Lucene3.6.2入门系列

文章15 |   阅读 6974 |   点赞0

来源：https://blog.csdn.net/jadyer/category_1571227.html

【Lucene3.6.2入门系列】第07节_高级搜索之普通Filter和自定义Filter

转x33g5p2x  于2021-12-24 转载在 其他  
字(7.4k)|赞(0)|评价(0)|浏览(428)

完整版见https://jadyer.github.io/2013/08/19/lucene-advanced-search-filter/

  1. package com.jadyer.lucene;
  3. import java.io.File;
  4. import java.io.IOException;
  5. import java.text.ParseException;
  6. import java.text.SimpleDateFormat;
  7. import java.util.Date;
  9. import org.apache.lucene.analysis.standard.StandardAnalyzer;
  10. import org.apache.lucene.document.Document;
  11. import org.apache.lucene.document.Field;
  12. import org.apache.lucene.document.NumericField;
  13. import org.apache.lucene.index.IndexReader;
  14. import org.apache.lucene.index.IndexWriter;
  15. import org.apache.lucene.index.IndexWriterConfig;
  16. import org.apache.lucene.queryParser.QueryParser;
  17. import org.apache.lucene.search.Filter;
  18. import org.apache.lucene.search.IndexSearcher;
  19. import org.apache.lucene.search.ScoreDoc;
  20. import org.apache.lucene.search.TopDocs;
  21. import org.apache.lucene.store.Directory;
  22. import org.apache.lucene.store.FSDirectory;
  23. import org.apache.lucene.util.Version;
  25. import com.jadyer.custom.MyFilter;
  27. /**
  28.  * 【Lucene3.6.2入门系列】第07节_高级搜索之普通Filter和自定义Filter
  29.  * @create Aug 19, 2013 11:13:40 AM
  30.  * @author 玄玉<http://blog.csdn.net/jadyer>
  31.  */
  32. public class AdvancedSearchByFilter {
  33. 	private Directory directory;
  34. 	private IndexReader reader;
  35. 	
  36. 	public AdvancedSearchByFilter(){
  37. 		/**文件大小*/
  38. 		int[] sizes = {90, 10, 20, 10, 60, 50};
  39. 		/**文件名*/
  40. 		String[] names = {"Michael.java", "Scofield.ini", "Tbag.txt", "Jack", "Jade", "Jadyer"};
  41. 		/**文件内容*/
  42. 		String[] contents = {"my java blog is http://blog.csdn.net/jadyer",
  43. 							 "my Java Website is http://www.jadyer.cn",
  44. 							 "my name is jadyer",
  45. 							 "I am a Java Developer",
  46. 							 "I am from Haerbin",
  47. 							 "I like java of Lucene"};
  48. 		/**文件日期*/
  49. 		Date[] dates = new Date[sizes.length];
  50. 		SimpleDateFormat sdf = new SimpleDateFormat("yyyyMMdd HH:mm:ss");
  51. 		IndexWriter writer = null;
  52. 		Document doc = null;
  53. 		try {
  54. 			dates[0] = sdf.parse("20130407 15:25:30");
  55. 			dates[1] = sdf.parse("20130407 16:30:45");
  56. 			dates[2] = sdf.parse("20130213 11:15:25");
  57. 			dates[3] = sdf.parse("20130808 09:30:55");
  58. 			dates[4] = sdf.parse("20130526 13:54:22");
  59. 			dates[5] = sdf.parse("20130701 17:35:34");
  60. 			directory = FSDirectory.open(new File("myExample/01_index/"));
  61. 			writer = new IndexWriter(directory, new IndexWriterConfig(Version.LUCENE_36, new StandardAnalyzer(Version.LUCENE_36)));
  62. 			writer.deleteAll();
  63. 			for(int i=0; i<sizes.length; i++){
  64. 				doc = new Document();
  65. 				doc.add(new NumericField("size",Field.Store.YES, true).setIntValue(sizes[i]));
  66. 				doc.add(new Field("name", names[i], Field.Store.YES, Field.Index.ANALYZED_NO_NORMS));
  67. 				doc.add(new Field("content", contents[i], Field.Store.NO, Field.Index.ANALYZED));
  68. 				doc.add(new NumericField("date", Field.Store.YES, true).setLongValue(dates[i].getTime()));
  69. 				//为每个文档添加一个fileID(与ScoreDoc.doc不同),专门在自定义Filter时使用
  70. 				doc.add(new Field("fileID", String.valueOf(i), Field.Store.YES, Field.Index.NOT_ANALYZED_NO_NORMS));
