LUCENE-10379: Count directly into the dense values array in FastTaxonomyFacetCounts#countAll

Co-authored-by: guofeng.my <[email protected]>

Author: gsmiller

lucene/CHANGES.txt

@@ -122,6 +122,9 @@ Optimizations
 
 * LUCENE-10356: Further optimize facet counting for single-valued TaxonomyFacetCounts. (Greg Miller)
 
+* LUCENE-10379: Count directly into the dense values array in FastTaxonomyFacetCounts#countAll.
+  (Guo Feng, Greg Miller)
+
 Changes in runtime behavior
 ---------------------
 

lucene/facet/src/java/org/apache/lucene/facet/taxonomy/FastTaxonomyFacetCounts.java

@@ -70,7 +70,7 @@ public FastTaxonomyFacetCounts(
     countAll(reader);
   }
 
-  private final void count(List<MatchingDocs> matchingDocs) throws IOException {
+  private void count(List<MatchingDocs> matchingDocs) throws IOException {
     for (MatchingDocs hits : matchingDocs) {
       SortedNumericDocValues multiValued =
           FacetUtils.loadOrdinalValues(hits.context.reader(), indexFieldName);
@@ -85,13 +85,27 @@ private final void count(List<MatchingDocs> matchingDocs) throws IOException {
           ConjunctionUtils.intersectIterators(Arrays.asList(hits.bits.iterator(), valuesIt));
 
       if (singleValued != null) {
-        while (it.nextDoc() != DocIdSetIterator.NO_MORE_DOCS) {
-          increment((int) singleValued.longValue());
+        if (values != null) {
+          while (it.nextDoc() != DocIdSetIterator.NO_MORE_DOCS) {
+            values[(int) singleValued.longValue()]++;
+          }
+        } else {
+          while (it.nextDoc() != DocIdSetIterator.NO_MORE_DOCS) {
+            sparseValues.addTo((int) singleValued.longValue(), 1);
+          }
         }
       } else {
-        while (it.nextDoc() != DocIdSetIterator.NO_MORE_DOCS) {
-          for (int i = 0; i < multiValued.docValueCount(); i++) {
-            increment((int) multiValued.nextValue());
+        if (values != null) {
+          while (it.nextDoc() != DocIdSetIterator.NO_MORE_DOCS) {
+            for (int i = 0; i < multiValued.docValueCount(); i++) {
+              values[(int) multiValued.nextValue()]++;
+            }
+          }
+        } else {
+          while (it.nextDoc() != DocIdSetIterator.NO_MORE_DOCS) {
+            for (int i = 0; i < multiValued.docValueCount(); i++) {
+              sparseValues.addTo((int) multiValued.nextValue(), 1);
+            }
           }
         }
       }
@@ -100,7 +114,8 @@ private final void count(List<MatchingDocs> matchingDocs) throws IOException {
     rollup();
   }
 
-  private final void countAll(IndexReader reader) throws IOException {
+  private void countAll(IndexReader reader) throws IOException {
+    assert values != null;
     for (LeafReaderContext context : reader.leaves()) {
       SortedNumericDocValues multiValued =
           FacetUtils.loadOrdinalValues(context.reader(), indexFieldName);
@@ -118,7 +133,7 @@ private final void countAll(IndexReader reader) throws IOException {
           if (liveDocs != null && liveDocs.get(doc) == false) {
             continue;
           }
-          increment((int) singleValued.longValue());
+          values[(int) singleValued.longValue()]++;
         }
       } else {
         for (int doc = multiValued.nextDoc();
@@ -128,7 +143,7 @@ private final void countAll(IndexReader reader) throws IOException {
             continue;
           }
           for (int i = 0; i < multiValued.docValueCount(); i++) {
-            increment((int) multiValued.nextValue());
+            values[(int) multiValued.nextValue()]++;
           }
         }
       }

lucene/facet/src/java/org/apache/lucene/facet/taxonomy/IntTaxonomyFacets.java

@@ -31,10 +31,21 @@
 /** Base class for all taxonomy-based facets that aggregate to a per-ords int[]. */
 public abstract class IntTaxonomyFacets extends TaxonomyFacets {
 
-  /** Per-ordinal value. */
-  private final int[] values;
+  /**
+   * Dense ordinal values.
+   *
+   * <p>We are making this and {@link #sparseValues} protected for some expert usage. e.g. It can be
+   * checked which is being used before a loop instead of calling {@link #increment} for each
+   * iteration.
+   */
+  protected final int[] values;
 
-  private final IntIntHashMap sparseValues;
+  /**
+   * Sparse ordinal values.
+   *
+   * @see #values for why protected.
+   */
+  protected final IntIntHashMap sparseValues;
 
   /** Sole constructor. */
   protected IntTaxonomyFacets(