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.IOException; 020import java.time.LocalDate; 021import java.time.ZoneOffset; 022import org.apache.lucene.analysis.core.WhitespaceAnalyzer; 023import org.apache.lucene.document.Document; 024import org.apache.lucene.document.Field; 025import org.apache.lucene.document.FloatPoint; 026import org.apache.lucene.document.IntPoint; 027import org.apache.lucene.document.LongPoint; 028import org.apache.lucene.document.StringField; 029import org.apache.lucene.facet.FacetResult; 030import org.apache.lucene.facet.Facets; 031import org.apache.lucene.facet.FacetsCollector; 032import org.apache.lucene.facet.FacetsCollectorManager; 033import org.apache.lucene.facet.facetset.DimRange; 034import org.apache.lucene.facet.facetset.ExactFacetSetMatcher; 035import org.apache.lucene.facet.facetset.FacetSet; 036import org.apache.lucene.facet.facetset.FacetSetDecoder; 037import org.apache.lucene.facet.facetset.FacetSetMatcher; 038import org.apache.lucene.facet.facetset.FacetSetsField; 039import org.apache.lucene.facet.facetset.MatchingFacetSetsCounts; 040import org.apache.lucene.facet.facetset.RangeFacetSetMatcher; 041import org.apache.lucene.index.DirectoryReader; 042import org.apache.lucene.index.IndexWriter; 043import org.apache.lucene.index.IndexWriterConfig; 044import org.apache.lucene.index.IndexWriterConfig.OpenMode; 045import org.apache.lucene.search.IndexSearcher; 046import org.apache.lucene.search.MatchAllDocsQuery; 047import org.apache.lucene.store.ByteBuffersDirectory; 048import org.apache.lucene.store.Directory; 049import org.apache.lucene.util.BytesRef; 050import org.apache.lucene.util.NumericUtils; 051 052/** 053 * Shows usage of indexing and searching {@link FacetSetsField} with a custom {@link FacetSet} 054 * implementation. Unlike the out of the box {@link FacetSet} implementations, this example shows 055 * how to mix and match dimensions of different types, as well as implementing a custom {@link 056 * FacetSetMatcher}. 057 */ 058public class CustomFacetSetExample { 059 060 private static final long MAY_SECOND_2022 = date("2022-05-02"); 061 private static final long JUNE_SECOND_2022 = date("2022-06-02"); 062 private static final long JULY_SECOND_2022 = date("2022-07-02"); 063 private static final float HUNDRED_TWENTY_DEGREES = fahrenheitToCelsius(120); 064 private static final float HUNDRED_DEGREES = fahrenheitToCelsius(100); 065 private static final float EIGHTY_DEGREES = fahrenheitToCelsius(80); 066 067 private final Directory indexDir = new ByteBuffersDirectory(); 068 069 /** Empty constructor */ 070 public CustomFacetSetExample() {} 071 072 /** Build the example index. */ 073 private void index() throws IOException { 074 IndexWriter indexWriter = 075 new IndexWriter( 076 indexDir, new IndexWriterConfig(new WhitespaceAnalyzer()).setOpenMode(OpenMode.CREATE)); 077 078 // Every document holds the temperature measures for a City by Date 079 080 Document doc = new Document(); 081 doc.add(new StringField("city", "city1", Field.Store.YES)); 082 doc.add( 083 FacetSetsField.create( 084 "temperature", 085 new TemperatureReadingFacetSet(MAY_SECOND_2022, HUNDRED_DEGREES), 086 new TemperatureReadingFacetSet(JUNE_SECOND_2022, EIGHTY_DEGREES), 087 new TemperatureReadingFacetSet(JULY_SECOND_2022, HUNDRED_TWENTY_DEGREES))); 088 indexWriter.addDocument(doc); 089 090 doc = new Document(); 091 doc.add(new StringField("city", "city2", Field.Store.YES)); 092 doc.add( 093 FacetSetsField.create( 094 "temperature", 095 new TemperatureReadingFacetSet(MAY_SECOND_2022, EIGHTY_DEGREES), 096 new