001/*
002 * Licensed to the Apache Software Foundation (ASF) under one or more
003 * contributor license agreements.  See the NOTICE file distributed with
004 * this work for additional information regarding copyright ownership.
005 * The ASF licenses this file to You under the Apache License, Version 2.0
006 * (the "License"); you may not use this file except in compliance with
007 * the License.  You may obtain a copy of the License at
008 *
009 *     http://www.apache.org/licenses/LICENSE-2.0
010 *
011 * Unless required by applicable law or agreed to in writing, software
012 * distributed under the License is distributed on an "AS IS" BASIS,
013 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
014 * See the License for the specific language governing permissions and
015 * limitations under the License.
016 */
017package org.apache.lucene.demo.facet;
018
019import java.io.Closeable;
020import java.io.IOException;
021import java.text.ParseException;
022import org.apache.lucene.analysis.core.WhitespaceAnalyzer;
023import org.apache.lucene.document.Document;
024import org.apache.lucene.document.DoublePoint;
025import org.apache.lucene.document.NumericDocValuesField;
026import org.apache.lucene.expressions.Expression;
027import org.apache.lucene.expressions.SimpleBindings;
028import org.apache.lucene.expressions.js.JavascriptCompiler;
029import org.apache.lucene.facet.DrillDownQuery;
030import org.apache.lucene.facet.DrillSideways;
031import org.apache.lucene.facet.FacetResult;
032import org.apache.lucene.facet.Facets;
033import org.apache.lucene.facet.FacetsCollector;
034import org.apache.lucene.facet.FacetsConfig;
035import org.apache.lucene.facet.range.DoubleRange;
036import org.apache.lucene.facet.range.DoubleRangeFacetCounts;
037import org.apache.lucene.facet.taxonomy.TaxonomyReader;
038import org.apache.lucene.index.DirectoryReader;
039import org.apache.lucene.index.IndexWriter;
040import org.apache.lucene.index.IndexWriterConfig;
041import org.apache.lucene.index.IndexWriterConfig.OpenMode;
042import org.apache.lucene.search.BooleanClause;
043import org.apache.lucene.search.BooleanQuery;
044import org.apache.lucene.search.DoubleValuesSource;
045import org.apache.lucene.search.IndexSearcher;
046import org.apache.lucene.search.MatchAllDocsQuery;
047import org.apache.lucene.search.Query;
048import org.apache.lucene.search.TopDocs;
049import org.apache.lucene.store.ByteBuffersDirectory;
050import org.apache.lucene.store.Directory;
051
052/**
053 * Shows simple usage of dynamic range faceting, using the expressions module to calculate distance.
054 */
055public class DistanceFacetsExample implements Closeable {
056
057  final DoubleRange ONE_KM = new DoubleRange("< 1 km", 0.0, true, 1.0, false);
058  final DoubleRange TWO_KM = new DoubleRange("< 2 km", 0.0, true, 2.0, false);
059  final DoubleRange FIVE_KM = new DoubleRange("< 5 km", 0.0, true, 5.0, false);
060  final DoubleRange TEN_KM = new DoubleRange("< 10 km", 0.0, true, 10.0, false);
061
062  private final Directory indexDir = new ByteBuffersDirectory();
063  private IndexSearcher searcher;
064  private final FacetsConfig config = new FacetsConfig();
065
066  /** The "home" latitude. */
067  public static final double ORIGIN_LATITUDE = 40.7143528;
068
069  /** The "home" longitude. */
070  public static final double ORIGIN_LONGITUDE = -74.0059731;
071
072  /**
073   * Mean radius of the Earth in KM
074   *
075   * <p>NOTE: this is approximate, because the earth is a bit wider at the equator than the poles.
