Lucene是一个高性能的,可扩展的信息检索库,能让你轻松到在自己的应用中增加全文检索功能。简单的说,Lucene就是实现了搜索引擎功能的基础库。
Lucene是一个高性能的,可扩展的信息检索库,能让你轻松到在自己的应用中增加全文检索功能。简单的说,Lucene就是实现了搜索引擎功能的基础库。搜索引擎就一个功能,根据网名输入的Query,找到相关的文档。文档成千上万上亿,直接字符串匹配顺序查找,慢的像蜗牛一样,谁都受不了,所以聪明人发明了索引(Index)。所谓索引,简单点说就是建立了一些词到一些文档的映射,能迅速的根据词找到相关的文档。所以Lucene就两个基本功能:(1)对文档建索引;(2)根据网名输入,通过索引迅速的找到相关的文档。
索引是现代搜索引擎的核心,建立索引的过程就是把源数据处理成非常方便查询的索引文件的过程。为什么索引这么重要呢,试想你现在要在大量的文档中搜索含有某个关键词的文档,那么如果不建立索引的话你就需要把这些文档顺序的读入内存,然后检查这个文章中是不是含有要查找的关键词,这样的话就会耗费非常多的时间,想想搜索引擎可是在毫秒级的时间内查找出要搜索的结果的。这就是由于建立了索引的原因,你可以把索引想象成这样一种数据结构,他能够使你快速的随机访问存储在索引中的关键词,进而找到该关键词所关联的文档。Lucene 采用的是一种称为反向索引(inverted index)的机制。反向索引就是说我们维护了一个词 / 短语表,对于这个表中的每个词 / 短语,都有一个链表描述了有哪些文档包含了这个词 / 短语。这样在用户输入查询条件的时候,就能非常快的得到搜索结果。
为了对文档进行索引,Lucene 提供了五个基础的类,他们分别是 Document, Field, IndexWriter, Analyzer, Directory。下面我们分别介绍一下这五个类的用途:
Document
Document 是用来描述文档的,这里的文档可以指一个 HTML 页面,一封电子邮件,或者是一个文本文件。一个 Document 对象由多个 Field 对象组成的。可以把一个 Document 对象想象成数据库中的一个记录,而每个 Field 对象就是记录的一个字段。
Field
Field 对象是用来描述一个文档的某个属性的,比如一封电子邮件的标题和内容可以用两个 Field 对象分别描述。
Analyzer
在一个文档被索引之前,首先需要对文档内容进行分词处理,这部分工作就是由 Analyzer 来做的。Analyzer 类是一个抽象类,它有多个实现。针对不同的语言和应用需要选择适合的 Analyzer。Analyzer 把分词后的内容交给 IndexWriter 来建立索引。
IndexWriter
IndexWriter 是 Lucene 用来创建索引的一个核心的类,他的作用是把一个个的 Document 对象加到索引中来。
Directory
这个类代表了 Lucene 的索引的存储的位置,这是一个抽象类,它目前有两个实现,第一个是 FSDirectory,它表示一个存储在文件系统中的索引的位置。第二个是 RAMDirectory,它表示一个存储在内存当中的索引的位置。
下面是一个简单的建索引的代码:
public class IndexFiles {
private IndexFiles() {}/** Index all text files under a directory. */
public static void main(String[] args) {
String usage = "java org.apache.lucene.demo.IndexFiles"
+ " [-index INDEX_PATH] [-docs DOCS_PATH] [-update]\n\n"
+ "This indexes the documents in DOCS_PATH, creating a Lucene index"
+ "in INDEX_PATH that can be searched with SearchFiles";
String indexPath = "index";
String docsPath = null;
boolean create = true;
for(int i=0;i<args.length;i++) {
//写索引的目录
if ("-index".equals(args[i])) {
indexPath = args[i+1];
i++;
//需建索引的文档目录
} else if ("-docs".equals(args[i])) {
docsPath = args[i+1];
i++;
//指明是更新还是新建
} else if ("-update".equals(args[i])) {
create = false;
}
}if (docsPath == null) {
System.err.println("Usage: " + usage);
System.exit(1);
}final File docDir = new File(docsPath);
if (!docDir.exists() || !docDir.canRead()) {
System.out.println("Document directory '" +docDir.getAbsolutePath()+ "' does not exist or is not readable, please check the path");
System.exit(1);
}
Date start = new Date();
try {
System.out.println("Indexing to directory '" + indexPath + "'...");
//写索引的目录
Directory dir = FSDirectory.open(new File(indexPath));
//实例化一个分析器
Analyzer analyzer = new StandardAnalyzer(Version.LUCENE_40);
//索引Writer配置,参数为分析器
IndexWriterConfig iwc = new IndexWriterConfig(Version.LUCENE_40, analyzer);if (create) {
// Create a new index in the directory, removing any
// previously indexed documents:
iwc.setOpenMode(OpenMode.CREATE);
} else {
// Add new documents to an existing index:
iwc.setOpenMode(OpenMode.CREATE_OR_APPEND);
}// Optional: for better indexing performance, if you
