Add access to internal representation of geo_point doc values (#86800)

Refactors GeoPoint docvalues wrappers to expose the internal SortedNumericDocValues
elastic · May 20, 2022 · 2e992fd · 2e992fd
1 parent 8a1ef1f
commit 2e992fd
Show file tree

Hide file tree

Showing 16 changed files with 129 additions and 265 deletions.
diff --git a/server/src/main/java/org/elasticsearch/common/geo/GeoPoint.java b/server/src/main/java/org/elasticsearch/common/geo/GeoPoint.java
@@ -197,6 +197,18 @@ public String getGeohash() {
         return Geohash.stringEncode(lon, lat);
     }
 
+    /** Return the point in Lucene encoded format used to stored points as doc values */
+    public long getEncoded() {
+        final int latitudeEncoded = GeoEncodingUtils.encodeLatitude(this.lat);
+        final int longitudeEncoded = GeoEncodingUtils.encodeLongitude(this.lon);
+        return (((long) latitudeEncoded) << 32) | (longitudeEncoded & 0xFFFFFFFFL);
+    }
+
+    /** reset the point using Lucene encoded format used to stored points as doc values */
+    public GeoPoint resetFromEncoded(long encoded) {
+        return reset(GeoEncodingUtils.decodeLatitude((int) (encoded >>> 32)), GeoEncodingUtils.decodeLongitude((int) encoded));
+    }
+
     @Override
     public boolean equals(Object o) {
         if (this == o) return true;

diff --git a/server/src/main/java/org/elasticsearch/index/fielddata/FieldData.java b/server/src/main/java/org/elasticsearch/index/fielddata/FieldData.java
@@ -14,7 +14,6 @@
 import org.apache.lucene.index.SortedNumericDocValues;
 import org.apache.lucene.index.SortedSetDocValues;
 import org.apache.lucene.util.BytesRef;
-import org.elasticsearch.common.geo.GeoPoint;
 
 import java.io.IOException;
 import java.util.ArrayList;
@@ -58,27 +57,6 @@ public static SortedNumericDoubleValues emptySortedNumericDoubles() {
         return singleton(emptyNumericDouble());
     }
 
-    public static GeoPointValues emptyGeoPoint() {
-        return new GeoPointValues() {
-            @Override
-            public boolean advanceExact(int doc) throws IOException {
-                return false;
-            }
-
-            @Override
-            public GeoPoint geoPointValue() {
-                throw new UnsupportedOperationException();
-            }
-        };
-    }
-
-    /**
-     * Return a {@link SortedNumericDoubleValues} that doesn't contain any value.
-     */
-    public static MultiGeoPointValues emptyMultiGeoPoints() {
-        return singleton(emptyGeoPoint());
-    }
-
     /**
      * Returns a {@link DocValueBits} representing all documents from <code>values</code> that have a value.
      */
@@ -234,23 +212,12 @@ public static NumericDoubleValues unwrapSingleton(SortedNumericDoubleValues valu
         return null;
     }
 
-    /**
-     * Returns a multi-valued view over the provided {@link GeoPointValues}.
-     */
-    public static MultiGeoPointValues singleton(GeoPointValues values) {
-        return new SingletonMultiGeoPointValues(values);
-    }
-
     /**
      * Returns a single-valued view of the {@link MultiGeoPointValues},
-     * if it was previously wrapped with {@link #singleton(GeoPointValues)},
-     * or null.
+     * if the wrapped {@link SortedNumericDocValues} is a singleton.
      */
     public static GeoPointValues unwrapSingleton(MultiGeoPointValues values) {
-        if (values instanceof SingletonMultiGeoPointValues) {
-            return ((SingletonMultiGeoPointValues) values).getGeoPointValues();
-        }
-        return null;
+        return values.getGeoPointValues();
     }
 
     /**

diff --git a/server/src/main/java/org/elasticsearch/index/fielddata/GeoPointScriptDocValues.java b/server/src/main/java/org/elasticsearch/index/fielddata/GeoPointScriptDocValues.java
@@ -8,20 +8,16 @@
 
 package org.elasticsearch.index.fielddata;
 
-import org.apache.lucene.geo.GeoEncodingUtils;
-import org.elasticsearch.common.geo.GeoPoint;
 import org.elasticsearch.script.GeoPointFieldScript;
 
 import java.util.Arrays;
 
-public final class GeoPointScriptDocValues extends MultiGeoPointValues {
+public final class GeoPointScriptDocValues extends AbstractSortedNumericDocValues {
     private final GeoPointFieldScript script;
-    private final GeoPoint point;
     private int cursor;
 
