Adds support for geo-bounds filtering in geogrid aggregations

It is fairly common to filter the geo point candidates in
geohash_grid and geotile_grid aggregations according to some
viewable bounding box. This change introduces the option of
specifying this filter directly in the tiling aggregation.

Author: talevy

docs/reference/aggregations/bucket/geohashgrid-aggregation.asciidoc

@@ -191,6 +191,62 @@ var bbox = geohash.decode_bbox('u17');
 --------------------------------------------------
 // NOTCONSOLE
 
+==== Requests with additional bounding box filtering
+
+The `geohash_grid` aggregation supports an optional `bounds` parameter
+that restricts the points considered to those that fall within the
+bounds provided. The `bounds` parameter accepts the bounding box in
+all the same <<query-dsl-geo-bounding-box-query-accepted-formats,accepted formats>> of the
+bounds specified in the Geo Bounding Box Query. This bounding box can be used with or
+without an additional `geo_bounding_box` query filtering the points prior to aggregating.
+It is an independent bounding box that can intersect with, be equal to, or be disjoint
+to any additional `geo_bounding_box` queries defined in the context of the aggregation.
+
+[source,console,id=geohashgrid-aggregation-with-bounds]
+--------------------------------------------------
+POST /museums/_search?size=0
+{
+    "aggregations" : {
+        "tiles-in-bounds" : {
+            "geohash_grid" : {
+                "field" : "location",
+                "precision" : 8,
+                "bounds": {
+                  "top_left" : "53.4375, 4.21875",
+                  "bottom_right" : "52.03125, 5.625"
+                }
+            }
+        }
+    }
+}
+--------------------------------------------------
+// TEST[continued]
+
+[source,console-result]
+--------------------------------------------------
+{
+    ...
+    "aggregations" : {
+        "tiles-in-bounds" : {
+           "buckets" : [
+               {
+                 "key" : "u173zy3j",
+                 "doc_count" : 1
+               },
+               {
+                 "key" : "u173zvfz",
+                 "doc_count" : 1
+               },
+               {
+                 "key" : "u173zt90",
+                 "doc_count" : 1
+               }
+           ]
+        }
+    }
+}
+--------------------------------------------------
+// TESTRESPONSE[s/\.\.\./"took": $body.took,"_shards": $body._shards,"hits":$body.hits,"timed_out":false,/]
 
 ==== Cell dimensions at the equator
 The table below shows the metric dimensions for cells covered by various string lengths of geohash.
@@ -230,6 +286,8 @@ precision::     Optional. The string length of the geohashes used to define
                 to precision levels higher than the supported 12 levels,
                 (e.g. for distances <5.6cm) the value is rejected.
 
+bounds:         Optional. The bounding box to filter the points in the bucket.
+
 size::          Optional. The maximum number of geohash buckets to return
                 (defaults to 10,000). When results are trimmed, buckets are
                 prioritised based on the volumes of documents they contain.

docs/reference/aggregations/bucket/geotilegrid-aggregation.asciidoc

@@ -162,6 +162,62 @@ POST /museums/_search?size=0
 --------------------------------------------------
 // TESTRESPONSE[s/\.\.\./"took": $body.took,"_shards": $body._shards,"hits":$body.hits,"timed_out":false,/]
 
