fix unit tests

Signed-off-by: Lantao Jin <ltjin@amazon.com>

Author: LantaoJin

docs/user/ppl/cmd/stats.rst

@@ -39,7 +39,7 @@ stats [bucket_nullable=bool] <aggregation>... [by-clause]
 
 * aggregation: mandatory. A aggregation function. The argument of aggregation must be field.
 
-* bucket_nullable: optional. Controls whether the stats command includes null buckets in group-by aggregations. When set to ``false``, the aggregation ignores records where the group-by field is null, resulting in faster performance by excluding null bucket. The default value of ``bucket_nullable`` is determined by ``plugins.ppl.syntax.legacy.preferred``:
+* bucket_nullable: optional (since 3.3.0). Controls whether the stats command includes null buckets in group-by aggregations. When set to ``false``, the aggregation ignores records where the group-by field is null, resulting in faster performance by excluding null bucket. The default value of ``bucket_nullable`` is determined by ``plugins.ppl.syntax.legacy.preferred``:
 
  * When ``plugins.ppl.syntax.legacy.preferred=true``, ``bucket_nullable`` defaults to ``true``
  * When ``plugins.ppl.syntax.legacy.preferred=false``, ``bucket_nullable`` defaults to ``false``
@@ -692,9 +692,11 @@ PPL query::
 Example 15: Ignore null bucket
 ==============================
 
+Note: This argument requires version 3.3.0 or above.
+
 PPL query::
 
-    os> source=accounts | stats bucket_nullable=false count() as cnt by email;
+    PPL> source=accounts | stats bucket_nullable=false count() as cnt by email;
     fetched rows / total rows = 3/3
     +-----+-----------------------+
     | cnt | email                 |

opensearch/src/main/java/org/opensearch/sql/opensearch/storage/script/aggregation/AggregationQueryBuilder.java

@@ -83,6 +83,7 @@ public AggregationQueryBuilder(ExpressionSerializer serializer) {
           new NoBucketAggregationParser(metrics.getRight()));
     } else if (groupByList.size() == 1 && !bucketNullable) {
       // one bucket, use values source bucket builder for getting better performance
+      // TODO for multiple buckets, use MultiTermsAggregationBuilder
       return Pair.of(
           Collections.singletonList(
               bucketBuilder.build(groupByList.getFirst()).subAggregations(metrics.getLeft())),

opensearch/src/main/java/org/opensearch/sql/opensearch/storage/script/aggregation/dsl/BucketAggregationBuilder.java

@@ -11,10 +11,13 @@
 import static org.opensearch.sql.opensearch.storage.script.aggregation.AggregationQueryBuilder.AGGREGATION_BUCKET_SIZE;
 
 import java.util.List;
+import java.util.stream.Collectors;
 import org.opensearch.search.aggregations.bucket.histogram.DateHistogramAggregationBuilder;
 import org.opensearch.search.aggregations.bucket.histogram.DateHistogramInterval;
 import org.opensearch.search.aggregations.bucket.histogram.HistogramAggregationBuilder;
+import org.opensearch.search.aggregations.bucket.terms.MultiTermsAggregationBuilder;
 import org.opensearch.search.aggregations.bucket.terms.TermsAggregationBuilder;
+import org.opensearch.search.aggregations.support.MultiTermsValuesSourceConfig;
 import org.opensearch.search.aggregations.support.ValueType;
 import org.opensearch.search.aggregations.support.ValuesSourceAggregationBuilder;
 import org.opensearch.sql.ast.expression.SpanUnit;
@@ -32,7 +35,7 @@ public BucketAggregationBuilder(ExpressionSerializer serializer) {
     this.helper = new AggregationBuilderHelper(serializer);
   }
 
-  /** Build the list of ValuesSourceAggregationBuilder. */
+  /** Build the ValuesSourceAggregationBuilder. */
   public ValuesSourceAggregationBuilder<?> build(NamedExpression expr) {
     if (expr.getDelegated() instanceof SpanExpression) {
       SpanExpression spanExpr = (SpanExpression) expr.getDelegated();
@@ -55,6 +58,34 @@ public ValuesSourceAggregationBuilder<?> build(NamedExpression expr) {
     }
   }
 