TemperatureReadingFacetSet(JUNE_SECOND_2022, HUNDRED_DEGREES), 097 new TemperatureReadingFacetSet(JULY_SECOND_2022, HUNDRED_TWENTY_DEGREES))); 098 indexWriter.addDocument(doc); 099 100 indexWriter.close(); 101 } 102 103 /** Counting documents which exactly match a given {@link FacetSet}. */ 104 private FacetResult exactMatching() throws IOException { 105 try (DirectoryReader indexReader = DirectoryReader.open(indexDir)) { 106 IndexSearcher searcher = new IndexSearcher(indexReader); 107 108 // MatchAllDocsQuery is for "browsing" (counts facets 109 // for all non-deleted docs in the index); normally 110 // you'd use a "normal" query: 111 FacetsCollector fc = searcher.search(new MatchAllDocsQuery(), new FacetsCollectorManager()); 112 113 // Count both "May 2022, 100 degrees" and "July 2022, 120 degrees" dimensions 114 Facets facets = 115 new MatchingFacetSetsCounts( 116 "temperature", 117 fc, 118 TemperatureReadingFacetSet::decodeTemperatureReading, 119 new ExactFacetSetMatcher( 120 "May 2022 (100f)", 121 new TemperatureReadingFacetSet(MAY_SECOND_2022, HUNDRED_DEGREES)), 122 new ExactFacetSetMatcher( 123 "July 2022 (120f)", 124 new TemperatureReadingFacetSet(JULY_SECOND_2022, HUNDRED_TWENTY_DEGREES))); 125 126 // Retrieve results 127 return facets.getAllChildren("temperature"); 128 } 129 } 130 131 /** Counting documents which match a certain degrees value for any date. */ 132 private FacetResult rangeMatching() throws IOException { 133 try (DirectoryReader indexReader = DirectoryReader.open(indexDir)) { 134 IndexSearcher searcher = new IndexSearcher(indexReader); 135 136 // MatchAllDocsQuery is for "browsing" (counts facets 137 // for all non-deleted docs in the index); normally 138 // you'd use a "normal" query: 139 FacetsCollector fc = searcher.search(new MatchAllDocsQuery(), new FacetsCollectorManager()); 140 141 // Count 80-100 degrees 142 Facets facets = 143 new MatchingFacetSetsCounts( 144 "temperature", 145 fc, 146 TemperatureReadingFacetSet::decodeTemperatureReading, 147 new RangeFacetSetMatcher( 148 "Eighty to Hundred Degrees", 149 DimRange.fromLongs(Long.MIN_VALUE, true, Long.MAX_VALUE, true), 150 DimRange.fromFloats(EIGHTY_DEGREES, true, HUNDRED_DEGREES, true))); 151 152 // Retrieve results 153 return facets.getAllChildren("temperature"); 154 } 155 } 156 157 /** 158 * Like {@link #rangeMatching()}, however this example demonstrates a custom {@link 159 * FacetSetMatcher} which only considers certain dimensions (in this case only the temperature 160 * one). 161 */ 162 private FacetResult customRangeMatching() throws IOException { 163 try (DirectoryReader indexReader = DirectoryReader.open(indexDir)) { 164 IndexSearcher searcher = new IndexSearcher(indexReader); 165 166 // MatchAllDocsQuery is for "browsing" (counts facets 167 // for all non-deleted docs in the index); normally 168 // you'd use a "normal" query: 169 FacetsCollector fc = searcher.search(new MatchAllDocsQuery(), new FacetsCollectorManager()); 170 171 // Count 80-100 degrees 172 Facets facets = 173 new MatchingFacetSetsCounts( 174 "temperature", 175 fc, 176 TemperatureReadingFacetSet::decodeTemperatureReading, 177 new TemperatureOnlyFacetSetMatcher( 178 "Eighty to Hundred Degrees", 179 DimRange.fromFloats(EIGHTY_DEGREES, true, HUNDRED_DEGREES, true))); 180 181 // Retrieve results 182 return facets.getAllChildren("temperature"); 183 } 184 } 185 186 private static long date(String dateString) { 187 return LocalDate.parse(dateString).atStartOfDay().toInstant(ZoneOffset.UTC).toEpochMilli(); 188 } 189 190 private static float fahrenheitToCelsius(int degrees) { 191 