// are indexing many documents, increase the RAM
// buffer. But if you do this, increase the max heap
// size to the JVM (eg add -Xmx512m or -Xmx1g):
//
// iwc.setRAMBufferSizeMB(256.0);//索引Writer,参数为索引目录以及索引Writer配置
IndexWriter writer = new IndexWriter(dir, iwc);
//建立索引,参数为索引Writer以及需建索引的文档目录
indexDocs(writer, docDir);// NOTE: if you want to maximize search performance,
// you can optionally call forceMerge here. This can be
// a terribly costly operation, so generally it's only
// worth it when your index is relatively static (ie
// you're done adding documents to it):
//
// writer.forceMerge(1);writer.close();
Date end = new Date();
System.out.println(end.getTime() - start.getTime() + " total milliseconds");} catch (IOException e) {
System.out.println(" caught a " + e.getClass() +
"\n with message: " + e.getMessage());
}
}/**
* Indexes the given file using the given writer, or if a directory is given,
* recurses over files and directories found under the given directory.
*
* NOTE: This method indexes one document per input file. This is slow. For good
* throughput, put multiple documents into your input file(s). An example of this is
* in the benchmark module, which can create "line doc" files, one document per line,
* using the
* <a href="../../../../../contrib-benchmark/org/apache/lucene/benchmark/byTask/tasks/WriteLineDocTask.html"
* >WriteLineDocTask</a>.
*
* @param writer Writer to the index where the given file/dir info will be stored
* @param file The file to index, or the directory to recurse into to find files to index
* @throws IOException If there is a low-level I/O error
*/
static void indexDocs(IndexWriter writer, File file)
throws IOException {if (file.canRead()) {
//如果是目录,进行递归调用
if (file.isDirectory()) {
String[] files = file.list();
if (files != null) {
for (int i = 0; i < files.length; i++) {
indexDocs(writer, new File(file, files[i]));
}
}
} else {
//对文件建索引
FileInputStream fis;
try {
fis = new FileInputStream(file);
} catch (FileNotFoundException fnfe) {
return;
}try {
//
Document doc = new Document();// Add the path of the file as a field named "path". Use a
// field that is indexed (i.e. searchable), but don't tokenize
// the field into separate words and don't index term frequency
// or positional information:
Field pathField = new StringField("path", file.getPath(), Field.Store.YES);
doc.add(pathField);// Add the last modified date of the file a field named "modified".
// Use a LongField that is indexed (i.e. efficiently filterable with
// NumericRangeFilter). This indexes to milli-second resolution, which
// is often too fine. You could instead create a number based on
// year/month/day/hour/minutes/seconds, down the resolution you require.
// For example the long value 2011021714 would mean
// February 17, 2011, 2-3 PM.
doc.add(new LongField("modified", file.lastModified(), Field.Store.NO));// Add the contents of the file to a field named "contents". Specify a Reader,
// so that the text of the file is tokenized and indexed, but not stored.
