lucene在时间范围内对关键字的查询和分页

  一般简单的全文检索只需要输入关键字，然后我们按时间倒序对结果排序。
  当数据量庞大的时候还是不够友好，所以需要加上时间范围的约束。
  本文是基于别人的例子改的，只为个人积累^_^
参考出处http://zfsn.iteye.com/blog/520363

public class SearchUtil {

	//索引所在文件夹
	private String indexDir;
	//查询起始时间
	private String sDateStr;
	//查询结束时间
	private String eDateStr;
	private RangeQuery rangeQuery;
	private Query query;
	private BooleanQuery booleanQuery;
	//时间格式化类型
	private static SimpleDateFormat sdf = new SimpleDateFormat("yyyyMMdd");

	public Map getSearchResult(String searchType, String key, Date sDate,
			Date eDate, int start, int limit) {
		
		booleanQuery = new BooleanQuery();
		// 搜索开始时间
		Date beginTime = new Date();
		ArrayList al = new ArrayList();
		HashMap map = new HashMap();
		// 获取记录数组
		List list = new ArrayList();
		Analyzer analyzer = new IKAnalyzer();
		Searcher searcher = null;
		Directory directory = null;
		try{
		// 定义解析器
		QueryParser parser = new QueryParser(searchType, analyzer);
		indexDir = IndexUtil.getValue(IndexUtil.INDEX_ROOT2);
		//indexDir 为了测试可以直接写成绝对路径
		directory = FSDirectory.open(new File(indexDir));
		searcher = new IndexSearcher(directory);
		searcher.setSimilarity(new IKSimilarity());
		
		// 若只有结束值起始值默认为1900年1月1日
		if ((sDate == null || "".equals(sDate))
				&& (eDate != null && !"".equals(eDate))) {
			Calendar calendar = Calendar.getInstance();
			calendar.set(1900, 0, 1);
			sDate = calendar.getTime();
		}
		// 若只有起始值结束值默认为当天
		if ((sDate != null && !"".equals(sDate))
				&& (eDate == null || "".equals(eDate))) {
			eDate = new Date();
		}
		if ((sDate != null && !"".equals(sDate))
				&& (eDate != null || !"".equals(eDate))) {

			// Lucene日期转换格式不准，改用format格式
			// sDateStr=DateTools.dateToString(sDate,
			// DateTools.Resolution.MINUTE);
			// eDateStr=DateTools.dateToString(eDate,
			// DateTools.Resolution.MINUTE);

			sDateStr = sdf.format(sDate);
			eDateStr = sdf.format(eDate);

			//时间范围查询
			Term tstart = new Term("createDate", sDateStr);
			Term tend = new Term("createDate", eDateStr);
			rangeQuery = new RangeQuery(tstart, tend, true);
		}

		if (key != null && !"".equals(key)) {
			query = parser.parse(key);
		}

		//按条件进行多条件查询
		if ((sDate != null && !"".equals(sDate))
				&& (eDate != null && !"".equals(eDate)) && (key != null)
				&& !"".equals(key)) {
			booleanQuery.add(rangeQuery, BooleanClause.Occur.MUST);
			booleanQuery.add(query, BooleanClause.Occur.MUST);
		} else if ((sDate != null && !"".equals(sDate))
				&& (eDate != null && !"".equals(eDate))) {
			booleanQuery.add(rangeQuery, BooleanClause.Occur.MUST);

		} else if (key != null && !"".equals(key)) {
			booleanQuery.add(query, BooleanClause.Occur.MUST);

		} else {
			booleanQuery.add(rangeQuery, BooleanClause.Occur.SHOULD);
			booleanQuery.add(query, BooleanClause.Occur.SHOULD);
		}

		// 索引排序条件
		 SortField[] sortfield = new SortField[] { SortField.FIELD_SCORE,
		 new SortField("createDate", SortField.STRING, true) };
		Sort sort = new Sort(sortfield);
		 TopFieldDocs docs = searcher.search(booleanQuery, null, 1, sort);
		 int totalNum = docs.totalHits;
		 docs = searcher.search(booleanQuery, null, totalNum, sort);
		ScoreDoc[] scoreDocs = docs.scoreDocs;

		//高亮设置
		SimpleHTMLFormatter simpleHtmlFormatter = new SimpleHTMLFormatter(
				"<FONT COLOR='RED'>", "</FONT>");
		Highlighter highlighter = new Highlighter(simpleHtmlFormatter,
				new QueryScorer(query));
		// 设置提取字符串长度
		highlighter.setTextFragmenter(new SimpleFragmenter(58));
			
			if (scoreDocs.length == 0) {
				System.out.println("没有符合条件的记录");
			} else {

				for (int i = start; i < scoreDocs.length && i < start + limit; i++) {
					
					ScoreDoc scdoc = scoreDocs[i];
					Document document = searcher.doc(scdoc.doc);

					TokenStream tokenStream = analyzer.tokenStream("",
							new StringReader(document.get("content")));
					//提取高亮显示内容
					String str = highlighter.getBestFragment(tokenStream, document
							.get("content"));
					System.out.println(str);
					System.out.println(document.getField("createDate").toString());
					al.add(str);
				}
				
				// 搜索完成时间
				Date endTime = new Date();
				// 搜索所耗时间
				long timeOfSearch = endTime.getTime() - beginTime.getTime();
				System.out.println("The time For indexsearch is " + timeOfSearch
						+ " ms");

				map.put("list", al);
				map.put("time", timeOfSearch);
				map.put("total", totalNum);
			}
		}catch (Exception e) {
			e.printStackTrace();
			
		}finally {
			if (searcher != null) {
				try {
					searcher.close();
				} catch (IOException e) {
					e.printStackTrace();
				}
			}
			if (directory != null) {
				try {
					directory.close();
				} catch (IOException e) {
					e.printStackTrace();
				}
			}
		}
		return map;
	}

	//测试方法
	public static void main(String[] args) {
		SearchUtil s = new SearchUtil();
		try {
			Date sDate = sdf.parse("20101228");
			Date eDate = sdf.parse("20101229");
			s.getSearchResult("content", "星期三", null, eDate, 0, 15);
		} catch (Exception e) {
			e.printStackTrace();
		}
	}
}

ps: lucene版本为2.9.3 不过3.0也是可以用的代码不需改变

评论

1 楼 diyunpeng 2012-12-03  

参考一下。。。。