076   * See http://en.wikipedia.org/wiki/Earth_radius
077   */
078  // see http://earth-info.nga.mil/GandG/publications/tr8350.2/wgs84fin.pdf
079  public static final double EARTH_RADIUS_KM = 6_371.0087714;
080
081  /** Empty constructor */
082  public DistanceFacetsExample() {}
083
084  /** Build the example index. */
085  public void index() throws IOException {
086    IndexWriter writer =
087        new IndexWriter(
088            indexDir, new IndexWriterConfig(new WhitespaceAnalyzer()).setOpenMode(OpenMode.CREATE));
089
090    // TODO: we could index in radians instead ... saves all the conversions in getBoundingBoxFilter
091
092    // Add documents with latitude/longitude location:
093    // we index these both as DoublePoints (for bounding box/ranges) and as NumericDocValuesFields
094    // (for scoring)
095    Document doc = new Document();
096    doc.add(new DoublePoint("latitude", 40.759011));
097    doc.add(new NumericDocValuesField("latitude", Double.doubleToRawLongBits(40.759011)));
098    doc.add(new DoublePoint("longitude", -73.9844722));
099    doc.add(new NumericDocValuesField("longitude", Double.doubleToRawLongBits(-73.9844722)));
100    writer.addDocument(doc);
101
102    doc = new Document();
103    doc.add(new DoublePoint("latitude", 40.718266));
104    doc.add(new NumericDocValuesField("latitude", Double.doubleToRawLongBits(40.718266)));
105    doc.add(new DoublePoint("longitude", -74.007819));
106    doc.add(new NumericDocValuesField("longitude", Double.doubleToRawLongBits(-74.007819)));
107    writer.addDocument(doc);
108
109    doc = new Document();
110    doc.add(new DoublePoint("latitude", 40.7051157));
111    doc.add(new NumericDocValuesField("latitude", Double.doubleToRawLongBits(40.7051157)));
112    doc.add(new DoublePoint("longitude", -74.0088305));
113    doc.add(new NumericDocValuesField("longitude", Double.doubleToRawLongBits(-74.0088305)));
114    writer.addDocument(doc);
115
116    // Open near-real-time searcher
117    searcher = new IndexSearcher(DirectoryReader.open(writer));
118    writer.close();
119  }
120
121  // TODO: Would be nice to augment this example with documents containing multiple "locations",
122  // adding the ability to compute distance facets for the multi-valued case (see LUCENE-10245)
123  private DoubleValuesSource getDistanceValueSource() {
124    Expression distance;
125    try {
126      distance =
127          JavascriptCompiler.compile(
128              "haversin(" + ORIGIN_LATITUDE + "," + ORIGIN_LONGITUDE + ",latitude,longitude)");
129    } catch (ParseException pe) {
130      // Should not happen
131      throw new RuntimeException(pe);
132    }
133    SimpleBindings bindings = new SimpleBindings();
134    bindings.add("latitude", DoubleValuesSource.fromDoubleField("latitude"));
135    bindings.add("longitude", DoubleValuesSource.fromDoubleField("longitude"));
136
137    return distance.getDoubleValuesSource(bindings);
138  }
139
140  /**
141   * Given a latitude and longitude (in degrees) and the maximum great circle (surface of the earth)
142   * distance, returns a simple Filter bounding box to "fast match" candidates.
143   */
144  public static Query getBoundingBoxQuery(
145      double originLat, double originLng, double maxDistanceKM) {
146
147    // Basic bounding box geo math from
148    // http://JanMatuschek.de/LatitudeLongitudeBoundingCoordinates,
149    // licensed under creative commons 3.0:
150    // http://creativecommons.org/licenses/by/3.0
151
152    // TODO: maybe switch to recursive prefix tree instead
153    // (in lucene/spatial)?  It should be more efficient
154    // since it's a 2D trie...
155
156    // Degrees -> Radians:
157    double originLatRadians = Math.toRadians(originLat);
158    double originLngRadians = Math.toRadians(originLng);
159
160    double angle = maxDistanceKM / EARTH_RADIUS_KM;
161
162    double minLat = originLatRadians - angle;
163    double maxLat = originLatRadians + angle;
164
165    double minLng;
166    double maxLng;
167    if (minLat > Math.toRadians(-90) && maxLat < Math.toRadians(90)) {
168      double delta = Math.asin(Math.sin(angle) / Math.cos(originLatRadians));
169      minLng = originLngRadians - delta;
170      if (minLng < Math.toRadians(-180)) {
171        minLng += 2 * Math.PI;
172      }
173      maxLng = originLngRadians + delta;
174      if (maxLng > Math.toRadians(180)) {
175        maxLng -= 2 * Math.PI;
176      }
177    } else {
178      // The query includes a pole!