  71. 				writer.addDocument(doc);
  72. 			}
  73. 		} catch (Exception e) {
  74. 			e.printStackTrace();
  75. 		} finally {
  76. 			if(null != writer){
  77. 				try {
  78. 					writer.close();
  79. 				} catch (IOException ce) {
  80. 					ce.printStackTrace();
  81. 				}
  82. 			}
  83. 		}
  84. 	}
  85. 	
  86. 	/**
  87. 	 * 获取IndexReader实例
  88. 	 */
  89. 	private IndexReader getIndexReader(){
  90. 		try {
  91. 			if(reader == null){
  92. 				reader = IndexReader.open(directory);
  93. 			}else{
  94. 				//if the index was changed since the provided reader was opened, open and return a new reader; else,return null
  95. 				//如果当前reader在打开期间index发生改变,则打开并返回一个新的IndexReader,否则返回null
  96. 				IndexReader ir = IndexReader.openIfChanged(reader);
  97. 				if(ir != null){
  98. 					reader.close(); //关闭原reader
  99. 					reader = ir;    //赋予新reader
  100. 				}
  101. 			}
  102. 			return reader;
  103. 		}catch(Exception e) {
  104. 			e.printStackTrace();
  105. 		}
  106. 		return null; //发生异常则返回null
  107. 	}
  108. 	
  109. 	
  110. 	/**
  111. 	 * 搜索过滤
  112. 	 */
  113. 	public void searchByFilter(String expr, Filter filter){
  114. 		IndexSearcher searcher = new IndexSearcher(this.getIndexReader());
  115. 		QueryParser parser = new QueryParser(Version.LUCENE_36, "content", new StandardAnalyzer(Version.LUCENE_36));
  116. 		TopDocs tds = null;
  117. 		try {
  118. 			if(null == filter){
  119. 				tds = searcher.search(parser.parse(expr), 10);
  120. 			}else{
  121. 				tds = searcher.search(parser.parse(expr), filter, 10);
  122. 			}
  123. 			for(ScoreDoc sd : tds.scoreDocs){
  124. 				Document doc = searcher.doc(sd.doc);
  125. 				System.out.print("文档编号=" + sd.doc + "  文档权值=" + doc.getBoost() + "  文档评分=" + sd.score + "    ");
  126. 				System.out.println("fileID=" + doc.get("fileID") + "  size=" + doc.get("size") + "  date=" + new SimpleDateFormat("yyyyMMdd HH:mm:ss").format(new Date(Long.parseLong(doc.get("date")))) + "  name=" + doc.get("name"));
  127. 			}
  128. 		} catch (Exception e) {
  129. 			e.printStackTrace();
  130. 		} finally {
  131. 			if(searcher != null){
  132. 				try {
  133. 					searcher.close();
  134. 				} catch (IOException e) {
  135. 					e.printStackTrace();
  136. 				}
  137. 			}
  138. 		}
  139. 	}
  140. 	
  141. 	
  142. 	/**
  143. 	 * 测试一下过滤效果
  144. 	 */
  145. 	public static void main(String[] args) throws ParseException {
  146. 		AdvancedSearchByFilter advancedSearch = new AdvancedSearchByFilter();
  147. //		//过滤文件名首字母从'h'到'n'的记录(注意hn要小写)
  148. //		advancedSearch.searchByFilter("Java", new TermRangeFilter("name", "h", "n", true, true));
  149. //		//过滤文件大小在30到80以内的记录
  150. //		advancedSearch.searchByFilter("Java", NumericRangeFilter.newIntRange("size", 30, 80, true, true));
  151. //		//过滤文件日期在20130701 00:00:00到20130808 23:59:59之间的记录
  152. //		Long min = Long.valueOf(new SimpleDateFormat("yyyyMMdd").parse("20130701").getTime());
  153. //		Long max = Long.valueOf(new SimpleDateFormat("yyyyMMdd HH:mm:ss").parse("20130808 23:59:59").getTime());
  154. //		advancedSearch.searchByFilter("Java", NumericRangeFilter.newLongRange("date", min, max, true, true));
  155. //		//过滤文件名以'ja'打头的(注意ja要小写)
  156. //		advancedSearch.searchByFilter("Java", new QueryWrapperFilter(new WildcardQuery(new Term("name", "ja*"))));
  157. 		//自定义Filter
  158. 		advancedSearch.searchByFilter("Java", new MyFilter());
  159. 	}
  160. }

