lucene搜索简单的实例

时间：2021-07-20 03:09:08

最近项目要用到lucene做全文检索，刚开始迷茫的，回去看了一下，自己写了简单的实例，希望对一些初学者能所帮助，高手请绕道。


项目环境：

1.导入jar包

KAnalyzer3.2.0Stable.jar
lucene-analyzers-3.0.1.jar
lucene-core-3.0.1.jar
lucene-highlighter-3.0.1.jar
lucene-memory-3.0.1.jar


2.写一个demo

  a.创建一个实体

DROP TABLE IF EXISTS `article`;
CREATE TABLE `article` (
  `id` int(11) NOT NULL AUTO_INCREMENT,
  `title` varchar(20) DEFAULT NULL,
  `content` varchar(5000) DEFAULT NULL,
  PRIMARY KEY (`id`)
) ENGINE=InnoDB DEFAULT CHARSET=utf8;


public class Article {
private int id;
private String title;
private String content;

................//get/set/toString方法省略

}

b.创建一个提供公共方法的类：

//提供封装分页数据的类：

@SuppressWarnings("rawtypes")
public class QueryResult {
private List list;
private int count;

..............get/set省略

}


//提供创建索引，目录的类

public class LuceneUtils{

private static Directory directory;// 建立索引库存储目录
private static Analyzer analyzer;// 创建分词器

private static IndexWriter indexWriter; // 在程序启动是初始化,建立索引

private static IndexSearcher indexSearcher;// 查询


static {
try {
// 加载配置文件lucene.properties，该文件中是创建索引库的路径"path=D:\\IindexSearch
Properties prop = new Properties();
InputStream inStream = LuceneUtils.class.getClassLoader().getResourceAsStream("lucene.properties");
//InputStream inStream = ClassLoader.getSystemResourceAsStream("lucene.properties");
prop.load(inStream);

directory = FSDirectory.open(new File(prop.getProperty("path")));
analyzer = new StandardAnalyzer(Version.LUCENE_30);
// 在程序启动是初始化,建立索引
indexWriter = new IndexWriter(directory, analyzer, MaxFieldLength.LIMITED);

//程序退出时关闭资源
Runtime.getRuntime().addShutdownHook(new Thread(){
public void run(){
try {
indexWriter.close();
} catch (Exception e) {
e.printStackTrace();
} 
}
});
} catch (Exception e) {
e.printStackTrace();
}
}


public static Document objectToDocument(Object obj) {
Article article = (Article) obj;
// 将文档转为domcment
Document doc = new Document();
String idstr = NumericUtils.intToPrefixCoded(article.getId());
doc.add(new Field("id", idstr, Store.YES, Index.NOT_ANALYZED));
doc.add(new Field("title", article.getTitle(), Store.YES, Index.ANALYZED));
doc.add(new Field("content", article.getContent(), Store.YES, Index.ANALYZED));
return doc;
}


public static Object documentToObject(Document doc) {
Article article = new Article();
//将Document转为Article
//将字符串转化为数字
int id = NumericUtils.prefixCodedToInt(doc.get("id"));
article.setId(id);
article.setTitle(doc.get("title"));
article.setContent(doc.get("content"));
return article;
}


public static IndexWriter getIndexWriter() {
return indexWriter;
}


public static IndexSearcher getIndexSearch() {
// 执行查询
try {
indexSearcher = new IndexSearcher(directory);
} catch (Exception e) {
throw new RuntimeException(e);
}
return indexSearcher;
}


public static Directory getDirectory() {
return directory;
}


public static Analyzer getAnalyzer() {
return analyzer;
}
}


c: 创建增删改查方法

public class IndexDao {


/**
* //保存到索引库
* 
* @return
* @throws Exception
*/
public void save(Article article) {
try {
// 将Aritcle转为Documnet
Document doc = LuceneUtils.objectToDocument(article);


// 建立索引
IndexWriter indexWriter = LuceneUtils.getIndexWriter();
indexWriter.addDocument(doc);
indexWriter.commit();
} catch (Exception e) {
throw new RuntimeException(e);
}
}


/**
* 删除索引库 Term 表示制定列中包含的关键字
* 
* @return
* @throws Exception
*/
public void delete(Article article) {

String idStr = NumericUtils.intToPrefixCoded(article.getId());
Term term = new Term("id", idStr);
try {
// 建立索引
IndexWriter indexWriter = LuceneUtils.getIndexWriter();
indexWriter.deleteDocuments(term);// 删除指定Term总重的documnet数据
indexWriter.commit();
} catch (Exception e) {
throw new RuntimeException(e);
}
}


/**
* 修改索引库
* 
* @return
* @throws Exception
*/
public void update(Article article) {

// 创建Term
String idStr = NumericUtils.intToPrefixCoded(article.getId());
Term term = new Term("id", idStr);


