Add discard_punctuation to nori token filter (#4591) (#4618)

Add discard_punctuation to nori token filter Co-authored-by: Stuart Cam <stuart.cam@elastic.co>
elastic · Apr 17, 2020 · 5b074c3 · 5b074c3
1 parent 2e7c9bc
commit 5b074c3
Show file tree

Hide file tree

Showing 3 changed files with 30 additions and 0 deletions.
diff --git a/src/Nest/Analysis/Analyzers/NoriAnalyzer.cs b/src/Nest/Analysis/Analyzers/NoriAnalyzer.cs
@@ -8,6 +8,7 @@ namespace Nest
 	/// <para> - nori_tokenizer</para>
 	/// <para> - nori_part_of_speech token filter</para>
 	/// <para> - nori_readingform token filter</para>
+	/// <para> - nori_number token filter</para>
 	/// <para> - lowercase token filter</para>
 	/// </summary>
 	public interface INoriAnalyzer : IAnalyzer

diff --git a/src/Nest/Analysis/Tokenizers/NoriTokenizer.cs b/src/Nest/Analysis/Tokenizers/NoriTokenizer.cs
@@ -30,6 +30,12 @@ public interface INoriTokenizer : ITokenizer
 		[DataMember(Name = "decompound_mode")]
 		NoriDecompoundMode? DecompoundMode { get; set; }
 
+		/// <summary>
+		/// Whether punctuation should be discarded from the output. Defaults to `true`.
+		/// </summary>
+		[DataMember(Name = "discard_punctuation")]
+		bool? DiscardPunctuation { get; set; }
+
 		/// <summary>
 		/// The Nori tokenizer uses the mecab-ko-dic dictionary by default. A user_dictionary with custom nouns (NNG) may be
 		/// appended to
@@ -57,6 +63,9 @@ public class NoriTokenizer : TokenizerBase, INoriTokenizer
 		/// <inheritdoc cref="INoriTokenizer.DecompoundMode" />
 		public NoriDecompoundMode? DecompoundMode { get; set; }
 
+		/// <inheritdoc cref="INoriTokenizer.DiscardPunctuation" />
+		public bool? DiscardPunctuation { get; set; }
+
 		/// <inheritdoc cref="INoriTokenizer.UserDictionary" />
 		public string UserDictionary { get; set; }
 
@@ -73,6 +82,7 @@ public class NoriTokenizerDescriptor
 		NoriDecompoundMode? INoriTokenizer.DecompoundMode { get; set; }
 		string INoriTokenizer.UserDictionary { get; set; }
 		IEnumerable<string> INoriTokenizer.UserDictionaryRules { get; set; }
+		bool? INoriTokenizer.DiscardPunctuation { get; set; }
 
 		/// <inheritdoc cref="INoriTokenizer.DecompoundMode" />
 		public NoriTokenizerDescriptor DecompoundMode(NoriDecompoundMode? mode) => Assign(mode, (a, v) => a.DecompoundMode = v);
@@ -85,5 +95,8 @@ public class NoriTokenizerDescriptor
 
 		/// <inheritdoc cref="INoriTokenizer.UserDictionaryRules" />
 		public NoriTokenizerDescriptor UserDictionaryRules(IEnumerable<string> rules) => Assign(rules, (a, v) => a.UserDictionaryRules = v);
+
+		/// <inheritdoc cref="INoriTokenizer.DiscardPunctuation" />
+		public NoriTokenizerDescriptor DiscardPunctuation(bool? discard = true) => Assign(discard, (a, v) => a.DiscardPunctuation = v);
 	}
 }
diff --git a/tests/Tests/Analysis/Tokenizers/TokenizerTests.cs b/tests/Tests/Analysis/Tokenizers/TokenizerTests.cs
@@ -332,5 +332,21 @@ public class CharGroupTests : TokenizerAssertionBase<CharGroupTests>
 
 			public override string Name => "char_group";
 		}
+
+		[SkipVersion("<7.7.0", "discard_punctuation introduced in 7.7.0")]
+		public class DiscardPunctuationTests : TokenizerAssertionBase<DiscardPunctuationTests>
+		{
+			public override FuncTokenizer Fluent => (n, t) => t.Nori(n, e => e
+				.DiscardPunctuation()
+			);
+
+			public override ITokenizer Initializer => new NoriTokenizer
+			{
+				DiscardPunctuation = true
+			};
+
+			public override object Json => new { type = "nori_tokenizer", discard_punctuation = true };
+			public override string Name => "nori";
+		}
 	}
 }