Lucene基本应用示例

以下两段代码分别展示如何创建lucene索引和通过lucene进行检索：

1.为指定的数据表创建索引

indexDirPath是索引文件的存放路径。clearDir是指明是否在每次建立索引时清空文件夹。batchIndexingDao是专来提取数据的DAO接口，它有一个public <T> List<T> getBatchRecords(String tableName, int beginIndex, int count);方法用来分批次的从表中提取数据。此接口一般由某具体Dao实现。tableName，用于需要做索引的数据表。其实这个参数是不应该出现的，它应该由某个具体的DAO封装起来了。这里是由于程序中的特殊原因。indexDocWrapper是很有意思的一个接口，它很像spirngjdbc的rowmapper，由于把DAO返回的实体对象包裹或者说适配成lucene的Document对象。

/** * Creates index. This is generic method for creating index. * * @param <T> the object type which will be indexed. * @param indexFileDir the index file's directory. * @param clearDir whether clear the direcory first when creating index. * @param batchIndexingDao the batch indexing dao, it's used to load objects from database when indexing. * @param tableName the table name. there are many topic and reply daily tables. * @param indexDocWrapper the wrapper to wrap an object into document. */ public <T> void createIndex(String indexDirPath, boolean clearDir, BatchIndexingDao batchIndexingDao ,String tableName , IndexDocWrapper<T> indexDocWrapper){ IndexWriter writer = null; try { File indexFileDir = new File(indexDirPath); if(indexFileDir.exists()&&clearDir==true){ FileUtils.cleanDirectory(indexFileDir); logger.info("The specified direcory:" + indexDirPath +" has cleared."); } writer = new IndexWriter(FSDirectory.open(indexFileDir), new IKAnalyzer(), true, IndexWriter.MaxFieldLength.LIMITED); writer.setRAMBufferSizeMB(Constants.INDEX_WRITER_RAM_BUFFER); int recordIndex = 0; List<T> objects = null; do { try { objects = batchIndexingDao.getBatchRecords(tableName, recordIndex, Constants.BATCH_INDEX_SIZE); } catch (BadSqlGrammarException e) { logger.error(tableName+" does not exist!"); continue; } if (objects != null && !objects.isEmpty()) { for (T object : objects) { writer.addDocument(indexDocWrapper.createDoc(object)); } logger.info(recordIndex + objects.size() + " records has added!"); recordIndex += Constants.BATCH_INDEX_SIZE; } } while (objects != null && !objects.isEmpty()); logger.info("Starts to optimize..."); writer.optimize(); logger.info("Optimize finished."); } catch (CorruptIndexException e) { logger.error(e.getMessage()); e.printStackTrace(); } catch (LockObtainFailedException e) { logger.error(e.getMessage()); e.printStackTrace(); } catch (IOException e) { logger.error(e.getMessage()); e.printStackTrace(); } finally { if(writer!=null){ try { writer.close(); logger.info("Topic indexing work finished."); } catch (CorruptIndexException e) { e.printStackTrace(); } catch (IOException e) { e.printStackTrace(); } } } }

2. 根据给定的关键字检索命中的对象

indexDirPath是索引文件的存放路径。field是将被检索的Lucene的Document的field.expression是lucene表达式。indexObjectWrapper和indexDocWrapper是对应的，当是把lucene的Document提取出来封装到对象中的包裹器。beginIndex和count是当检索结果很多时用来分页的。

/** * Search. This is generic method for searching with paging. * * @param <T> the searching target object type. * @param indexFileDir the index file dir * @param field the document field for searching. * @param expression the expression * @param indexObjectWrapper the index object wrapper * @param beginIndex the begin index * @param count the count * @return the list */ public <T> List<T> search(String indexFileDir, String field, String expression, IndexObjectWrapper<T> indexObjectWrapper, int beginIndex, int count){ IndexSearcher indexSearcher = null; try { logger.info("Starts to search..."); if(expression==null||"".equals(expression.trim())){ throw new NullPointerException("The lucene expression is null or empty!"); } indexSearcher = new IndexSearcher(FSDirectory.open(new File(indexFileDir))); Query query = new QueryParser(Version.LUCENE_CURRENT, field, new IKAnalyzer()).parse(expression); TopDocs topDocs = indexSearcher.search(query, Integer.MAX_VALUE); ScoreDoc[] scoreDocs = topDocs.scoreDocs; logger.info("Total Hits: "+topDocs.totalHits); List<T> objects = new ArrayList<T>(); int endIndex = beginIndex+count; for (int i = beginIndex; i < endIndex; i++) { Document targetDoc = indexSearcher.doc(scoreDocs[i].doc); objects.add(indexObjectWrapper.createObject(targetDoc)); } return objects; } catch (Exception e) { logger.info(e.getMessage()); e.printStackTrace(); return null; } finally { if(indexSearcher!=null){ try { indexSearcher.close(); } catch (IOException e) { e.printStackTrace(); } } } }