How to project the result of an aggregate query result with Panache and MongoDB?

Consider this method:

public static Uni<List<ReactivePanacheMongoEntityBase>> getAverage(String productId) {
    List<Bson> pipeline = new ArrayList<>();
    Bson match = Aggregates.match(Filters.eq("productId", new ObjectId(productId)));
    Bson group = Aggregates.group("$productId", new BsonField("Average", new Document("$avg", "$score")));
    pipeline.add(match);
    pipeline.add(group);
    return Rating.mongoCollection().aggregate(pipeline).collect().asList();
}

How to I project the results to a custom class? The project() only exists in conjunction with find() 🤔

In MongoDB, you can use the project aggregation stage to specify the shape of the documents returned by the aggregation pipeline. To map the result to a custom Java class, you can use a combination of the project stage and the codec registry.

Let's create a custom class to represent the result. Suppose you want to map the result to a class called AverageRating.

public class AverageRating {
    private ObjectId productId;
    private Double average;

    // Getters and setters
}

Modify the getAverage method to include the project stage and to map the results to AverageRating class. You can use the map method of the Uni class for this.

import org.bson.Document;
import org.bson.codecs.configuration.CodecRegistries;
import org.bson.codecs.configuration.CodecRegistry;
import org.bson.codecs.pojo.PojoCodecProvider;
import io.smallrye.mutiny.Uni;

import java.util.ArrayList;
import java.util.List;

import static com.mongodb.client.model.Projections.fields;
import static com.mongodb.client.model.Projections.computed;

public static Uni<List<AverageRating>> getAverage(String productId) {
    List<Bson> pipeline = new ArrayList<>();
    Bson match = Aggregates.match(Filters.eq("productId", new ObjectId(productId)));
    Bson group = Aggregates.group("$productId", new BsonField("average", new Document("$avg", "$score")));
    Bson project = Aggregates.project(fields(computed("productId", "$_id"), computed("average", "$average")));
    pipeline.add(match);
    pipeline.add(group);
    pipeline.add(project);

    CodecRegistry pojoCodecRegistry = CodecRegistries.fromRegistries(
            MongoClientSettings.getDefaultCodecRegistry(),
            CodecRegistries.fromProviders(PojoCodecProvider.builder().automatic(true).build())
    );

    return Rating.mongoCollection()
                 .withCodecRegistry(pojoCodecRegistry)
                 .aggregate(pipeline, AverageRating.class)
                 .collect()
                 .asList();
}

In the above example, the project stage specifies the shape of the documents returned by the aggregation pipeline. It is also using the codec registry to map the result to the AverageRating POJO (Plain Old Java Object). Note that the aggregate method is now parameterized with AverageRating.class which indicates the type of class the results should be mapped to.