// 准备document
Document doc = LuceneUtils.objectToDocument(article);
try {
// 建立索引
IndexWriter indexWriter = LuceneUtils.getIndexWriter();
indexWriter.updateDocument(term, doc);// 删除指定Term总重的documnet数据
indexWriter.commit();


// 先删除，在创建
// indexWriter.deleteDocuments(term);
// indexWriter.addDocument(doc);
} catch (Exception e) {
throw new RuntimeException(e);
}
}


/**
* 查询索引库
* 
* @return
* @throws Exception
*/
public QueryResult query(String queryString, int first, int max) {
IndexSearcher indexSearcher = null;
try {
// MultiFieldQueryParser：表示可以根据多个字段查询
int totail = first + max;
// 1.把字符串转为Query对象
QueryParser parser = new MultiFieldQueryParser(Version.LUCENE_30, new String[] { "id", "title", "content" },
LuceneUtils.getAnalyzer());
Query query = parser.parse(queryString);

// 2.执行查询
indexSearcher = LuceneUtils.getIndexSearch();
// 指定排序条件
Sort sort = new Sort(new SortField("id", SortField.INT));// 按照id升序
TopDocs topDocs = indexSearcher.search(query, null, totail, sort);// 查询并返回最多的前n条数据
int count = topDocs.totalHits;// 总记录数
ScoreDoc[] scoreDoc = topDocs.scoreDocs;// 最多前n条结果数据


// 生成高亮显示器；设置前缀，后缀，摘要的大小
Formatter formatter = new SimpleHTMLFormatter("<font color='red'>", "</font>");
Scorer scorer = new QueryScorer(query);// 查询条件
Highlighter highlighter = new Highlighter(formatter, scorer);
highlighter.setTextFragmenter(new SimpleFragmenter(100));// 设置摘要的大小


// 3.取出数据
int endIndex = Math.min(totail, scoreDoc.length);
List<Article> list = new ArrayList<Article>();
for (int i = 0; i < endIndex; i++) {
// float score = scoreDoc[i].score;//平均得分
int docId = scoreDoc[i].doc;
Document doc = indexSearcher.doc(docId);


// 进行高亮操作,当没有找到关键词时，返回为null
String text = highlighter.getBestFragment(LuceneUtils.getAnalyzer(), "title", doc.get("title"));
if (text != null) {
doc.getField("title").setValue(text);
}
// 将Document转为Article
Article article = (Article) LuceneUtils.documentToObject(doc);
list.add(article);
}
QueryResult queryResult = new QueryResult(list, count);
return queryResult;
} catch (Exception e) {
throw new RuntimeException(e);
}
}



c;测试dao方法：


@Test
public void testSave() {
// 创建文档对象
Article article = new Article();
for (int i = 0; i < 20; i++) {
article.setId(i);
article.setTitle("Lucene搜索的方式");
article.setContent("全文检索是计算机程序通过扫描文章中的每一个词，对每一个词建立一个索引，指明该词在文章中出现的次数和位置。");
indexDao.save(article);
}
}


@Test
public void testDelete() {
Article article = new Article();
article.setId(1);
indexDao.delete(article);
}


@Test
public void testUpdate() {
// 创建文档对象
Article article = new Article();
article.setId(1);
article.setTitle("Lucene搜索的方式");
article.setContent("跟新索引库测试是否正确");
indexDao.update(article);
}

@Test
@SuppressWarnings("unchecked")
public void testQuery() {
String queryString = "Lucene";
QueryResult queryResult = indexDao.searchAndOrderBy(queryString, 0, 10);
System.out.println("count---------->" + queryResult.getCount());
List<Article> list = (List<Article>)queryResult.getList();
for(Article article:list){
System.err.println("list--------->" + article.toString());
}



到此lucenes测试例子完成。以下是本人使用过程中进行改造封装后，制作的模板，由于增删改中都要用到一些相同的代码，并且重复比较多，为此本人将重复的制作为一个方法，其他的不能确定的提供一个借口提供一个抽象方法，在调用者需要封装那些实体，即可调用接口中的方法并重写即可。此封装主要用到的是模板设计模式。

如下：

public class ComsLuceneUtils {
public interface CallBackQuery {

//不分页

//public abstract  List  documentToObject(Document doc, List list) throws Exception;

//分页方法
public abstract QueryResult documentToObject(Document doc, List list, int count) throws Exception;


public abstract Document objectToDocument(Object obj) throws Exception;
}


/**
* //保存到索引库
* CallBackQuery ：保存函数
* @return
* @throws Exception
*/
public static void save(Article article, CallBackQuery callBack) {
try {
// 将Aritcle转为Documnet
Document doc = callBack.objectToDocument(article);


// 建立索引
IndexWriter indexWriter = LuceneUtils.getIndexWriter();
indexWriter.addDocument(doc);
indexWriter.commit();
} catch (Exception e) {
throw new RuntimeException(e);
}
}