// Note that FileReader expects the file to be in UTF-8 encoding.
// If that's not the case searching for special characters will fail.
doc.add(new TextField("contents", new BufferedReader(new InputStreamReader(fis, "UTF-8"))));if (writer.getConfig().getOpenMode() == OpenMode.CREATE) {
// New index, so we just add the document (no old document can be there):
System.out.println("adding " + file);
writer.addDocument(doc);
} else {
// Existing index (an old copy of this document may have been indexed) so
// we use updateDocument instead to replace the old one matching the exact
// path, if present:
System.out.println("updating " + file);
writer.updateDocument(new Term("path", file.getPath()), doc);
}
} finally {
fis.close();
}
}
}
}
}
利用 Lucene 进行搜索就像建立索引一样也是非常方便的。在上面一部分中,我们已经为一个目录下的文本文档建立好了索引,现在我们就要在这个索引上进行搜索以找到包含某个关键词或短语的文档。Lucene 提供了几个基础的类来完成这个过程,它们分别是呢 IndexSearcher, Term, Query, TermQuery, Hits. 下面我们分别介绍这几个类的功能。
Query
这是一个抽象类,他有多个实现,比如 TermQuery, BooleanQuery, PrefixQuery. 这个类的目的是把用户输入的查询字符串封装成 Lucene 能够识别的 Query。
Term
Term 是搜索的基本单位,一个 Term 对象有两个 String 类型的域组成。生成一个 Term 对象可以有如下一条语句来完成:Term term = new Term(“fieldName”,”queryWord”); 其中第一个参数代表了要在文档的哪一个 Field 上进行查找,第二个参数代表了要查询的关键词。
TermQuery
TermQuery 是抽象类 Query 的一个子类,它同时也是 Lucene 支持的最为基本的一个查询类。生成一个 TermQuery 对象由如下语句完成: TermQuery termQuery = new TermQuery(new Term(“fieldName”,”queryWord”)); 它的构造函数只接受一个参数,那就是一个 Term 对象。
IndexSearcher
IndexSearcher 是用来在建立好的索引上进行搜索的。它只能以只读的方式打开一个索引,所以可以有多个 IndexSearcher 的实例在一个索引上进行操作。
Hits
Hits 是用来保存搜索的结果的。
下面是简单的查询代码:
public class SearchFiles {
private SearchFiles() {}
/** Simple command-line based search demo. */
public static void main(String[] args) throws Exception {
String usage =
"Usage:\tjava org.apache.lucene.demo.SearchFiles [-index dir] [-field f] [-repeat n] [-queries file] [-query string] [-raw] [-paging hitsPerPage]\n\nSee http://lucene.apache.org/core/4_1_0/demo/ for details.";
if (args.length > 0 && ("-h".equals(args[0]) || "-help".equals(args[0]))) {
System.out.println(usage);
System.exit(0);
}String index = "index";
String field = "contents";
String queries = null;
int repeat = 0;
boolean raw = false;
String queryString = null;
int hitsPerPage = 10;
for(int i = 0;i < args.length;i++) {
if ("-index".equals(args[i])) {
index = args[i+1];
i++;
} else if ("-field".equals(args[i])) {
field = args[i+1];
i++;
} else if ("-queries".equals(args[i])) {
queries = args[i+1];
i++;
} else if ("-query".equals(args[i])) {
queryString = args[i+1];
i++;
} else if ("-repeat".equals(args[i])) {
repeat = Integer.parseInt(args[i+1]);
i++;
} else if ("-raw".equals(args[i])) {
raw = true;
} else if ("-paging".equals(args[i])) {
hitsPerPage = Integer.parseInt(args[i+1]);
if (hitsPerPage <= 0) {
System.err.println("There must be at least 1 hit per page.");
System.exit(1);
}
i++;
}
}
IndexReader reader = DirectoryReader.open(FSDirectory.open(new File(index)));
IndexSearcher searcher = new IndexSearcher(reader);
Analyzer analyzer = new StandardAnalyzer(Version.LUCENE_40);BufferedReader in = null;
if (queries != null) {
in = new BufferedReader(new InputStreamReader(new FileInputStream(queries), "UTF-8"));
} else {
in = new BufferedReader(new InputStreamReader(System.in, "UTF-8"));
}
QueryParser parser = new QueryParser(Version.LUCENE_40, field, analyzer);
while (true) {
if (queries == null && queryString == null) { // prompt the user
