MongoDB $aggregate $push multiple fields in Java Spring Data
Solution 1:
You can directly pass the BasicDbObject to any of the aggregation pipeline stage.
Aggregation agg = newAggregation(
group("title").
push(new BasicDBObject
("version", "$version").append
("author", "$author").append
("dateAdded", "$dateAdded")).as("versions"));
Solution 2:
Note: MongoDB versión 2.6 - 3.0 (compatible Java driver)
I've extended org.springframework.data.mongodb.core.aggregation.AggregationOperation
class to implement custom toDBObject
method:
public class GenericAggregationOperation implements AggregationOperation {
private String operator;
private DBObject query;
/**
* Default constructor.
*
* @param operator MongoDB operator ($group, $sort, $project, etc..)
* @param query MongoDB aggregation query step string
*/
public GenericAggregationOperation(String operator, String query) {
this(operator, (DBObject) JSON.parse(query));
}
/**
* Default constructor.
*
* @param operator MongoDB operator ($group, $sort, $project, etc..)
* @param query MongoDB aggregation query step DBObject
*/
public GenericAggregationOperation(String operator, DBObject query) {
this.operator = operator;
this.query = query;
}
@Override
public DBObject toDBObject(AggregationOperationContext context) {
return new BasicDBObject(operator, query);
}
}
In your case, it would be:
List<AggregationOperation> list = new ArrayList<AggregationOperation>();
lista.add(new GenericAggregationOperation("$group", "{ \"_id\": \"$title\", \"versions\": { \"$push\": { \"version\":\"$version\", \"author\":\"$author\", \"dateAdded\":\"$dateAdded\"}} }"));
TypedAggregation<EpisodeIndexDto> agg = Aggregation.newAggregation(
YourClassTitleVersion.class, list);
mongoOperations.aggregate(agg, YourClassTitleVersion.class,
YourClassTitleVersion.class).getMappedResults();
Hope this helps you.
Solution 3:
In the new version spring-data-mongodb:2.x.x the AggregationOperation need to return Document instead of DBObject, so the updated class will be:
import com.mongodb.BasicDBObject;
import com.mongodb.DBObject;
import org.bson.Document;
import org.springframework.data.mongodb.core.aggregation.AggregationOperation;
import org.springframework.data.mongodb.core.aggregation.AggregationOperationContext;
public class GenericAggregationOperation implements AggregationOperation {
private String operator;
private DBObject query;
public GenericAggregationOperation(String operator, DBObject query) {
this.operator = operator;
this.query = query;
}
public GenericAggregationOperation(String operator, String query) {
this(operator, BasicDBObject.parse(query));
}
@Override
public Document toDocument(AggregationOperationContext context) {
return new Document(operator, query);
}
}
Plus to make it easier to be used I'll add utility interface (java 8+, for java 7 or lower you can convert this to class utils instead):
import com.mongodb.DBObject;
import org.springframework.data.mongodb.core.aggregation.AggregationOperation;
public interface GenericAggregationUtils {
static AggregationOperation aggregate(String operation, String query) {
return new GenericAggregationOperation(operation, query);
}
static AggregationOperation aggregate(String operation, DBObject query) {
return new GenericAggregationOperation(operation, query);
}
}
And then we can static import the interface into our class:
import static com.example.mongodb.aggregation.GenericAggregationUtils.*;
And use it in the aggregation pipeline together with other spring data AggregationOperation like this:
Aggregation aggregation = newAggregation(YourDocCollection.class,
aggregate("$group", "{ \"_id\": \"$title\", \"versions\": { \"$push\": { \"version\":\"$version\", \"author\":\"$author\", \"dateAdded\":\"$dateAdded\"}} }"),
sort(Sort.Direction.ASC, "title"),
...
);