Terms aggs that run as filters ignore shard_min_doc_count (#69323)

This change handles `shard_min_doc_count` for terms aggregation that run as filters. Closes #69312
elastic · Feb 22, 2021 · 27c31fa · 27c31fa
1 parent 914a85a
commit 27c31fa
Show file tree

Hide file tree

Showing 2 changed files with 34 additions and 0 deletions.
diff --git a/.../org/elasticsearch/search/aggregations/bucket/terms/StringTermsAggregatorFromFilters.java b/.../org/elasticsearch/search/aggregations/bucket/terms/StringTermsAggregatorFromFilters.java
@@ -159,6 +159,9 @@ protected boolean lessThan(OrdBucket a, OrdBucket b) {
             };
             OrdBucket spare = null;
             for (InternalFilters.InternalBucket b : filters.getBuckets()) {
+                if (b.getDocCount() < bucketCountThresholds.getShardMinDocCount()) {
+                    continue;
+                }
                 if (spare == null) {
                     spare = new OrdBucket(showTermDocCountError, format);
                 } else {
@@ -194,6 +197,9 @@ protected boolean lessThan(OrdBucket a, OrdBucket b) {
         } else {
             buckets = new ArrayList<>(filters.getBuckets().size());
             for (InternalFilters.InternalBucket b : filters.getBuckets()) {
+                if (b.getDocCount() < bucketCountThresholds.getShardMinDocCount()) {
+                    continue;
+                }
                 buckets.add(buildBucket(b));
             }
             Collections.sort(buckets, reduceOrder.comparator());

diff --git a/...rc/test/java/org/elasticsearch/search/aggregations/bucket/terms/TermsAggregatorTests.java b/...rc/test/java/org/elasticsearch/search/aggregations/bucket/terms/TermsAggregatorTests.java
@@ -102,6 +102,7 @@
 import java.io.IOException;
 import java.net.InetAddress;
 import java.util.ArrayList;
+import java.util.Arrays;
 import java.util.Collections;
 import java.util.Comparator;
 import java.util.HashMap;
@@ -258,6 +259,33 @@ public void testSimple() throws Exception {
         }, fieldType);
     }
 
+    public void testStringShardMinDocCount() throws IOException {
+        MappedFieldType fieldType = new KeywordFieldMapper.KeywordFieldType("string", true, true, null);
+        for (TermsAggregatorFactory.ExecutionMode executionMode : TermsAggregatorFactory.ExecutionMode.values()) {
+            TermsAggregationBuilder aggregationBuilder = new TermsAggregationBuilder("_name")
+                .field("string")
+                .executionHint(executionMode.toString())
+                .size(2)
+                .minDocCount(2)
+                .shardMinDocCount(2)
+                .order(BucketOrder.key(true));
+            testCase(aggregationBuilder, new MatchAllDocsQuery(), iw -> {
+                // force single shard/segment
+                iw.addDocuments(Arrays.asList(
+                    doc(fieldType, "a", "b"),
+                    doc(fieldType, "", "c", "d"),
+                    doc(fieldType, "b", "d"),
+                    doc(fieldType, "b")));
+            }, (InternalTerms<?, ?> result) -> {
+                assertEquals(2, result.getBuckets().size());
+                assertEquals("b", result.getBuckets().get(0).getKeyAsString());
+                assertEquals(3L, result.getBuckets().get(0).getDocCount());
+                assertEquals("d", result.getBuckets().get(1).getKeyAsString());
+                assertEquals(2L, result.getBuckets().get(1).getDocCount());
+            }, fieldType);
+        }
+    }
+
     public void testManyUniqueTerms() throws Exception {
         MappedFieldType fieldType = new KeywordFieldMapper.KeywordFieldType("string", randomBoolean(), true, null);
         TermsAggregationBuilder aggregationBuilder = new TermsAggregationBuilder("_name")