主题：程序代码——可以对Heritrix进行索引和检索的Lucene程序

import java.awt.Dimension;
import java.awt.Toolkit;
import java.awt.event.ActionEvent;
import java.awt.event.ActionListener;
import java.io.File;
import java.io.FileReader;
import java.io.IOException;

import javax.swing.JFileChooser;
import javax.swing.JFrame;
import javax.swing.JMenu;
import javax.swing.JMenuBar;
import javax.swing.JMenuItem;
import javax.swing.JOptionPane;
import javax.swing.JScrollPane;
import javax.swing.JTextArea;
import javax.swing.ScrollPaneConstants;

import org.apache.lucene.analysis.standard.StandardAnalyzer;
import org.apache.lucene.document.Document;
import org.apache.lucene.document.Field;
import org.apache.lucene.index.IndexWriter;
import org.apache.lucene.queryParser.QueryParser;
import org.apache.lucene.search.Hits;
import org.apache.lucene.search.IndexSearcher;
import org.apache.lucene.search.Query;
import org.apache.lucene.store.Directory;
import org.apache.lucene.store.FSDirectory;

//测试类
public class Exec {
    public static void main(String[] args) {
        // 创建窗体类变量
        DemoWindow dw = new DemoWindow("基于Lucene的文件全文搜索系统");

        // 将窗体的宽度和高度分别设置为屏幕宽度和屏幕高度的1/3，左上角位置也设置为屏幕宽度和屏幕高度的1/3处
        Toolkit theKit = dw.getToolkit();
        Dimension wndSize = theKit.getScreenSize();
        dw.setBounds(wndSize.width / 3, wndSize.height / 3, wndSize.width / 3,
                wndSize.height / 3);

        // 点击关闭按钮可以退出程序
        dw.setDefaultCloseOperation(JFrame.EXIT_ON_CLOSE);

        // 设置窗体为可见
        dw.setVisible(true);
    }
}

// 界面窗体
class DemoWindow extends JFrame implements ActionListener {
    // 显示命中文件名称的文本区
    JTextArea jta = new JTextArea();

    // 设置文本区的滚动条
    int v = ScrollPaneConstants.VERTICAL_SCROLLBAR_AS_NEEDED;
    int h = ScrollPaneConstants.HORIZONTAL_SCROLLBAR_AS_NEEDED;
    JScrollPane jsp = new JScrollPane(jta, v, h);

    // 菜单
    JMenuBar menuBar = new JMenuBar();
    JMenu index = new JMenu("建立索引");
    JMenuItem indexSource = new JMenuItem("指定需要索引的目录");
    JMenuItem indexDestination = new JMenuItem("指定存放索引的目录");
    JMenuItem createIndex = new JMenuItem("建立索引");
    JMenu search = new JMenu("搜索");
    JMenuItem keywordsSearch = new JMenuItem("关键词检索");

    // Lucene索引器
    Indexer indexer = null;

    // Lucene搜索器
    Searcher searcher = new Searcher();

    // 需要建立索引的文件路径
    String indexSourceStr = null;

    // 存放索引的路径
    String indexDestinationStr = null;

    // 构造函数
    public DemoWindow(String title) {
        super(title);

        // 添加文本区
        add(jsp);

        // 添加菜单
        index.add(indexSource);
        index.add(indexDestination);
        index.addSeparator();
        index.add(createIndex);
        search.add(keywordsSearch);
        menuBar.add(index);
        menuBar.add(search);
        setJMenuBar(menuBar);

        // 取消部分菜单项的可用功能，这些菜单会随着操作的不断推进而逐次可用
        indexDestination.setEnabled(false);
        createIndex.setEnabled(false);
        keywordsSearch.setEnabled(false);

        // 添加事件监听器
        indexSource.addActionListener(this);
        indexDestination.addActionListener(this);
        createIndex.addActionListener(this);
        keywordsSearch.addActionListener(this);
    }

    // 响应单击菜单
    public void actionPerformed(ActionEvent e) {
        // 设置需要建立索引的文件所在目录
        if (e.getSource() == indexSource) {
            // 设置文件选择对话框
            JFileChooser jfc = new JFileChooser();