     GeoPointScriptDocValues(GeoPointFieldScript script) {
         this.script = script;
-        this.point = new GeoPoint();
     }
 
     @Override
@@ -41,10 +37,7 @@ public int docValueCount() {
     }
 
     @Override
-    public GeoPoint nextValue() {
-        final long value = script.values()[cursor++];
-        final int lat = (int) (value >>> 32);
-        final int lon = (int) (value & 0xFFFFFFFF);
-        return point.reset(GeoEncodingUtils.decodeLatitude(lat), GeoEncodingUtils.decodeLongitude(lon));
+    public long nextValue() {
+        return script.values()[cursor++];
     }
 }
diff --git a/server/src/main/java/org/elasticsearch/index/fielddata/GeoPointScriptFieldData.java b/server/src/main/java/org/elasticsearch/index/fielddata/GeoPointScriptFieldData.java
@@ -9,6 +9,7 @@
 package org.elasticsearch.index.fielddata;
 
 import org.apache.lucene.index.LeafReaderContext;
+import org.apache.lucene.index.SortedNumericDocValues;
 import org.apache.lucene.search.SortField;
 import org.elasticsearch.common.util.BigArrays;
 import org.elasticsearch.index.fielddata.plain.AbstractLeafGeoPointFieldData;
@@ -92,7 +93,7 @@ public LeafGeoPointFieldData load(LeafReaderContext context) {
         GeoPointFieldScript script = leafFactory.newInstance(context);
         return new AbstractLeafGeoPointFieldData(toScriptFieldFactory) {
             @Override
-            public MultiGeoPointValues getGeoPointValues() {
+            public SortedNumericDocValues getSortedNumericDocValues() {
                 return new GeoPointScriptDocValues(script);
             }
 

diff --git a/server/src/main/java/org/elasticsearch/index/fielddata/GeoPointValues.java b/server/src/main/java/org/elasticsearch/index/fielddata/GeoPointValues.java
@@ -8,25 +8,36 @@
 
 package org.elasticsearch.index.fielddata;
 
+import org.apache.lucene.index.NumericDocValues;
 import org.elasticsearch.common.geo.GeoPoint;
 
 import java.io.IOException;
 
 /**
  * Per-document geo-point values.
  */
-public abstract class GeoPointValues {
+public final class GeoPointValues {
+
+    private final GeoPoint point = new GeoPoint();
+    private final NumericDocValues values;
+
+    GeoPointValues(NumericDocValues values) {
+        this.values = values;
+    }
 
     /**
      * Advance this instance to the given document id
      * @return true if there is a value for this document
      */
-    public abstract boolean advanceExact(int doc) throws IOException;
+    public boolean advanceExact(int doc) throws IOException {
+        return values.advanceExact(doc);
+    }
 
     /**
      * Get the {@link GeoPoint} associated with the current document.
      * The returned {@link GeoPoint} might be reused across calls.
      */
-    public abstract GeoPoint geoPointValue();
-
+    public GeoPoint geoPointValue() throws IOException {
+        return point.resetFromEncoded(values.longValue());
+    }
 }
diff --git a/server/src/main/java/org/elasticsearch/index/fielddata/LeafGeoPointFieldData.java b/server/src/main/java/org/elasticsearch/index/fielddata/LeafGeoPointFieldData.java
@@ -7,14 +7,25 @@
  */
 package org.elasticsearch.index.fielddata;
 
+import org.apache.lucene.index.SortedNumericDocValues;
+
 /**
  * {@link LeafFieldData} specialization for geo points.
  */
-public interface LeafGeoPointFieldData extends LeafFieldData {
+public abstract class LeafGeoPointFieldData implements LeafFieldData {
 
     /**
      * Return geo-point values.
      */
-    MultiGeoPointValues getGeoPointValues();
+    public final MultiGeoPointValues getGeoPointValues() {
+        return new MultiGeoPointValues(getSortedNumericDocValues());
+    }
+
+    /**
+     * Return the internal representation of geo_point doc values as a {@link SortedNumericDocValues}.
+     * A point is encoded as a long that can be decoded by using
+     * {@link org.elasticsearch.common.geo.GeoPoint#resetFromEncoded(long)}
+     */
+    public abstract SortedNumericDocValues getSortedNumericDocValues();
 
 }
diff --git a/server/src/main/java/org/elasticsearch/index/fielddata/MultiGeoPointValues.java b/server/src/main/java/org/elasticsearch/index/fielddata/MultiGeoPointValues.java
@@ -7,6 +7,9 @@
  */
 package org.elasticsearch.index.fielddata;
 
