You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@lucene.apache.org by mk...@apache.org on 2025/04/05 19:09:03 UTC

(lucene) branch main updated: KeywordField.newSetQuery() to reuse prefixed terms in IndexOrDocValuesQuery (#14435)

This is an automated email from the ASF dual-hosted git repository.

mkhl pushed a commit to branch main
in repository https://gitbox.apache.org/repos/asf/lucene.git


The following commit(s) were added to refs/heads/main by this push:
     new 219cd5eaebc KeywordField.newSetQuery() to reuse prefixed terms in IndexOrDocValuesQuery (#14435)
219cd5eaebc is described below

commit 219cd5eaebca686c1df93226f47924337ee1a783
Author: Mikhail Khludnev <mk...@users.noreply.github.com>
AuthorDate: Sat Apr 5 22:08:58 2025 +0300

    KeywordField.newSetQuery() to reuse prefixed terms in IndexOrDocValuesQuery (#14435)
    
    * KeywordField.newSetQuery() reuses prefixed terms.
    
    fix #14425
---
 lucene/CHANGES.txt                                 |  2 ++
 .../org/apache/lucene/document/KeywordField.java   |  6 ++----
 .../org/apache/lucene/search/TermInSetQuery.java   | 24 +++++++++++++++++++++-
 3 files changed, 27 insertions(+), 5 deletions(-)

diff --git a/lucene/CHANGES.txt b/lucene/CHANGES.txt
index 961d52c8eca..867f53b1e06 100644
--- a/lucene/CHANGES.txt
+++ b/lucene/CHANGES.txt
@@ -73,6 +73,8 @@ Optimizations
 
 * GITHUB#14268: PointInSetQuery early exit on non-matching segments. (hanbj)
 
+* GITHUB#14425: KeywordField.newSetQuery() reuses prefixed terms (Mikhail Khludnev)
+
 Bug Fixes
 ---------------------
 (No changes)
diff --git a/lucene/core/src/java/org/apache/lucene/document/KeywordField.java b/lucene/core/src/java/org/apache/lucene/document/KeywordField.java
index 31159cc5384..2ed22b2e141 100644
--- a/lucene/core/src/java/org/apache/lucene/document/KeywordField.java
+++ b/lucene/core/src/java/org/apache/lucene/document/KeywordField.java
@@ -22,7 +22,6 @@ import org.apache.lucene.index.DocValuesType;
 import org.apache.lucene.index.IndexOptions;
 import org.apache.lucene.index.Term;
 import org.apache.lucene.search.ConstantScoreQuery;
-import org.apache.lucene.search.IndexOrDocValuesQuery;
 import org.apache.lucene.search.MultiTermQuery;
 import org.apache.lucene.search.Query;
 import org.apache.lucene.search.SortField;
@@ -175,9 +174,8 @@ public class KeywordField extends Field {
   public static Query newSetQuery(String field, Collection<BytesRef> values) {
     Objects.requireNonNull(field, "field must not be null");
     Objects.requireNonNull(values, "values must not be null");
-    Query indexQuery = new TermInSetQuery(field, values);
-    Query dvQuery = new TermInSetQuery(MultiTermQuery.DOC_VALUES_REWRITE, field, values);
-    return new IndexOrDocValuesQuery(indexQuery, dvQuery);
+    return TermInSetQuery.newIndexOrDocValuesQuery(
+        MultiTermQuery.CONSTANT_SCORE_BLENDED_REWRITE, field, values);
   }
 
   /**
diff --git a/lucene/core/src/java/org/apache/lucene/search/TermInSetQuery.java b/lucene/core/src/java/org/apache/lucene/search/TermInSetQuery.java
index c82df0ac1eb..4c8f4c9b62e 100644
--- a/lucene/core/src/java/org/apache/lucene/search/TermInSetQuery.java
+++ b/lucene/core/src/java/org/apache/lucene/search/TermInSetQuery.java
@@ -84,8 +84,30 @@ public class TermInSetQuery extends MultiTermQuery implements Accountable {
     termDataHashCode = termData.hashCode();
   }
 
+  /**
+   * Creates a new {@link IndexOrDocValuesQuery} combining two {@link TermInSetQuery} with the same
+   * terms allowing to pack them only once that's faster. Doc Values query always uses {@link
+   * MultiTermQuery#DOC_VALUES_REWRITE}.
+   *
+   * @param field field name for indexed and doc values queries.
+   * @param indexRewriteMethod rewrite method used for indexed query.
+   * @param terms collection of {@link BytesRef}. Note: passing {@link SortedSet} with default
+   *     comparator let to bypass terms sorting.
+   */
+  public static IndexOrDocValuesQuery newIndexOrDocValuesQuery(
+      RewriteMethod indexRewriteMethod, String field, Collection<BytesRef> terms) {
+    PrefixCodedTerms packed = packTerms(field, terms);
+    Query indexQuery = new TermInSetQuery(indexRewriteMethod, field, packed);
+    Query dvQuery = new TermInSetQuery(MultiTermQuery.DOC_VALUES_REWRITE, field, packed);
+    return new IndexOrDocValuesQuery(indexQuery, dvQuery);
+  }
+
   private TermInSetQuery(String field, PrefixCodedTerms termData) {
-    super(field, MultiTermQuery.CONSTANT_SCORE_BLENDED_REWRITE);
+    this(MultiTermQuery.CONSTANT_SCORE_BLENDED_REWRITE, field, termData);
+  }
+
+  private TermInSetQuery(RewriteMethod rewrite, String field, PrefixCodedTerms termData) {
+    super(field, rewrite);
     this.field = field;
     this.termData = termData;
     termDataHashCode = termData.hashCode();