|
1 |
| package photospace.search; |
|
2 |
| |
|
3 |
| import java.io.*; |
|
4 |
| import java.util.*; |
|
5 |
| import org.apache.commons.logging.*; |
|
6 |
| import org.apache.lucene.index.*; |
|
7 |
| import photospace.meta.*; |
|
8 |
| |
|
9 |
| public class Indexer |
|
10 |
| implements Serializable |
|
11 |
| { |
|
12 |
| private static final Log log = LogFactory.getLog(Indexer.class); |
|
13 |
| |
|
14 |
| private SearchIndex index; |
|
15 |
| private DocumentFactory factory = new DocumentFactory(); |
|
16 |
| |
|
17 |
1
| public void init() throws IOException
|
|
18 |
| { |
|
19 |
1
| try
|
|
20 |
| { |
|
21 |
1
| index.getReader();
|
|
22 |
| } |
|
23 |
| catch (Exception e) |
|
24 |
| { |
|
25 |
1
| log.info("Initializing search index.");
|
|
26 |
1
| index(new ArrayList(), true);
|
|
27 |
| } |
|
28 |
| } |
|
29 |
| |
|
30 |
15
| public int index(Collection docs, boolean fromScratch) throws IOException
|
|
31 |
| { |
|
32 |
15
| Date start = new Date();
|
|
33 |
| |
|
34 |
15
| Set metas = new HashSet(docs);
|
|
35 |
| |
|
36 |
15
| File destination;
|
|
37 |
14
| if (fromScratch) destination = new File(System.getProperty("java.io.tmpdir"), "photospace-index");
|
|
38 |
1
| else destination = index.getIndex();
|
|
39 |
| |
|
40 |
0
| if (log.isDebugEnabled()) log.debug("Writing index to " + destination);
|
|
41 |
| |
|
42 |
14
| if (!destination.exists()) destination.mkdirs();
|
|
43 |
| |
|
44 |
15
| IndexWriter writer = new IndexWriter(destination, new MetaAnalyzer(), fromScratch);
|
|
45 |
| |
|
46 |
15
| int initialDocs = writer.docCount();
|
|
47 |
| |
|
48 |
15
| try
|
|
49 |
| { |
|
50 |
15
| writer.mergeFactor = 1000;
|
|
51 |
15
| writer.maxMergeDocs = 10000;
|
|
52 |
15
| writer.setUseCompoundFile(true);
|
|
53 |
| |
|
54 |
15
| addDocuments((Meta[]) metas.toArray(new Meta[] {}), writer);
|
|
55 |
| |
|
56 |
15
| writer.optimize();
|
|
57 |
| } |
|
58 |
| finally |
|
59 |
| { |
|
60 |
15
| writer.close();
|
|
61 |
| } |
|
62 |
| |
|
63 |
15
| index.closeReader();
|
|
64 |
14
| if (fromScratch) index.update(destination);
|
|
65 |
| |
|
66 |
15
| int indexedDocs = writer.docCount() - initialDocs;
|
|
67 |
15
| log.info("Indexed " + indexedDocs + " documents in " + (new Date().getTime() - start.getTime()) + " milliseconds");
|
|
68 |
15
| return indexedDocs;
|
|
69 |
| } |
|
70 |
| |
|
71 |
25
| private void addDocuments(Meta[] metas, IndexWriter writer)
|
|
72 |
| throws IOException |
|
73 |
| { |
|
74 |
25
| for (int i = 0; i < metas.length; i++)
|
|
75 |
| { |
|
76 |
43
| Meta meta = metas[i];
|
|
77 |
43
| try
|
|
78 |
| { |
|
79 |
43
| writer.addDocument(factory.createDocument(meta));
|
|
80 |
| } |
|
81 |
| catch (Exception e) |
|
82 |
| { |
|
83 |
0
| log.error("Exception creating document for " + meta, e);
|
|
84 |
| } |
|
85 |
43
| if (meta instanceof FolderMeta)
|
|
86 |
| { |
|
87 |
10
| addDocuments(((FolderMeta) meta).getFiles(), writer);
|
|
88 |
| } |
|
89 |
| } |
|
90 |
| } |
|
91 |
| |
|
92 |
1
| public void merge(Collection metas)
|
|
93 |
| throws Exception |
|
94 |
| { |
|
95 |
1
| delete(metas);
|
|
96 |
1
| index(metas, false);
|
|
97 |
| } |
|
98 |
| |
|
99 |
1
| public void delete(Collection metas)
|
|
100 |
| throws IOException |
|
101 |
| { |
|
102 |
1
| Date start = new Date();
|
|
103 |
1
| synchronized (index)
|
|
104 |
| { |
|
105 |
1
| IndexReader reader = index.getReader();
|
|
106 |
1
| try
|
|
107 |
| { |
|
108 |
1
| for (Iterator i = metas.iterator(); i.hasNext();)
|
|
109 |
| { |
|
110 |
1
| Meta meta = (Meta) i.next();
|
|
111 |
1
| reader.delete(new Term("path", meta.getPath()));
|
|
112 |
| } |
|
113 |
| } |
|
114 |
| finally |
|
115 |
| { |
|
116 |
1
| index.closeReader();
|
|
117 |
| } |
|
118 |
| } |
|
119 |
1
| log.info("Deleted " + metas.size() + " documents in " + (new Date().getTime() - start.getTime()) + " milliseconds");
|
|
120 |
| } |
|
121 |
| |
|
122 |
13
| public void setIndex(SearchIndex index)
|
|
123 |
| { |
|
124 |
13
| this.index = index;
|
|
125 |
| } |
|
126 |
| |
|
127 |
| } |
