|
1
2
3
4 package net.nutch.searcher;
5
6 import java.io.IOException  ;
7 import java.io.File ;
8
9 import java.util.ArrayList ;
10 import java.util.Enumeration ;
11
12 import org.apache.lucene.index.IndexReader;
13 import org.apache.lucene.index.MultiReader;
14
15 import org.apache.lucene.search.MultiSearcher;
16 import org.apache.lucene.search.TopDocs;
17 import org.apache.lucene.search.ScoreDoc;
18 import org.apache.lucene.search.FieldCache;
19
20 import org.apache.lucene.document.Document;
21 import org.apache.lucene.document.Field;
22
23 import net.nutch.io.*;
24 import net.nutch.util.*;
25 import net.nutch.db.*;
26 import net.nutch.fetcher.*;
27 import net.nutch.linkdb.*;
28 import net.nutch.indexer.*;
29 import net.nutch.analysis.NutchDocumentAnalyzer;
30
31
33 public class IndexSearcher implements Searcher, HitDetailer {
34
35 private org.apache.lucene.search.Searcher luceneSearcher;
36
37 private String [] sites;
38
39 private LuceneQueryOptimizer optimizer = new LuceneQueryOptimizer
40 (NutchConf.getInt("searcher.filter.cache.size", 16),
41 NutchConf.getFloat("searcher.filter.cache.threshold", 0.05f));
42
43
44 public IndexSearcher(File [] segmentDirs) throws IOException {
45 IndexReader[] readers = new IndexReader[segmentDirs.length];
46 for (int i = 0; i < segmentDirs.length; i++) {
47 readers[i] = IndexReader.open(new File (segmentDirs[i], "index"));
48 }
49 init(new MultiReader(readers));
50 }
51
52
54 public IndexSearcher(String index)
55 throws IOException {
56 init(IndexReader.open(index));
57 }
58
59 private void init(IndexReader reader) throws IOException {
60 this.sites = FieldCache.DEFAULT.getStrings(reader, "site");
61 this.luceneSearcher = new org.apache.lucene.search.IndexSearcher(reader);
62 this.luceneSearcher.setSimilarity(new NutchSimilarity());
63 }
64
65 public Hits search(Query query, int numHits) throws IOException {
66
67 org.apache.lucene.search.BooleanQuery luceneQuery =
68 QueryFilters.filter(query);
69
70 return translateHits
71 (optimizer.optimize(luceneQuery, luceneSearcher, numHits));
72 }
73
74 public String getExplanation(Query query, Hit hit) throws IOException {
75 return luceneSearcher.explain(QueryFilters.filter(query),
76 hit.getIndexDocNo()).toHtml();
77 }
78
79 public HitDetails getDetails(Hit hit) throws IOException {
80 ArrayList fields = new ArrayList ();
81 ArrayList values = new ArrayList ();
82
83 Document doc = luceneSearcher.doc(hit.getIndexDocNo());
84
85 Enumeration e = doc.fields();
86 while (e.hasMoreElements()) {
87 Field field = (Field)e.nextElement();
88 fields.add(field.name());
89 values.add(field.stringValue());
90 }
91
92 return new HitDetails((String [])fields.toArray(new String [fields.size()]),
93 (String [])values.toArray(new String [values.size()]));
94 }
95
96 public HitDetails[] getDetails(Hit[] hits) throws IOException {
97 HitDetails[] results = new HitDetails[hits.length];
98 for (int i = 0; i < hits.length; i++)
99 results[i] = getDetails(hits[i]);
100 return results;
101 }
102
103 private Hits translateHits(TopDocs topDocs) throws IOException {
104 ScoreDoc[] scoreDocs = topDocs.scoreDocs;
105 int length = scoreDocs.length;
106 Hit[] hits = new Hit[length];
107 for (int i = 0; i < length; i++) {
108 int doc = scoreDocs[i].doc;
109 hits[i] = new Hit(doc, scoreDocs[i].score, sites[doc]);
110 }
111 return new Hits(topDocs.totalHits, hits);
112 }
113
114 }
115 |
Popular Tags |
Java API By Example, From Geeks To Geeks.