+  /** Build the MultiTermsAggregationBuilder. */
+  public MultiTermsAggregationBuilder buildMultipleTerms(List<NamedExpression> exprs) {
+    MultiTermsAggregationBuilder sourceBuilder =
+        new MultiTermsAggregationBuilder(
+            exprs.stream().map(NamedExpression::getName).collect(Collectors.joining("_")));
+    sourceBuilder.terms(
+        exprs.stream()
+            .map(
+                expr -> {
+                  MultiTermsValuesSourceConfig.Builder config =
+                      new MultiTermsValuesSourceConfig.Builder();
+                  config.setFieldName(expr.getName());
+                  // Time types values are converted to LONG in ExpressionAggregationScript::execute
+                  if ((expr.getDelegated().type() instanceof OpenSearchDateType
+                          && List.of(TIMESTAMP, TIME, DATE)
+                              .contains(
+                                  ((OpenSearchDateType) expr.getDelegated().type())
+                                      .getExprCoreType()))
+                      || List.of(TIMESTAMP, TIME, DATE).contains(expr.getDelegated().type())) {
+                    config.setUserValueTypeHint(ValueType.LONG);
+                  }
+                  return config.build();
+                })
+            .toList());
+    sourceBuilder.size(AGGREGATION_BUCKET_SIZE);
+    return sourceBuilder;
+  }
+
   public static ValuesSourceAggregationBuilder<?> buildHistogram(
       String name, String field, Double value, SpanUnit unit) {
     switch (unit) {

opensearch/src/test/java/org/opensearch/sql/opensearch/request/AggregateAnalyzerTest.java

@@ -138,8 +138,7 @@ void analyze_aggCall_simple() throws ExpressionNotAnalyzableException {
             List.of(countCall, avgCall, sumCall, minCall, maxCall), ImmutableBitSet.of());
     Project project = createMockProject(List.of(0));
     Pair<List<AggregationBuilder>, OpenSearchAggregationResponseParser> result =
-        AggregateAnalyzer.analyze(
-            aggregate, project, rowType, fieldTypes, outputFields, null, true);
+        AggregateAnalyzer.analyze(aggregate, project, rowType, fieldTypes, outputFields, null);
     assertEquals(
         "[{\"cnt\":{\"value_count\":{\"field\":\"_index\"}}},"
             + " {\"avg\":{\"avg\":{\"field\":\"a\"}}},"
@@ -220,8 +219,7 @@ void analyze_aggCall_extended() throws ExpressionNotAnalyzableException {
             List.of(varSampCall, varPopCall, stddevSampCall, stddevPopCall), ImmutableBitSet.of());
     Project project = createMockProject(List.of(0));
     Pair<List<AggregationBuilder>, OpenSearchAggregationResponseParser> result =
-        AggregateAnalyzer.analyze(
-            aggregate, project, rowType, fieldTypes, outputFields, null, true);
+        AggregateAnalyzer.analyze(aggregate, project, rowType, fieldTypes, outputFields, null);
     assertEquals(
         "[{\"var_samp\":{\"extended_stats\":{\"field\":\"a\",\"sigma\":2.0}}},"
             + " {\"var_pop\":{\"extended_stats\":{\"field\":\"a\",\"sigma\":2.0}}},"
@@ -260,8 +258,7 @@ void analyze_groupBy() throws ExpressionNotAnalyzableException {
     Aggregate aggregate = createMockAggregate(List.of(aggCall), ImmutableBitSet.of(0, 1));
     Project project = createMockProject(List.of(0, 1));
     Pair<List<AggregationBuilder>, OpenSearchAggregationResponseParser> result =
-        AggregateAnalyzer.analyze(
-            aggregate, project, rowType, fieldTypes, outputFields, null, true);
+        AggregateAnalyzer.analyze(aggregate, project, rowType, fieldTypes, outputFields, null);
 
     assertEquals(
         "[{\"composite_buckets\":{\"composite\":{\"size\":1000,\"sources\":["
@@ -304,7 +301,7 @@ void analyze_aggCall_TextWithoutKeyword() {
             ExpressionNotAnalyzableException.class,
             () ->
                 AggregateAnalyzer.analyze(
-                    aggregate, project, rowType, fieldTypes, List.of("sum"), null, true));
+                    aggregate, project, rowType, fieldTypes, List.of("sum"), null));
     assertEquals("[field] must not be null: [sum]", exception.getCause().getMessage());
   }
 
@@ -331,14 +328,15 @@ void analyze_groupBy_TextWithoutKeyword() {
             ExpressionNotAnalyzableException.class,
             () ->
                 AggregateAnalyzer.analyze(
-                    aggregate, project, rowType, fieldTypes, outputFields, null, true));
+                    aggregate, project, rowType, fieldTypes, outputFields, null));
     assertEquals("[field] must not be null", exception.getCause().getMessage());
   }
 
   private Aggregate createMockAggregate(List<AggregateCall> calls, ImmutableBitSet groups) {
     Aggregate agg = mock(Aggregate.class);
     when(agg.getGroupSet()).thenReturn(groups);
     when(agg.getAggCallList()).thenReturn(calls);
+    when(agg.getHints()).thenReturn(ImmutableList.of());
     return agg;
   }