下面是自定义的MyFilter.java

  1. package com.jadyer.custom;
  3. import java.io.IOException;
  5. import org.apache.lucene.index.IndexReader;
  6. import org.apache.lucene.index.Term;
  7. import org.apache.lucene.index.TermDocs;
  8. import org.apache.lucene.search.DocIdSet;
  9. import org.apache.lucene.search.Filter;
  10. import org.apache.lucene.util.OpenBitSet;
  12. /**
  13.  * 自定义Filter
  14.  * @see ------------------------------------------------------------------------------------------
  15.  * @see 本例的应用场景
  16.  * @see 假设很多的数据,然后删除了其中的某几条数据,此时在接受搜索请求时为保证不会搜索到已删除的数据
  17.  * @see 那么可以更新索引,但更新索引会消耗很多时间(因为数据量大),而又要保证已删除的数据不会被搜索到
  18.  * @see 此时就可以自定义Filter,原理即搜索过程中,当发现此记录为已删除记录,则不添加到返回的搜索结果集中
  19.  * @see ------------------------------------------------------------------------------------------
  20.  * @see 自定义Filter步骤如下
  21.  * @see 1)继承Filter类并重写getDocIdSet()方法
  22.  * @see 2)根据实际过滤要求返回新的DocIdSet对象
  23.  * @see ------------------------------------------------------------------------------------------
  24.  * @see DocIdSet小解
  25.  * @see 这里Filter干的活其实就是创建一个DocIdSet,而DocIdSet其实就是一个数组,可以理解为其中只存放0或1的值
  26.  * @see 每个搜索出来的Document都有一个文档编号,所以搜索出来多少个Document,那么DocIdSet中就会有多少条记录
  27.  * @see 而DocIdSet中每一条记录的索引号与文档编号是一一对应的
  28.  * @see 所以当DocIdSet中的记录为1时,则对应文档编号的Document就会被添加到TopDocs中,为0就会被过滤掉
  29.  * @see ------------------------------------------------------------------------------------------
  30.  * @create Aug 6, 2013 7:28:53 PM
  31.  * @author 玄玉<http://blog.csdn.net/jadyer>
  32.  */
  33. public class MyFilter extends Filter {
  34. 	private static final long serialVersionUID = -8955061358165068L;
  35. 	
  36. 	//假设这是已删除记录的fileID值的集合
  37. 	private String[] deleteFileIDs = {"1", "3"};
  38. 	
  39. 	@Override
  40. 	public DocIdSet getDocIdSet(IndexReader reader) throws IOException {
  41. 		//创建一个DocIdSet的子类OpenBitSet(创建之后默认所有元素都是0),传的参数就是本次"搜索到的"元素数目
  42. 		OpenBitSet obs = new OpenBitSet(reader.maxDoc());
  43. 		//先把元素填满,即全部设置为1
  44. 		obs.set(0, reader.maxDoc());
  45. 		//用于保存已删除元素的文档编号
  46. 		int[] docs = new int[1];
  47. 		for(String deleteDataID : deleteFileIDs){
  48. 			//获取已删除元素对应的TermDocs
  49. 			TermDocs tds = reader.termDocs(new Term("fileID", deleteDataID));
  50. 			//将已删除元素的文档编号放到docs中,将其出现的频率放到freqs中,最后返回查询出来的元素数目
  51. 			int count = tds.read(docs, new int[1]);
  52. 			if(count == 1){
  53. 				//将这个位置docs[0]的元素删除
  54. 				obs.clear(docs[0]);
  55. 			}
  56. 		}
  57. 		return obs;
  58. 	}
  59. }

版权说明 : 本文为转载文章, 版权归原作者所有 版权申明

原文链接 : https://blog.csdn.net/jadyer/article/details/10065993

内容来源于网络，如有侵权，请联系作者删除！

Filtersolrlucene搜索过滤

相关文章

    查看更多
    • 蜀ICP备13028337号-1 大数据知识库 https://www.saoniuhuo.com © All rights reserved
    • 本站内容来源互联网,如果侵犯您的权益请联系我们删除, 联系方式：448109455@qq.com