+==== Requests with additional bounding box filtering
+
+The `geotile_grid` aggregation supports an optional `bounds` parameter
+that restricts the points considered to those that fall within the
+bounds provided. The `bounds` parameter accepts the bounding box in
+all the same <<query-dsl-geo-bounding-box-query-accepted-formats,accepted formats>> of the
+bounds specified in the Geo Bounding Box Query. This bounding box can be used with or
+without an additional `geo_bounding_box` query filtering the points prior to aggregating.
+It is an independent bounding box that can intersect with, be equal to, or be disjoint
+to any additional `geo_bounding_box` queries defined in the context of the aggregation.
+
+[source,console,id=geotilegrid-aggregation-with-bounds]
+--------------------------------------------------
+POST /museums/_search?size=0
+{
+    "aggregations" : {
+        "tiles-in-bounds" : {
+            "geotile_grid" : {
+                "field" : "location",
+                "precision" : 22,
+                "bounds": {
+                  "top_left" : "52.4, 4.9",
+                  "bottom_right" : "52.3, 5.0"
+                }
+            }
+        }
+    }
+}
+--------------------------------------------------
+// TEST[continued]
+
+[source,console-result]
+--------------------------------------------------
+{
+    ...
+    "aggregations" : {
+        "tiles-in-bounds" : {
+           "buckets" : [
+               {
+                 "key" : "22/2154412/1378379",
+                 "doc_count" : 1
+               },
+               {
+                 "key" : "22/2154385/1378332",
+                 "doc_count" : 1
+               },
+               {
+                 "key" : "22/2154259/1378425",
+                 "doc_count" : 1
+               }
+           ]
+        }
+    }
+}
+--------------------------------------------------
+// TESTRESPONSE[s/\.\.\./"took": $body.took,"_shards": $body._shards,"hits":$body.hits,"timed_out":false,/]
 
 ==== Options
 
@@ -172,6 +228,8 @@ precision::     Optional. The integer zoom of the key used to define
                 cells/buckets in the results. Defaults to 7.
                 Values outside of [0,29] will be rejected.
 
+bounds:         Optional. The bounding box to filter the points in the bucket.
+
 size::          Optional. The maximum number of geohash buckets to return
                 (defaults to 10,000). When results are trimmed, buckets are
                 prioritised based on the volumes of documents they contain.

docs/reference/query-dsl/geo-bounding-box-query.asciidoc

@@ -84,6 +84,7 @@ be executed in memory or indexed. See <<geo-bbox-type,Type>> below for further d
 Default is `memory`.
 |=======================================================================
 
+[[query-dsl-geo-bounding-box-query-accepted-formats]]
 [float]
 ==== Accepted Formats
 

server/src/main/java/org/elasticsearch/common/geo/GeoBoundingBox.java

@@ -68,6 +68,11 @@ public GeoBoundingBox(StreamInput input) throws IOException {
         this.bottomRight = input.readGeoPoint();
     }
 
+    public boolean isUnbounded() {
+        return Double.isNaN(topLeft.lon()) || Double.isNaN(topLeft.lat())
+            || Double.isNaN(bottomRight.lon()) || Double.isNaN(bottomRight.lat());
+    }
+
     public GeoPoint topLeft() {
         return topLeft;
     }
@@ -120,6 +125,26 @@ public XContentBuilder toXContentFragment(XContentBuilder builder, boolean build
         return builder;
     }
 
+    /**
+     * If the bounding box crosses the date-line (left greater-than right) then the
+     * longitude of the point need only to be higher than the left or lower
+     * than the right. Otherwise, it must be both.
+     *
+     * @param lon the longitude of the point
+     * @param lat the latitude of the point
+     * @return whether the point (lon, lat) is in the specified bounding box
+     */
+    public boolean pointInBounds(double lon, double lat) {
+        if (lat >= bottom() && lat <= top()) {
+            if (left() <= right()) {
+                return lon >= left() && lon <= right();
+            } else {
+                return lon >= left() || lon <= right();
+            }
+        }
+        return false;
+    }
+
     @Override
     public void writeTo(StreamOutput out) throws IOException {
         out.writeGeoPoint(topLeft);

server/src/main/java/org/elasticsearch/common/geo/GeoUtils.java

@@ -93,6 +93,7 @@ public static boolean isValidLongitude(double longitude) {
         return true;
     }
 
+
     /**
      * Calculate the width (in meters) of geohash cells at a specific level
      * @param level geohash level must be greater or equal to zero

server/src/main/java/org/elasticsearch/search/aggregations/bucket/composite/GeoTileGridValuesSourceBuilder.java

@@ -19,7 +19,10 @@
 
 package org.elasticsearch.search.aggregations.bucket.composite;
 
