好久没有更新了,今天说说最近为这个项目加的一个新功能吧,即全文检索Lucene!至于Lucene到底是什么东西,大家可以在自己学习一下,我这里只说说是怎样将其配置到我的项目中的.大家如果对我这个项目不是很了解,可以先看看前面几个帖子.
其实说到Lucene,我也是第一次接触,以前听说过,但没有用过.搞了两天,先是看视频,看完了还是不会配到我的S2SH框架中,没办法,咱人笨.最后找项目,配合视频,终于搞定了.Lucene主要是操作其自己的索引,所以其实和数据库没有多少关系,但为了和我的项目框架保持协调性,我依然采用了dao-->service-->action这种模式,其实完全没有必要,我个人觉得有一个service层就足够了.好了,废话补多少了,直接上代码:
public class SearchDaoImpl implements SearchDao {
public void save(Document doc, IndexWriter indexWriter) {
try {
indexWriter.addDocument(doc);
} catch (Exception e) {
e.printStackTrace();
}finally{
try {
indexWriter.optimize();
} catch (Exception e) {
e.printStackTrace();
}
}
}
public void delete(Term term , IndexWriter indexWriter) {
try {
indexWriter.deleteDocuments(term);
} catch (Exception e) {
e.printStackTrace();
}finally{
try {
indexWriter.close();
} catch (Exception e) {
e.printStackTrace();
}
}
}
public void update(Term term, Document doc, IndexWriter indexWriter) {
try {
indexWriter.updateDocument(term, doc);
} catch (Exception e) {
e.printStackTrace();
}finally{
try {
indexWriter.close();
} catch (Exception e) {
e.printStackTrace();
}
}
}
//返回查询结果集
public TopDocs search(Query query, IndexSearcher indexSearcher) throws Exception {
Filter filter = null;
return indexSearcher.search(query, filter, 10000);
}
大家可以看到,我其实连过滤器都没有写,主要就是为了熟悉这个框架和S2SH项目的整合使用.
public class ForumSearchServiceImpl implements ForumSearchService {
private SearchDao searchDao;
//保存索引
public void saveForumIndex(List forumList) throws Exception{
File indexFile = new File("D:\\index");
Analyzer analyzer = new PaodingAnalyzer();
IndexWriter indexWriter = new IndexWriter(indexFile, analyzer, true, MaxFieldLength.LIMITED);
List<Forum> list =forumList;
//对所有的论坛进行索引创建
for(Forum forum : list){
Document doc = new Document();
String id = forum.getId();
Long bid = forum.getBoard().getId();
String title = forum.getTitle();
String detail = forum.getDetail();
String postTime = BBSNSUtil.formatDateTime(new Date(forum.getPostTime()), Constant.formatDate);
//因为要搜索和高亮,所以index是tokennized,TermVector是WITH_POSITIONS_OFFSETS
doc.add(new Field("title", title, Field.Store.YES, Field.Index.ANALYZED, Field.TermVector.WITH_POSITIONS_OFFSETS));
//利用htmlparser得到新闻内容html的纯文本
Parser parser = new Parser();
parser.setInputHTML(detail);
String strings = parser.parse(null).elementAt(0).toPlainTextString().trim();
System.out.println("-------"+strings);
if(!(strings.length()==0)){
doc.add(new Field("detail", strings, Field.Store.YES, Field.Index.ANALYZED, Field.TermVector.WITH_POSITIONS_OFFSETS));
}else{
String str = "内容为视频或者图片,不包含文字";
doc.add(new Field("detail", str, Field.Store.YES, Field.Index.ANALYZED, Field.TermVector.WITH_POSITIONS_OFFSETS));
System.out.println(str);
}
//添加时间至文档,因为要按照此字段降序排列排序,所以tokenzied,不用高亮所以TermVector是no就行了
doc.add(new Field("postTime", postTime, Field.Store.YES, Field.Index.ANALYZED, Field.TermVector.NO));
//添加主键至文档,不分词,不高亮。
doc.add(new Field("id" , id , Field.Store.YES , Field.Index.NO , Field.TermVector.NO));
doc.add(new Field("bid" , bid.toString() , Field.Store.YES , Field.Index.NO , Field.TermVector.NO));
searchDao.save(doc, indexWriter);
}
indexWriter.close();
}
//搜索方法
public PageList searchFourm(String which, String keyWord, Pages pages) throws Exception{
System.out.println("pages.getSpage()="+pages.getSpage());
