AggregationQuery doesn't respect namespace from `com.google.cloud.datastore.Datastore` object

[jainsahab]

Given

I have 2 entities stored for Task kind under the namespace ns1 and i configure a com.google.cloud.datastore.Datastore service with ns1 namespace.

Scenario

When I try to run entity query with or without namespace through the com.google.cloud.datastore.Datastore object configured with namespace, i get same results, whereas in case of AggregationQuery i don't.

Code example

import static com.google.cloud.datastore.aggregation.Aggregation.count;
import static com.google.common.collect.Iterables.getOnlyElement;

import com.google.cloud.datastore.AggregationQuery;
import com.google.cloud.datastore.Datastore;
import com.google.cloud.datastore.DatastoreOptions;
import com.google.cloud.datastore.Entity;
import com.google.cloud.datastore.EntityQuery;
import com.google.cloud.datastore.Query;
import java.util.ArrayList;
import java.util.Iterator;
import java.util.List;

public class Playground {

  public static void main(String[] args) {
    DatastoreOptions datastoreOptions = DatastoreOptions.newBuilder()
        .setNamespace("ns1")
        .build();
    Datastore datastore = datastoreOptions.getService();

    EntityQuery entityQueryWithNamespace = Query.newEntityQueryBuilder()
        .setKind("Task")
        .setNamespace("ns1")
        .build();

    EntityQuery entityQueryWithoutNamespace = Query.newEntityQueryBuilder()
        .setKind("Task")
        .build();

    AggregationQuery aggregationQueryWithNamespace = Query.newAggregationQueryBuilder()
        .over(Query.newKeyQueryBuilder().setKind("Task").build())
        .addAggregation(count().as("my_count"))
        .setNamespace("ns1")
        .build();

    AggregationQuery aggregationQueryWithoutNamespace = Query.newAggregationQueryBuilder()
        .over(Query.newKeyQueryBuilder().setKind("Task").build())
        .addAggregation(count().as("my_count"))
        .build();

    System.out.println("Total elements from entity query with namespace " + toList(
        datastore.run(entityQueryWithNamespace)).size());  // 2
    System.out.println("Total elements from entity query without namespace " + toList(
        datastore.run(entityQueryWithoutNamespace)).size()); // 2

    System.out.println("Total elements from aggregation query with namespace " + getOnlyElement(
        datastore.runAggregation(aggregationQueryWithNamespace)).get("my_count")); // 2
    System.out.println("Total elements from aggregation query without namespace " + getOnlyElement(
        datastore.runAggregation(aggregationQueryWithoutNamespace)).get("my_count")); // 0
  }

  private static List<Entity> toList(Iterator<Entity> results) {
    ArrayList<Entity> list = new ArrayList<>();
    while (results.hasNext()) {
      list.add(results.next());
    }
    return list;
  }

}

Problem

last query execution of aggregationQueryWithoutNamespace returns 0 count.

Explanation

If a namespace is configured in AggragationQuery the library will use that namespace to run that query, and if it's not present, it'll completely ignore the one configured in com.google.cloud.datastore.Datastore service and simply run the query in default namespace and hence the diff in output.

java-datastore/google-cloud-datastore/src/main/java/com/google/cloud/datastore/execution/request/AggregationQueryRequestProtoPreparer.java

Lines 92 to 99 in 412be61

	private PartitionId getPartitionId(AggregationQuery aggregationQuery) {
	PartitionId.Builder builder =
	PartitionId.newBuilder().setProjectId(datastoreOptions.getProjectId());
	if (aggregationQuery.getNamespace() != null) {
	builder.setNamespaceId(aggregationQuery.getNamespace());
	}
	return builder.build();
	}

This part of code needs to be changed, where we can give priority to the namespace from AggregationQuery followed by com.google.cloud.datastore.Datastore service to keep it consistent with the Entity query.

Thanks!