tweak keyword field to use binary DV

rishabhmaurya · Dec 12, 2024 · 61d7250 · 61d7250
1 parent 05513df
commit 61d7250
Showing 1 changed file with 18 additions and 3 deletions.
diff --git a/server/src/main/java/org/opensearch/index/mapper/KeywordFieldMapper.java b/server/src/main/java/org/opensearch/index/mapper/KeywordFieldMapper.java
@@ -34,6 +34,7 @@
 
 import org.apache.lucene.analysis.TokenStream;
 import org.apache.lucene.analysis.tokenattributes.CharTermAttribute;
+import org.apache.lucene.document.BinaryDocValuesField;
 import org.apache.lucene.document.Field;
 import org.apache.lucene.document.FieldType;
 import org.apache.lucene.document.SortedSetDocValuesField;
@@ -61,6 +62,7 @@
 import org.opensearch.index.analysis.NamedAnalyzer;
 import org.opensearch.index.compositeindex.datacube.DimensionType;
 import org.opensearch.index.fielddata.IndexFieldData;
+import org.opensearch.index.fielddata.plain.BinaryIndexFieldData;
 import org.opensearch.index.fielddata.plain.SortedSetOrdinalsIndexFieldData;
 import org.opensearch.index.query.QueryShardContext;
 import org.opensearch.index.similarity.SimilarityProvider;
@@ -167,6 +169,13 @@ public static class Builder extends ParametrizedFieldMapper.Builder {
             false
         );
 
+        private final Parameter<Boolean> useBinaryDV = Parameter.boolParam(
+            "use_binary_doc_value",
+            true,
+            m -> toType(m).useBinaryDocValue,
+            false
+        );
+
         private final Parameter<Map<String, String>> meta = Parameter.metaParam();
         private final Parameter<Float> boost = Parameter.boostParam();
 
@@ -333,7 +342,8 @@ NamedAnalyzer normalizer() {
         @Override
         public IndexFieldData.Builder fielddataBuilder(String fullyQualifiedIndexName, Supplier<SearchLookup> searchLookup) {
             failIfNoDocValues();
-            return new SortedSetOrdinalsIndexFieldData.Builder(name(), CoreValuesSourceType.BYTES);
+            return new BinaryIndexFieldData.Builder(name() + ".binary", CoreValuesSourceType.BYTES);
+            // return new SortedSetOrdinalsIndexFieldData.Builder(name(), CoreValuesSourceType.BYTES);
         }
 
         @Override
@@ -661,6 +671,7 @@ public Query wildcardQuery(
     private final SimilarityProvider similarity;
     private final String normalizerName;
     private final boolean splitQueriesOnWhitespace;
+    private final boolean useBinaryDocValue;
 
     private final IndexAnalyzers indexAnalyzers;
 
@@ -684,7 +695,7 @@ protected KeywordFieldMapper(
         this.similarity = builder.similarity.getValue();
         this.normalizerName = builder.normalizer.getValue();
         this.splitQueriesOnWhitespace = builder.splitQueriesOnWhitespace.getValue();
-
+        this.useBinaryDocValue = builder.useBinaryDV.getValue();
         this.indexAnalyzers = builder.indexAnalyzers;
     }
 
@@ -741,7 +752,11 @@ protected void parseCreateField(ParseContext context) throws IOException {
         }
 
         if (fieldType().hasDocValues()) {
-            context.doc().add(new SortedSetDocValuesField(fieldType().name(), binaryValue));
+            if (useBinaryDocValue) {
+                context.doc().add(new BinaryDocValuesField(fieldType().name() + ".binary", binaryValue));
+            } else {
+                context.doc().add(new SortedSetDocValuesField(fieldType().name(), binaryValue));
+            }
         }
     }