简单的运用Lucene进行检索

文章来源：http://www.blogjava.net/wangdei/archive/2008/06/17/208696.html

     Lucene不是一个完整的全文索引应用，而是是一个用Java写的全文索引引擎工具包，它可以方便的嵌入到各种应用中实现针对应用的全文索引/检索功能。
为了更快的体验lucene.本文作者写了一个比较简单的类.大家可以上 BT下载 或是 小说520网看看其效果.

 

public class BtLucene {
    
        private static Logger logger = Logger.getLogger(BtLucene.class);
        
        public static String[] StopStrs = {"BT285","BT软件","BT电影","BT下载"};
        
        /**
         * 查询
         * @param queryStr
         * @param lucennePath
         * @return
         * @throws Exception
         */
        public LuceneModel query(String queryStr,String lucennePath) throws Exception {        
            String queryUTF8 = URLDecoder.decode(queryStr,"UTF-8"); 
            LuceneModel luceneModel = new LuceneModel();
            List<LuceneInfo> lucneneInfoList = new ArrayList<LuceneInfo>();
            long begin = System.currentTimeMillis();
            Document doc = new Document();
            StandardAnalyzer analyzer = new StandardAnalyzer(StopStrs);
            // 一段简单的检索代码
            QueryParser queryParser = new QueryParser("title", analyzer);
            Query query = queryParser.parse(queryUTF8);
            // 检索
            Searcher searcher = new IndexSearcher(lucennePath);// "index"指定索引文件位置
            Hits hits = searcher.search(query);
            int size = hits.length();
            // 打印结果值集
            if(logger.isDebugEnabled()){
                logger.debug("result size is " + size);
            }
            luceneModel.setSize(size);
            for (int i = 0; i < size; i++) {                        
                LuceneInfo lucneneInfo = new LuceneInfo();
                doc = hits.doc(i);
                String id = doc.get("id");
                String title = doc.get("title");
                String[] splitTitle = title.split(queryUTF8);
                lucneneInfo.setId(id);
                if(splitTitle.length >1)
                    lucneneInfo.setTitle(splitTitle[0]+ "" +queryUTF8 +"" + splitTitle[1]);
                else
                    lucneneInfo.setTitle(splitTitle[0]+ "" +queryUTF8 +"");
                lucneneInfo.setTrip(title);
                lucneneInfoList.add(lucneneInfo);
                if(i==200)
                    break;
            }
            long needsTime = (System.currentTimeMillis()-begin);
            long compiteTime = needsTime/1000;
            luceneModel.setTime(String.valueOf(compiteTime));
            luceneModel.setLuceneInfoList(lucneneInfoList);
            logger.info("query the " + queryUTF8 + " needs " + needsTime +" ms" );
            return luceneModel;
        }

        
        /**
         * 建立索引
         * @throws Exception
         */
        public void batchCreate() throws Exception {
            Configure.propertiesConfigure();
            BtBatchContentCreate contentCreate = new BtBatchContentCreate();
            ClassPathXmlApplicationContext appContext = new ClassPathXmlApplicationContext(
                    "./mysqlContext.xml");
            WNewsDAO newsDAO = (WNewsDAO) appContext.getBean("wNewDaoProxy");
            Bt285DAO bt285DAO = (Bt285DAO) appContext.getBean("bt285DAO");
            contentCreate.setNewsDAO(newsDAO);
            contentCreate.setBt285DAO(bt285DAO);
            LieService lieService = (LieService) appContext.getBean("lieService");
            contentCreate.setLieService(lieService);
            StandardAnalyzer analyzer = new StandardAnalyzer(StopStrs);
            IndexWriter writer = new IndexWriter(Configure.getCreateBtLucenePath(), analyzer, true);

            for (int i = 1; i < 214; i++) {
                Page page = new Page();
                logger.info("i=" + i);
                page.setPageIndex(i);
                page.setPageSize(1000);
                List<Bt285> list = bt285DAO.findPageByQuery(
                                    "select t from Bt285 t ",null, page);
                for (Bt285 news : list) {
                    logger.debug("news Id=" + news.getId());        
                    // 添加一条文档
                    Document doc = new Document();
                    String title = news.getTitle();
                    String newTitle = null;
                    if(title == null)
                        title ="no title";
                    newTitle = title.replace("|BT285.cn|BT下载|BT电影|BT软件", "");
                    doc.add(new Field("id", String.valueOf(news.getId()), Field.Store.YES, Field.Index.NO));
                    doc.add(new Field("title", newTitle, Field.Store.YES,
                            Field.Index.TOKENIZED));
                    doc.setBoost(news.getId() * 10);
                    writer.addDocument(doc);
                }
            }
            writer.optimize();
            writer.close();
        }
        public static void main(String[] args) throws Exception {
            System.out.println("server begin!");
            Configure.propertiesConfigure();
            BtLucene action = new BtLucene();
            //action.batchCreate();        
            String path = Configure.getCreateBtLucenePath();
            action.query("nba",path);
            System.out.println(URLEncoder.encode("天兆","UTF-8"));//%E5%A4%A9%E5%85%86
            System.out.println("server finish!");
        }
}

 搜索法证先锋II 出来的效果.
 
[TVB连续剧][法证先锋II][粤语中字][TV-RMVB]