+import org.elasticsearch.Version;
 import org.elasticsearch.common.ParseField;
+import org.elasticsearch.common.geo.GeoBoundingBox;
+import org.elasticsearch.common.geo.GeoPoint;
 import org.elasticsearch.common.io.stream.StreamInput;
 import org.elasticsearch.common.io.stream.StreamOutput;
 import org.elasticsearch.common.xcontent.ObjectParser;
@@ -45,6 +48,8 @@ public class GeoTileGridValuesSourceBuilder extends CompositeValuesSourceBuilder
     static {
         PARSER = new ObjectParser<>(GeoTileGridValuesSourceBuilder.TYPE);
         PARSER.declareInt(GeoTileGridValuesSourceBuilder::precision, new ParseField("precision"));
+        PARSER.declareField(((p, builder, context) -> builder.geoBoundingBox(GeoBoundingBox.parseBoundingBox(p))),
+            GeoBoundingBox.BOUNDS_FIELD, ObjectParser.ValueType.OBJECT);
         CompositeValuesSourceParserHelper.declareValuesSourceFields(PARSER, ValueType.NUMERIC);
     }
 
@@ -53,6 +58,7 @@ static GeoTileGridValuesSourceBuilder parse(String name, XContentParser parser)
     }
 
     private int precision = GeoTileGridAggregationBuilder.DEFAULT_PRECISION;
+    private GeoBoundingBox geoBoundingBox = new GeoBoundingBox(new GeoPoint(Double.NaN, Double.NaN), new GeoPoint(Double.NaN, Double.NaN));
 
     GeoTileGridValuesSourceBuilder(String name) {
         super(name);
@@ -61,13 +67,21 @@ static GeoTileGridValuesSourceBuilder parse(String name, XContentParser parser)
     GeoTileGridValuesSourceBuilder(StreamInput in) throws IOException {
         super(in);
         this.precision = in.readInt();
+        if (in.getVersion().onOrAfter(Version.V_8_0_0)) {
+            this.geoBoundingBox = new GeoBoundingBox(in);
+        }
     }
 
     public GeoTileGridValuesSourceBuilder precision(int precision) {
         this.precision = GeoTileUtils.checkPrecisionRange(precision);
         return this;
     }
 
+    public GeoTileGridValuesSourceBuilder geoBoundingBox(GeoBoundingBox geoBoundingBox) {
+        this.geoBoundingBox = geoBoundingBox;
+        return this;
+    }
+
     @Override
     public GeoTileGridValuesSourceBuilder format(String format) {
         throw new IllegalArgumentException("[format] is not supported for [" + TYPE + "]");
@@ -76,11 +90,17 @@ public GeoTileGridValuesSourceBuilder format(String format) {
     @Override
     protected void innerWriteTo(StreamOutput out) throws IOException {
         out.writeInt(precision);
+        if (out.getVersion().onOrAfter(Version.V_8_0_0)) {
+            geoBoundingBox.writeTo(out);
+        }
     }
 
     @Override
     protected void doXContentBody(XContentBuilder builder, Params params) throws IOException {
         builder.field("precision", precision);
+        if (geoBoundingBox.isUnbounded() == false) {
+            geoBoundingBox.toXContent(builder, params);
+        }
     }
 
     @Override
@@ -90,7 +110,7 @@ String type() {
 
     @Override
     public int hashCode() {
-        return Objects.hash(super.hashCode(), precision);
+        return Objects.hash(super.hashCode(), precision, geoBoundingBox);
     }
 
     @Override
@@ -99,7 +119,8 @@ public boolean equals(Object obj) {
         if (obj == null || getClass() != obj.getClass()) return false;
         if (super.equals(obj) == false) return false;
         GeoTileGridValuesSourceBuilder other = (GeoTileGridValuesSourceBuilder) obj;
-        return precision == other.precision;
+        return Objects.equals(precision,other.precision)
+            && Objects.equals(geoBoundingBox, other.geoBoundingBox);
     }
 