179      minLat = Math.max(minLat, Math.toRadians(-90));
180      maxLat = Math.min(maxLat, Math.toRadians(90));
181      minLng = Math.toRadians(-180);
182      maxLng = Math.toRadians(180);
183    }
184
185    BooleanQuery.Builder f = new BooleanQuery.Builder();
186
187    // Add latitude range filter:
188    f.add(
189        DoublePoint.newRangeQuery("latitude", Math.toDegrees(minLat), Math.toDegrees(maxLat)),
190        BooleanClause.Occur.FILTER);
191
192    // Add longitude range filter:
193    if (minLng > maxLng) {
194      // The bounding box crosses the international date
195      // line:
196      BooleanQuery.Builder lonF = new BooleanQuery.Builder();
197      lonF.add(
198          DoublePoint.newRangeQuery("longitude", Math.toDegrees(minLng), Double.POSITIVE_INFINITY),
199          BooleanClause.Occur.SHOULD);
200      lonF.add(
201          DoublePoint.newRangeQuery("longitude", Double.NEGATIVE_INFINITY, Math.toDegrees(maxLng)),
202          BooleanClause.Occur.SHOULD);
203      f.add(lonF.build(), BooleanClause.Occur.MUST);
204    } else {
205      f.add(
206          DoublePoint.newRangeQuery("longitude", Math.toDegrees(minLng), Math.toDegrees(maxLng)),
207          BooleanClause.Occur.FILTER);
208    }
209
210    return f.build();
211  }
212
213  /** User runs a query and counts facets. */
214  public FacetResult search() throws IOException {
215
216    FacetsCollector fc = new FacetsCollector();
217
218    searcher.search(new MatchAllDocsQuery(), fc);
219
220    Facets facets =
221        new DoubleRangeFacetCounts(
222            "field",
223            getDistanceValueSource(),
224            fc,
225            getBoundingBoxQuery(ORIGIN_LATITUDE, ORIGIN_LONGITUDE, 10.0),
226            ONE_KM,
227            TWO_KM,
228            FIVE_KM,
229            TEN_KM);
230
231    return facets.getTopChildren(10, "field");
232  }
233
234  /** User drills down on the specified range. */
235  public TopDocs drillDown(DoubleRange range) throws IOException {
236
237    // Passing no baseQuery means we drill down on all
238    // documents ("browse only"):
239    DrillDownQuery q = new DrillDownQuery(null);
240    final DoubleValuesSource vs = getDistanceValueSource();
241    q.add(
242        "field",
243        range.getQuery(getBoundingBoxQuery(ORIGIN_LATITUDE, ORIGIN_LONGITUDE, range.max), vs));
244    DrillSideways ds =
245        new DrillSideways(searcher, config, (TaxonomyReader) null) {
246          @Override
247          protected Facets buildFacetsResult(
248              FacetsCollector drillDowns,
249              FacetsCollector[] drillSideways,
250              String[] drillSidewaysDims)
251              throws IOException {
252            assert drillSideways.length == 1;
253            return new DoubleRangeFacetCounts(
254                "field", vs, drillSideways[0], ONE_KM, TWO_KM, FIVE_KM, TEN_KM);
255          }
256        };
257    return ds.search(q, 10).hits;
258  }
259
260  @Override
261  public void close() throws IOException {
262    searcher.getIndexReader().close();
263    indexDir.close();
264  }
265
266  /** Runs the search and drill-down examples and prints the results. */
267  public static void main(String[] args) throws Exception {
268    DistanceFacetsExample example = new DistanceFacetsExample();
269    example.index();
270
271    System.out.println("Distance facet counting example:");
272    System.out.println("-----------------------");
273    System.out.println(example.search());
274
275    System.out.println("Distance facet drill-down example (field/< 2 km):");
276    System.out.println("---------------------------------------------");
277    TopDocs hits = example.drillDown(example.TWO_KM);
278    System.out.println(hits.totalHits + " totalHits");
279
280    example.close();
281  }
282}