System.out.println("Enter query: ");
}String line = queryString != null ? queryString : in.readLine();
if (line == null || line.length() == -1) {
break;
}line = line.trim();
if (line.length() == 0) {
break;
}
Query query = parser.parse(line);
System.out.println("Searching for: " + query.toString(field));
if (repeat > 0) { // repeat & time as benchmark
Date start = new Date();
for (int i = 0; i < repeat; i++) {
searcher.search(query, null, 100);
}
Date end = new Date();
System.out.println("Time: "+(end.getTime()-start.getTime())+"ms");
}doPagingSearch(in, searcher, query, hitsPerPage, raw, queries == null && queryString == null);
if (queryString != null) {
break;
}
}
reader.close();
}/**
* This demonstrates a typical paging search scenario, where the search engine presents
* pages of size n to the user. The user can then go to the next page if interested in
* the next hits.
*
* When the query is executed for the first time, then only enough results are collected
* to fill 5 result pages. If the user wants to page beyond this limit, then the query
* is executed another time and all hits are collected.
*
*/
public static void doPagingSearch(BufferedReader in, IndexSearcher searcher, Query query,
int hitsPerPage, boolean raw, boolean interactive) throws IOException {
// Collect enough docs to show 5 pages
TopDocs results = searcher.search(query, 5 * hitsPerPage);
ScoreDoc[] hits = results.scoreDocs;
int numTotalHits = results.totalHits;
System.out.println(numTotalHits + " total matching documents");int start = 0;
int end = Math.min(numTotalHits, hitsPerPage);
while (true) {
if (end > hits.length) {
System.out.println("Only results 1 - " + hits.length +" of " + numTotalHits + " total matching documents collected.");
System.out.println("Collect more (y/n) ?");
String line = in.readLine();
if (line.length() == 0 || line.charAt(0) == 'n') {
break;
}hits = searcher.search(query, numTotalHits).scoreDocs;
}
end = Math.min(hits.length, start + hitsPerPage);
for (int i = start; i < end; i++) {
if (raw) { // output raw format
System.out.println("doc="+hits[i].doc+" score="+hits[i].score);
continue;
}Document doc = searcher.doc(hits[i].doc);
String path = doc.get("path");
if (path != null) {
System.out.println((i+1) + ". " + path);
String title = doc.get("title");
if (title != null) {
System.out.println(" Title: " + doc.get("title"));
}
} else {
System.out.println((i+1) + ". " + "No path for this document");
}
}if (!interactive || end == 0) {
break;
}if (numTotalHits >= end) {
boolean quit = false;
while (true) {
System.out.print("Press ");
if (start - hitsPerPage >= 0) {
System.out.print("(p)revious page, ");
}
if (start + hitsPerPage < numTotalHits) {
System.out.print("(n)ext page, ");
}
System.out.println("(q)uit or enter number to jump to a page.");
String line = in.readLine();
if (line.length() == 0 || line.charAt(0)=='q') {
quit = true;
break;
}
if (line.charAt(0) == 'p') {
start = Math.max(0, start - hitsPerPage);
break;
} else if (line.charAt(0) == 'n') {
if (start + hitsPerPage < numTotalHits) {
start+=hitsPerPage;
}
break;
} else {
int page = Integer.parseInt(line);
if ((page - 1) * hitsPerPage < numTotalHits) {
start = (page - 1) * hitsPerPage;
break;
} else {
System.out.println("No such page");
}
}
}
if (quit) break;
end = Math.min(numTotalHits, start + hitsPerPage);
}
}
}
}