return (degrees - 32.0f) * 5.f / 9.f; 192 } 193 194 /** Runs the exact matching example. */ 195 public FacetResult runExactMatching() throws IOException { 196 index(); 197 return exactMatching(); 198 } 199 200 /** Runs the range matching example. */ 201 public FacetResult runRangeMatching() throws IOException { 202 index(); 203 return rangeMatching(); 204 } 205 206 /** Runs the custom range matching example. */ 207 public FacetResult runCustomRangeMatching() throws IOException { 208 index(); 209 return customRangeMatching(); 210 } 211 212 /** Runs the search and drill-down examples and prints the results. */ 213 public static void main(String[] args) throws Exception { 214 CustomFacetSetExample example = new CustomFacetSetExample(); 215 216 System.out.println("Exact Facet Set matching example:"); 217 System.out.println("-----------------------"); 218 FacetResult result = example.runExactMatching(); 219 System.out.println("Temperature Reading: " + result); 220 221 System.out.println("Range Facet Set matching example:"); 222 System.out.println("-----------------------"); 223 result = example.runRangeMatching(); 224 System.out.println("Temperature Reading: " + result); 225 226 System.out.println("Custom Range Facet Set matching example:"); 227 System.out.println("-----------------------"); 228 result = example.runCustomRangeMatching(); 229 System.out.println("Temperature Reading: " + result); 230 } 231 232 /** 233 * A {@link FacetSet} which encodes a temperature reading in a date (long) and degrees (celsius; 234 * float). 235 */ 236 public static class TemperatureReadingFacetSet extends FacetSet { 237 238 private static final int SIZE_PACKED_BYTES = Long.BYTES + Float.BYTES; 239 240 private final long date; 241 private final float degrees; 242 243 /** Constructor */ 244 public TemperatureReadingFacetSet(long date, float degrees) { 245 super(2); // We encode two dimensions 246 247 this.date = date; 248 this.degrees = degrees; 249 } 250 251 @Override 252 public long[] getComparableValues() { 253 return new long[] {date, NumericUtils.floatToSortableInt(degrees)}; 254 } 255 256 @Override 257 public int packValues(byte[] buf, int start) { 258 LongPoint.encodeDimension(date, buf, start); 259 // Encode 'degrees' as a sortable integer. 260 FloatPoint.encodeDimension(degrees, buf, start + Long.BYTES); 261 return sizePackedBytes(); 262 } 263 264 @Override 265 public int sizePackedBytes() { 266 return SIZE_PACKED_BYTES; 267 } 268 269 /** 270 * An implementation of {@link FacetSetDecoder#decode(BytesRef, int, long[])} for {@link 271 * TemperatureReadingFacetSet}. 272 */ 273 public static int decodeTemperatureReading(BytesRef bytesRef, int start, long[] dest) { 274 dest[0] = LongPoint.decodeDimension(bytesRef.bytes, start); 275 // Decode the degrees as a sortable integer. 276 dest[1] = IntPoint.decodeDimension(bytesRef.bytes, start + Long.BYTES); 277 return SIZE_PACKED_BYTES; 278 } 279 } 280 281 /** 282 * A {@link FacetSetMatcher} which matches facet sets only by their temperature dimension, 283 * ignoring the date. 284 */ 285 public static class TemperatureOnlyFacetSetMatcher extends FacetSetMatcher { 286 287 private final DimRange temperatureRange; 288 289 /** Constructor */ 290 protected TemperatureOnlyFacetSetMatcher(String label, DimRange temperatureRange) { 291 super(label, 1); // We only evaluate one dimension 292 293 this.temperatureRange = temperatureRange; 294 } 295 296 @Override 297 public boolean matches(long[] dimValues) { 298 return temperatureRange.min <= dimValues[1] && temperatureRange.max >= dimValues[1]; 299 } 300 } 301}