QueryResult queryResult = this.searchFourm(which, keyWord, pages.getSpage(), pages.getPerPageNum());
PageList pl = new PageList();
if(pages.getTotalNum()==-1){
pages.setTotalNum(queryResult.getRecordCount());
}
pages.executeCount();
queryResult = this.searchFourm(which, keyWord, pages.getSpage(), pages.getPerPageNum());
System.out.println("queryResult.getRecordList()="+queryResult.getRecordList());
pl.setObjectList(queryResult.getRecordList());
pl.setPages(pages);
return pl;
}
public QueryResult searchFourm(String which, String keyWord, int firstResult, int maxResult) throws Exception {
File indexFile = new File("D:\\index");
IndexReader reader = IndexReader.open(indexFile);
//庖丁解牛分词器
Analyzer analyzer = new PaodingAnalyzer();
//指定对content还是title进行查询
QueryParser queryParser = new QueryParser(which, analyzer);
IndexSearcher indexSearcher = new IndexSearcher(reader);
//对用户的输入进行查询
Query query = queryParser.parse(keyWord);
TopDocs topDocs = searchDao.search(query, indexSearcher);
int recordCount = topDocs.totalHits;
//高亮htmlFormatter对象
SimpleHTMLFormatter sHtmlF = new SimpleHTMLFormatter("<font color='red'>", "</font>");
//高亮对象
Highlighter highlighter = new Highlighter(sHtmlF, new QueryScorer(query));
//设置高亮附近的字数
highlighter.setTextFragmenter(new SimpleFragmenter(100));
List<ForumSearch> recordList = new ArrayList<ForumSearch>();
// 3,取出当前页的数据
int end = Math.min(firstResult + maxResult, topDocs.totalHits);
for(int i = firstResult; i < end; i++){
ScoreDoc scoreDoc = topDocs.scoreDocs[i];
int docSn = scoreDoc.doc;//文档内部编号
Document doc = indexSearcher.doc(docSn); // 根据编号取出相应的文档
//取得该条索引文档
ForumSearch fs = new ForumSearch();
String title = doc.get("title");
String detail = doc.get("detail");
String id = doc.get("id");
String bid = doc.get("bid");
String postTime = doc.get("postTime");
if(which.equals("title")){
String bestFragment = highlighter.getBestFragment(analyzer, which, title);
//获得高亮后的标题内容
fs.setTitle(bestFragment);
//如果内容不足100字,全部设置
if(detail.length()<100){
fs.setDetail(detail);
}else{
fs.setDetail(detail.substring(0, 100));
}
}else{
//如果查询内容
String bestFragment = highlighter.getBestFragment(analyzer, which, detail);
//取得高亮内容并设置
fs.setDetail(bestFragment);
fs.setTitle(title);
}
//设置日期
fs.setPostTime(postTime);
fs.setId(id);
fs.setBid(bid);
recordList.add(fs);
}
return new QueryResult(recordCount, recordList);
}
public SearchDao getSearchDao() {
return searchDao;
}
public void setSearchDao(SearchDao searchDao) {
this.searchDao = searchDao;
}
}
ForumSearchServiceImpl只写了两个方法,一个就是创建索引,另一个就是搜索.其中为了返回数据方便,定义了一个QueryResult结构,里面只有两个变量,这里就不粘代码了,能看懂的人自然知道那两个变量是什么,注意,这里面的方法模仿了Hibernate分页,接下来将我的Action层的代码贴出来:
public String createIndex() throws Exception{
List<Forum> forumList = forumService.listForums();
forumSearchService.saveForumIndex(forumList);
return SUCCESS;
}
public String search() throws Exception {
Pages pages = new Pages();
pages.setPerPageNum(10);
pages.setPage(this.getPage());
pages.setFileName(basePath + "forumSearch.bbsns?action=search"+"&keyWord="+keyWord+"&which="+which);
this.setPageList(forumSearchService.searchFourm(which, keyWord, pages));
return "result";
}
其实至于搜索和创建索引,大家可以完善很多功能,比如多关键字,多条件查询,方法只要实现好,创建好索引就行了,可以实现高级搜索等功能.最近有点懒,大概写了这么点,哎...
好了,大概差不多了,就到这里吧,如果各位有什么问题,欢迎留言探讨.
原创首发,谢谢支持!