     @Override
@@ -112,7 +133,7 @@ protected CompositeValuesSourceConfig innerBuild(QueryShardContext queryShardCon
             ValuesSource.GeoPoint geoPoint = (ValuesSource.GeoPoint) orig;
             // is specified in the builder.
             final MappedFieldType fieldType = config.fieldContext() != null ? config.fieldContext().fieldType() : null;
-            CellIdSource cellIdSource = new CellIdSource(geoPoint, precision, GeoTileUtils::longEncode);
+            CellIdSource cellIdSource = new CellIdSource(geoPoint, precision, geoBoundingBox, GeoTileUtils::longEncode);
             return new CompositeValuesSourceConfig(name, fieldType, cellIdSource, DocValueFormat.GEOTILE, order(),
                 missingBucket(), script() != null);
         } else {

server/src/main/java/org/elasticsearch/search/aggregations/bucket/geogrid/CellIdSource.java

@@ -20,6 +20,7 @@
 
 import org.apache.lucene.index.LeafReaderContext;
 import org.apache.lucene.index.SortedNumericDocValues;
+import org.elasticsearch.common.geo.GeoBoundingBox;
 import org.elasticsearch.index.fielddata.AbstractSortingNumericDocValues;
 import org.elasticsearch.index.fielddata.MultiGeoPointValues;
 import org.elasticsearch.index.fielddata.SortedBinaryDocValues;
@@ -36,11 +37,13 @@ public class CellIdSource extends ValuesSource.Numeric {
     private final ValuesSource.GeoPoint valuesSource;
     private final int precision;
     private final GeoPointLongEncoder encoder;
+    private final GeoBoundingBox geoBoundingBox;
 
-    public CellIdSource(GeoPoint valuesSource, int precision, GeoPointLongEncoder encoder) {
+    public CellIdSource(GeoPoint valuesSource,int precision, GeoBoundingBox geoBoundingBox, GeoPointLongEncoder encoder) {
         this.valuesSource = valuesSource;
         //different GeoPoints could map to the same or different hashing cells.
         this.precision = precision;
+        this.geoBoundingBox = geoBoundingBox;
         this.encoder = encoder;
     }
 
@@ -55,7 +58,7 @@ public boolean isFloatingPoint() {
 
     @Override
     public SortedNumericDocValues longValues(LeafReaderContext ctx) {
-        return new CellValues(valuesSource.geoPointValues(ctx), precision, encoder);
+        return new CellValues(valuesSource.geoPointValues(ctx), precision, geoBoundingBox, encoder);
     }
 
     @Override
@@ -81,21 +84,28 @@ private static class CellValues extends AbstractSortingNumericDocValues {
         private MultiGeoPointValues geoValues;
         private int precision;
         private GeoPointLongEncoder encoder;
+        private GeoBoundingBox geoBoundingBox;
 
-        protected CellValues(MultiGeoPointValues geoValues, int precision, GeoPointLongEncoder encoder) {
+        protected CellValues(MultiGeoPointValues geoValues, int precision, GeoBoundingBox geoBoundingBox, GeoPointLongEncoder encoder) {
             this.geoValues = geoValues;
             this.precision = precision;
             this.encoder = encoder;
+            this.geoBoundingBox = geoBoundingBox;
         }
 
         @Override
         public boolean advanceExact(int docId) throws IOException {
             if (geoValues.advanceExact(docId)) {
-                resize(geoValues.docValueCount());
-                for (int i = 0; i < docValueCount(); ++i) {
+                int docValueCount = geoValues.docValueCount();
+                resize(docValueCount);
+                int j = 0;
+                for (int i = 0; i < docValueCount; i++) {
                     org.elasticsearch.common.geo.GeoPoint target = geoValues.nextValue();
-                    values[i] = encoder.encode(target.getLon(), target.getLat(), precision);
+                    if (geoBoundingBox.isUnbounded() || geoBoundingBox.pointInBounds(target.getLon(), target.getLat())) {
+                        values[j++] = encoder.encode(target.getLon(), target.getLat(), precision);
+                    }
                 }
+                resize(j);
                 sort();
                 return true;
             } else {