Lucene 的四大索引查询 ——bool 域搜索 通配符 范围搜索

Lucene 的四大索引查询  
清单1：使用布尔操作符  

Java代码     

1. //Test boolean operator  
2. blic void testOperator(String indexDirectory) throws Exception{  
3.  Directory dir = FSDirectory.getDirectory(indexDirectory,false);  
4.  IndexSearcher indexSearcher = new IndexSearcher(dir);  
5.  String[] searchWords = {"Java AND Lucene", "Java NOT Lucene", "Java OR Lucene",   
6.                   "+Java +Lucene", "+Java -Lucene"};  
7.  Analyzer language = new StandardAnalyzer();  
8.  Query query;  
9.  for(int i = 0; i < searchWords.length; i++){  
10.     query = QueryParser.parse(searchWords[i], "title", language);  
11.     Hits results = indexSearcher.search(query);  
12.     System.out.println(results.length() + "search results for query " + searchWords[i]);  
13.  }

域搜索(Field Search)  

Lucene 支持域搜索，你可以指定一次查询是在哪些域(Field)上进行。例如，如果索引的文档包含两个域，Title 和 Content，你就可以使用查询 “Title: Lucene AND Content: Java” 来返回所有在 Title 域上包含 Lucene 并且在 Content 域上包含 Java 的文档。清单 2 显示了如何利用 Lucene 的 API 来实现域搜索。  


清单2：实现域搜索  

Java代码     

1. //Test field search  
2. public void testFieldSearch(String indexDirectory) throws Exception{  
3.     Directory dir = FSDirectory.getDirectory(indexDirectory,false);  
4.     IndexSearcher indexSearcher = new IndexSearcher(dir);  
5.     String searchWords = "title:Lucene AND content:Java";  
6.     Analyzer language = new StandardAnalyzer();  
7.     Query query = QueryParser.parse(searchWords, "title", language);  
8.     Hits results = indexSearcher.search(query);  
9.     System.out.println(results.length() + "search results for query " + searchWords);  
10. }  
11.

通配符搜索(Wildcard Search)  

Lucene 支持两种通配符：问号（？）和星号（*）。你可以使用问号（？）来进行单字符的通配符查询，或者利用星号（*）进行多字符的通配符查询。例如，如果你想搜索 tiny 或者 tony，你就可以使用查询语句 “t?ny”；如果你想查询 Teach, Teacher 和 Teaching，你就可以使用查询语句 “Teach*”。清单3 显示了通配符查询的过程。  


清单3：进行通配符查询  

Java代码     

1. //Test wildcard search  
2. public void testWildcardSearch(String indexDirectory)throws Exception{  
3.    Directory dir = FSDirectory.getDirectory(indexDirectory,false);  
4.    IndexSearcher indexSearcher = new IndexSearcher(dir);  
5.    String[] searchWords = {"tex*", "tex?", "?ex*"};  
6.    Query query;  
7.    for(int i = 0; i < searchWords.length; i++){  
8.       query = new WildcardQuery(new Term("title",searchWords[i]));  
9.       Hits results = indexSearcher.search(query);  
10.       System.out.println(results.length() + "search results for query " + searchWords[i]);  
11.    }  
12. }  
13.

模糊查询  

Lucene 提供的模糊查询基于编辑距离算法(Edit distance algorithm)。你可以在搜索词的尾部加上字符 ~ 来进行模糊查询。例如，查询语句 “think~” 返回所有包含和 think 类似的关键词的文档。清单 4 显示了如果利用 Lucene 的 API 进行模糊查询的代码。  


清单4：实现模糊查询  

Java代码     

1. //Test fuzzy search  
2. public void testFuzzySearch(String indexDirectory)throws Exception{  
3.    Directory dir = FSDirectory.getDirectory(indexDirectory,false);  
4.    IndexSearcher indexSearcher = new IndexSearcher(dir);  
5.    String[] searchWords = {"text", "funny"};  
6.    Query query;  
7.    for(int i = 0; i < searchWords.length; i++){  
8.       query = new FuzzyQuery(new Term("title",searchWords[i]));  
9.       Hits results = indexSearcher.search(query);  
10.       System.out.println(results.length() + "search results for query " + searchWords[i]);  
11.    }  
12. }  
13.

范围搜索(Range Search)  

范围搜索匹配某个域上的值在一定范围的文档。例如，查询 “age:[18 TO 35]” 返回所有 age 域上的值在 18 到 35 之间的文档。清单5显示了利用 Lucene 的 API 进行返回搜索的过程。  


清单5：测试范围搜索  

Java代码     

1. //Test range search  
2. public void testRangeSearch(String indexDirectory)throws Exception{  
3.     Directory dir = FSDirectory.getDirectory(indexDirectory,false);  
4.     IndexSearcher indexSearcher = new IndexSearcher(dir);  
5.     Term begin = new Term("birthDay","20000101");  
6.     Term end   = new Term("birthDay","20060606");  
7.     Query query = new RangeQuery(begin,end,true);  
8.     Hits results = indexSearcher.search(query);  
9.     System.out.println(results.length() + "search results is returned");  
10. }