Lucene
Java library for high-performance text search. Works by creating/searching an optimized index.
The code below has been tested with Lucene 2.3.2
Indexing
// create index IndexWriter writer = new IndexWriter(new File("/just/a/path/testIndex"), new StandardAnalyzer(), true); // create document Document doc = new Document(); doc.add(new Field("id", "5", Field.Store.YES, Field.Index.UN_TOKENIZED)); doc.add(new Field("name", "Just a name", Field.Store.YES, Field.Index.TOKENIZED)); doc.add(new Field("multifield", "toto\ntata", Field.Store.YES, Field.Index.TOKENIZED)); // you may want to delete in the index any already existing document with the same id writer.deleteDocuments(new Term("id", "5")); // add document to index writer.addDocument(doc); writer.flush(); // create/add other documents //...
- Field.Store.YES: if you need to get this value from the search result list. Otherwise it's just indexed and discarded.
- Field.Index.TOKENIZED: if this value needs to be indexed (it will be used in the search query). Otherwise you'll get it only from the results.
Searching
// get index IndexSearcher indexSearcher = new IndexSearcher("/just/a/path/testIndex"); // create query BooleanQuery booleanQuery = new BooleanQuery(); // add clause to query QueryParser parser = new QueryParser("name", new StandardAnalyzer()); booleanQuery.add(parser.parse("Just a name"), BooleanClause.Occur.MUST); // create/add other clauses to query //... // execute query on index Hits hits = indexSearcher.search(booleanQuery); // use results for(int i = 0; i < hits.length(); i++) { String id = hits.doc(i).get("id"); //... }
Download working example
Includes indexing, multi-criteria and multi-field search: Eclipse project with all dependencies.