+import org.apache.lucene.index.DocValues;
+import org.apache.lucene.index.NumericDocValues;
+import org.apache.lucene.index.SortedNumericDocValues;
 import org.elasticsearch.common.geo.GeoPoint;
 
 import java.io.IOException;
@@ -26,23 +29,32 @@
  * The set of values associated with a document might contain duplicates and
  * comes in a non-specified order.
  */
-public abstract class MultiGeoPointValues {
+public final class MultiGeoPointValues {
+
+    private final GeoPoint point = new GeoPoint();
+    private final SortedNumericDocValues numericValues;
 
     /**
      * Creates a new {@link MultiGeoPointValues} instance
      */
-    protected MultiGeoPointValues() {}
+    public MultiGeoPointValues(SortedNumericDocValues numericValues) {
+        this.numericValues = numericValues;
+    }
 
     /**
      * Advance this instance to the given document id
      * @return true if there is a value for this document
      */
-    public abstract boolean advanceExact(int doc) throws IOException;
+    public boolean advanceExact(int doc) throws IOException {
+        return numericValues.advanceExact(doc);
+    }
 
     /**
      * Return the number of geo points the current document has.
      */
-    public abstract int docValueCount();
+    public int docValueCount() {
+        return numericValues.docValueCount();
+    }
 
     /**
      * Return the next value associated with the current document. This must not be
@@ -52,6 +64,16 @@ protected MultiGeoPointValues() {}
      *
      * @return the next value for the current docID set to {@link #advanceExact(int)}.
      */
-    public abstract GeoPoint nextValue() throws IOException;
+    public GeoPoint nextValue() throws IOException {
+        return point.resetFromEncoded(numericValues.nextValue());
+    }
+
+    /**
+     * Returns a single-valued view of the {@link MultiGeoPointValues} if possible, otherwise null.
+     */
+    GeoPointValues getGeoPointValues() {
+        final NumericDocValues singleton = DocValues.unwrapSingleton(numericValues);
+        return singleton != null ? new GeoPointValues(singleton) : null;
+    }
 
 }
diff --git a/server/src/main/java/org/elasticsearch/index/fielddata/SingletonMultiGeoPointValues.java b/server/src/main/java/org/elasticsearch/index/fielddata/SingletonMultiGeoPointValues.java
diff --git a/.../src/main/java/org/elasticsearch/index/fielddata/plain/AbstractLeafGeoPointFieldData.java b/.../src/main/java/org/elasticsearch/index/fielddata/plain/AbstractLeafGeoPointFieldData.java
@@ -7,18 +7,14 @@
  */
 package org.elasticsearch.index.fielddata.plain;
 
-import org.apache.lucene.util.Accountable;
 import org.elasticsearch.index.fielddata.FieldData;
 import org.elasticsearch.index.fielddata.LeafGeoPointFieldData;
 import org.elasticsearch.index.fielddata.MultiGeoPointValues;
 import org.elasticsearch.index.fielddata.SortedBinaryDocValues;
 import org.elasticsearch.script.field.DocValuesScriptFieldFactory;
 import org.elasticsearch.script.field.ToScriptFieldFactory;
 
-import java.util.Collection;
-import java.util.Collections;
-
-public abstract class AbstractLeafGeoPointFieldData implements LeafGeoPointFieldData {
+public abstract class AbstractLeafGeoPointFieldData extends LeafGeoPointFieldData {
 
     protected final ToScriptFieldFactory<MultiGeoPointValues> toScriptFieldFactory;
 
@@ -35,27 +31,4 @@ public final SortedBinaryDocValues getBytesValues() {
     public DocValuesScriptFieldFactory getScriptFieldFactory(String name) {
         return toScriptFieldFactory.getScriptFieldFactory(getGeoPointValues(), name);
     }
-
-    public static LeafGeoPointFieldData empty(final int maxDoc, ToScriptFieldFactory<MultiGeoPointValues> toScriptFieldFactory) {
-        return new AbstractLeafGeoPointFieldData(toScriptFieldFactory) {
-
-            @Override
-            public long ramBytesUsed() {
-                return 0;
-            }
-
-            @Override
-            public Collection<Accountable> getChildResources() {
-                return Collections.emptyList();
-            }
-
-            @Override
-            public void close() {}
-
-            @Override
-            public MultiGeoPointValues getGeoPointValues() {
-                return FieldData.emptyMultiGeoPoints();
-            }
-        };
-    }
 }