestCost(1000)
+				.UserDictionaryRules(Inline)
 			);
 
 			public override ITokenizer Initializer => new KuromojiTokenizer
 			{
 				Mode = KuromojiTokenizationMode.Extended,
 				DiscardPunctuation = true,
 				NBestExamples = Example,
-				NBestCost = 1000
+				NBestCost = 1000,
+				UserDictionaryRules = new [] { Inline }
 			};
 
 			public override object Json => new
@@ -139,7 +143,8 @@ public class KuromojiTests : TokenizerAssertionBase<KuromojiTests>
 				mode = "extended",
 				nbest_cost = 1000,
 				nbest_examples = Example,
-				type = "kuromoji_tokenizer"
+				type = "kuromoji_tokenizer",
+				user_dictionary_rules = new [] { Inline }
 			};
 
 			public override string Name => "kuro";