Examples of org.elasticsearch.search.aggregations.support.OrderPath

Package org.elasticsearch.search.aggregations.support

Examples of org.elasticsearch.search.aggregations.support.OrderPath

org.elasticsearch.search.aggregations.support.OrderPath
A path that can be used to sort/order buckets (in some multi-bucket aggregations, eg terms & histogram) based on sub-aggregations. The path may point to either a single-bucket aggregation or a metrics aggregation. If the path points to a single-bucket aggregation, the sort will be applied based on the {@code doc_count} of the bucket. If thispath points to a metrics aggregation, if it's a single-value metrics (eg. avg, max, min, etc..) the sort will be applied on that single value. If it points to a multi-value metrics, the path should point out what metric should be the sort-by value.
The path has the following form:

{@code ['>'*]['.']}

Examples:
- {@code agg1>agg2>agg3} - where agg1, agg2 and agg3 are all single-bucket aggs (eg filter, nested, missing, etc..). Inthis case, the order will be based on the number of documents under {@code agg3}.
- {@code agg1>agg2>agg3} - where agg1 and agg2 are both single-bucket aggs and agg3 is a single-value metrics agg (eg avg, max, min, etc..).In this case, the order will be based on the value of {@code agg3}.
- {@code agg1>agg2>agg3.avg} - where agg1 and agg2 are both single-bucket aggs and agg3 is a multi-value metrics agg (eg stats, extended_stats, etc...).In this case, the order will be based on the avg value of {@code agg3}.

        this.bucketCountThresholds = bucketCountThresholds;
        this.order = InternalOrder.validate(order, this);
        this.subAggCollectMode = subAggCollectMode;
        // Don't defer any child agg if we are dependent on it for pruning results
        if (order instanceof Aggregation){
            OrderPath path = ((Aggregation) order).path();
            aggsUsedForSorting.add(path.resolveTopmostAggregator(this));
        } else if (order instanceof CompoundOrder) {
            CompoundOrder compoundOrder = (CompoundOrder) order;
            for (Terms.Order orderElement : compoundOrder.orderElements()) {
                if (orderElement instanceof Aggregation) {
                    OrderPath path = ((Aggregation) orderElement).path();
                    aggsUsedForSorting.add(path.resolveTopmostAggregator(this));
                }
            }
        }
    }

View Full Code Here

            }
            return order;
        } else if (!(order instanceof Aggregation)) {
            return order;
        }
        OrderPath path = ((Aggregation) order).path();
        path.validate(termsAggregator);
        return order;
    }

View Full Code Here

            // to avoid constructing buckets for ordering purposes (we can potentially have a lot of buckets and building
            // them will cause loads of redundant object constructions). The "special" comparators here will fetch the
            // sub aggregation values directly from the sub aggregators bypassing bucket creation. Note that the comparator
            // attached to the order will still be used in the reduce phase of the Aggregation.


            OrderPath path = path();
            final Aggregator aggregator = path.resolveAggregator(termsAggregator);
            final String key = path.tokens[path.tokens.length - 1].key;


            if (aggregator instanceof SingleBucketAggregator) {
                assert key == null : "this should be picked up before the aggregation is executed - on validate";
                return new Comparator<Terms.Bucket>() {

View Full Code Here

        public static void writeOrder(Terms.Order order, StreamOutput out) throws IOException {
            if (order instanceof Aggregation) {
                out.writeByte(order.id());
                Aggregation aggregationOrder = (Aggregation) order;
                out.writeBoolean(((MultiBucketsAggregation.Bucket.SubAggregationComparator) aggregationOrder.comparator).asc());
                OrderPath path = ((Aggregation) order).path();
                if (out.getVersion().onOrAfter(Version.V_1_1_0)) {
                    out.writeString(path.toString());
                } else {
                    // prev versions only supported sorting on a single level -> a single token;
                    OrderPath.Token token = path.lastToken();
                    out.writeString(token.name);
                    boolean hasValueName = token.key != null;
                    out.writeBoolean(hasValueName);
                    if (hasValueName) {
                        out.writeString(token.key);

View Full Code Here

TOP

Related Classes of org.elasticsearch.search.aggregations.support.OrderPath

org.elasticsearch.search.aggregations.Aggregation

org.elasticsearch.search.aggregations.Aggregator

org.elasticsearch.search.aggregations.bucket.terms.InternalOrder

org.elasticsearch.search.aggregations.bucket.terms.InternalOrder$Aggregation

org.elasticsearch.search.aggregations.bucket.terms.InternalOrder$Streams

org.elasticsearch.search.aggregations.bucket.terms.TermsAggregator

org.elasticsearch.search.aggregations.HasAggregations

org.elasticsearch.search.aggregations.AggregationExecutionException

org.elasticsearch.ElasticsearchIllegalArgumentException

All source code are property of their respective owners. Java is a trademark of Sun Microsystems, Inc and owned by ORACLE Inc. Contact coftware#gmail.com.