Optimize TopGroups.merge() maxScore computation for relevance sorting (#15879)

When docSort is Sort.RELEVANCE, derive each group's maxScore directly from mergedTopDocs.scoreDocs[0] instead of accumulating across shards via nonNANmax.

Also hoist docSort.equals(Sort.RELEVANCE) into a local boolean to avoid repeated equals() calls in the per-group and per-shard loops.

Signed-off-by: Binlong Gao <gbinlong@amazon.com>

Author: gaobinlong

lucene/CHANGES.txt

@@ -364,6 +364,8 @@ Optimizations
 
 * GITHUB#15779: Improve BytesRefHash.add performance by optimize rehash operation (tyronecai)
 
+* GITHUB#15879: Optimize TopGroups.merge() maxScore computation for relevance sorting (Binlong Gao)
+
 * GITHUB#15896: Skip heavy TreeSet opts for the first group in SearchGroup#merge (Binlong Gao)
 
 * GITHUB#15886: Cache frozen FieldType to skip redundant schema validation. (Tim Brooks)

lucene/grouping/src/java/org/apache/lucene/search/grouping/TopGroups.java

@@ -156,7 +156,8 @@ public static <T> TopGroups<T> merge(
     final GroupDocs<T>[] mergedGroupDocs = new GroupDocs[numGroups];
 
     final TopDocs[] shardTopDocs;
-    if (docSort.equals(Sort.RELEVANCE)) {
+    final boolean sortByRelevance = docSort.equals(Sort.RELEVANCE);
+    if (sortByRelevance) {
       shardTopDocs = new TopDocs[shardGroups.size()];
     } else {
       shardTopDocs = new TopFieldDocs[shardGroups.size()];
@@ -190,7 +191,7 @@ public static <T> TopGroups<T> merge(
         }
         */
 
-        if (docSort.equals(Sort.RELEVANCE)) {
+        if (sortByRelevance) {
           shardTopDocs[shardIDX] =
               new TopDocs(shardGroupDocs.totalHits(), shardGroupDocs.scoreDocs());
         } else {
@@ -203,15 +204,21 @@ public static <T> TopGroups<T> merge(
           shardTopDocs[shardIDX].scoreDocs[i].shardIndex = shardIDX;
         }
 
-        maxScore = nonNANmax(maxScore, shardGroupDocs.maxScore());
+        if (!sortByRelevance) {
+          maxScore = nonNANmax(maxScore, shardGroupDocs.maxScore());
+        }
         assert shardGroupDocs.totalHits().relation() == Relation.EQUAL_TO;
         totalHits += shardGroupDocs.totalHits().value();
         scoreSum += shardGroupDocs.score();
       }
 
       final TopDocs mergedTopDocs;
-      if (docSort.equals(Sort.RELEVANCE)) {
+      if (sortByRelevance) {
         mergedTopDocs = TopDocs.merge(docOffset + docTopN, shardTopDocs);
+        // When sorting by relevance, the highest-scoring doc is first, so we can
+        // derive maxScore directly instead of accumulating across shards.
+        maxScore =
+            mergedTopDocs.scoreDocs.length == 0 ? Float.NaN : mergedTopDocs.scoreDocs[0].score;
       } else {
         mergedTopDocs = TopDocs.merge(docSort, docOffset + docTopN, (TopFieldDocs[]) shardTopDocs);
       }