/**
* 删除索引库 Term 表示制定列中包含的关键字
* 
* @return
* @throws Exception
*/
public static void delete(Integer id) {


String idStr = NumericUtils.intToPrefixCoded(id);
Term term = new Term("id", idStr);
try {
// 建立索引
IndexWriter indexWriter = LuceneUtils.getIndexWriter();
indexWriter.deleteDocuments(term);// 删除指定Term总重的documnet数据
indexWriter.commit();
} catch (Exception e) {
throw new RuntimeException(e);
}
}


/**
* 修改索引库
* CallBackQuery ：更新函数
* @return
* @throws Exception
*/
public static void update(Article article, CallBackQuery callBack) {


// 创建Term
String idStr = NumericUtils.intToPrefixCoded(article.getId());
Term term = new Term("id", idStr);
try {
// 将Object转换为Document对象
Document doc = callBack.objectToDocument(article);


// 建立索引
IndexWriter indexWriter = LuceneUtils.getIndexWriter();
indexWriter.updateDocument(term, doc);// 删除指定Term数的documnet数据
indexWriter.commit();


// 先删除，在创建
// indexWriter.deleteDocuments(term);
// indexWriter.addDocument(doc);
} catch (Exception e) {
throw new RuntimeException(e);
}
}


/**查询分页方法
* 
* 查询索引库 querString: 查询字符串 
* first : 开始位置（分页功能）
* max : 最大数（分页功能） 
* parameter ：查询指定的字段 
* CallBackQuery ： 查询函数 
* QueryResult ： 返回结果集合
* @throws Exception
*/
public static QueryResult query(String queryString, int first, int max, String[] parameter, CallBackQuery callback) {
IndexSearcher indexSearcher = null;
try {
// MultiFieldQueryParser：表示可以根据多个字段查询
int totail = first + max;
// 1.把字符串转为Query对象
QueryParser parser = new MultiFieldQueryParser(Version.LUCENE_30, parameter, LuceneUtils.getAnalyzer());
Query query = parser.parse(queryString);


// 2.执行查询
indexSearcher = LuceneUtils.getIndexSearch();
// 指定排序条件
Sort sort = new Sort(new SortField("id", SortField.INT));// 按照id升序
TopDocs topDocs = indexSearcher.search(query, null, totail, sort);// 查询并返回最多的前n条数据
int count = topDocs.totalHits;// 总记录数
ScoreDoc[] scoreDoc = topDocs.scoreDocs;// 最多前n条结果数据


// 生成高亮显示器；设置前缀，后缀，摘要的大小
Formatter formatter = new SimpleHTMLFormatter("<font color='red'>", "</font>");
Scorer scorer = new QueryScorer(query);// 查询条件
Highlighter highlighter = new Highlighter(formatter, scorer);
highlighter.setTextFragmenter(new SimpleFragmenter(100));// 设置摘要的大小


// 3.取出数据
int endIndex = Math.min(totail, scoreDoc.length);
List<Article> list = new ArrayList<Article>();
for (int i = 0; i < endIndex; i++) {// start
// float score = scoreDoc[i].score;//平均得分
int docId = scoreDoc[i].doc;
Document doc = indexSearcher.doc(docId);


// 进行高亮操作
if (parameter.length > 0) {
for (int j = 0; j < parameter.length; j++) {
// 进行高亮操作,当没有找到关键词时，返回为null
String text = highlighter.getBestFragment(LuceneUtils.getAnalyzer(), parameter[j], doc.get(parameter[j]));
if (text != null) {
doc.getField(parameter[j]).setValue(text);
}
}
}


// 将Document转为Article
callback.documentToObject(doc, list, count);
}// end


// 返回结果集
QueryResult queryResult = new QueryResult(list, count);
return queryResult;
} catch (Exception e) {
throw new RuntimeException(e);
}
}

}



测试方法：

@Test
@SuppressWarnings("unchecked")
public void testQuery() {
String queryString = "Lucene";
String[] param = new String[]{"id","title","content"};

try {
QueryResult queryResult=ComsLuceneUtils.query(queryString, 0, 20, param, new cn.net.yixun.util.ComsLuceneUtils.CallBackQuery(){
public QueryResult documentToObject(Document doc,List list,int count)throws Exception{
Article article = new Article();
//将Document转为Article
//将字符串转化为数字
int id = NumericUtils.prefixCodedToInt(doc.get("id"));
article.setId(id);
article.setTitle(doc.get("title"));
article.setContent(doc.get("content"));
list.add(article);
QueryResult queryResult=new QueryResult(list, count);
return queryResult;
} 
public Document objectToDocument(Object obj)throws Exception{
return null;};
});

List<Article> list = (List<Article>)queryResult.getList();
for(Article article:list){
System.err.println("list--------->" + article.toString());
}
} catch (Exception e) {
e.printStackTrace();
}
}









标签：

相关文章