            // 该文件选择对话框只能打开文件目录
            jfc.setFileSelectionMode(JFileChooser.DIRECTORIES_ONLY);
            if (jfc.showOpenDialog(this) == JFileChooser.APPROVE_OPTION) {
                // 将“指定存放索引的目录”菜单项设置为可用
                indexDestination.setEnabled(true);

                // 保存需要建立索引的文件路径
                indexSourceStr = jfc.getSelectedFile().getPath();
            }
        }
        // 设置保存索引的目录
        else if (e.getSource() == indexDestination) {
            // 设置文件选择对话框
            JFileChooser jfc = new JFileChooser();

            // 该文件选择对话框只能打开文件目录
            jfc.setFileSelectionMode(JFileChooser.DIRECTORIES_ONLY);
            if (jfc.showOpenDialog(this) == JFileChooser.APPROVE_OPTION) {
                // 将“建立索引”菜单项设置为可用
                createIndex.setEnabled(true);

                // 保存存放索引的路径
                indexDestinationStr = jfc.getSelectedFile().getPath();
            }
        }
        // 处理建立索引
        else if (e.getSource() == createIndex) {
            // 建立索引器类变量
            indexer = new Indexer(indexSourceStr, indexDestinationStr);

            // 在新线程中建立索引
            Thread t = new Thread(new ProcessIndex(indexer, keywordsSearch));
            t.start();
        }
        // 处理搜索
        else if (e.getSource() == keywordsSearch) {
            // 显示输入信息对话框
            String queryWords = JOptionPane.showInputDialog("请输入搜索词语：");

            // 在文本区输出命中的文件名称结果
            jta.setText(searcher.getResults(queryWords, indexDestinationStr));
        }
    }
}

// 建立索引的线程（由于索引比较耗时，所以单独使用一个线程来建立索引）
class ProcessIndex implements Runnable {
    // 索引器类变量
    Indexer indexer;

    // 当索引结束后，设置该菜单项可用
    JMenuItem jmi;

    // 构造函数
    public ProcessIndex(Indexer indexer, JMenuItem jmi) {
        this.indexer = indexer;
        this.jmi = jmi;
    }

    // 线程的工作方法，启动索引工作，并在索引建立后，设置该菜单项可用
    public void run() {
        try {
            indexer.start();
            jmi.setEnabled(true);
        } catch (Exception e) {
            System.out.println(e.getMessage());
        }
    }
}

// 索引器
class Indexer {
    // 需要建立索引的文件路径
    String indexSourceStr;

    // 存放索引的路径
    String indexDestinationStr;

    // 构造函数
    public Indexer(String indexSourceStr, String indexDestinationStr) {
        this.indexSourceStr = indexSourceStr;
        this.indexDestinationStr = indexDestinationStr;
    }

    // 调用index建立索引的函数
    public void start() {
        try {
            File dataDir = new File(indexSourceStr);
            File indexDir = new File(indexDestinationStr);
            index(indexDir, dataDir);
        } catch (Exception e) {
            System.out.println(e.getMessage());
        }
    }

    // 调用indexDirectory函数建立索引的函数
    public void index(File indexDir, File dataDir) throws IOException {
        // 建立索引写入器
        // indexDir表示索引的存放地点
        // new StandardAnalyzer()表示分析器
        // true表示如已有索引则覆盖之
        IndexWriter writer = new IndexWriter(indexDir, new StandardAnalyzer(),
                true);

        // 表示生成多文件索引文件（默认为复合索引）
        writer.setUseCompoundFile(false);

        // 递归调用indexDirectory函数
        indexDirectory(writer, dataDir);

        // 优化索引
        writer.optimize();
        writer.close();
    }

    // 被递归调用的建立索引函数
    private void indexDirectory(IndexWriter writer, File dir)
            throws IOException {
        // 得到文件列表
        File[] files = dir.listFiles();

        // 循环所有文件
        for (int i = 0; i < files.length; i++) {
            File f = files[i];

            // 如果是目录，则递归调用自己，否则如果文件扩展名为html，则调用indexFile函数对此文件建立索引
            if (f.isDirectory()) {
                indexDirectory(writer, f);
            } else if (f.getName().endsWith(".html")) {
                indexFile(writer, f);
            }
        }
    }

    // 对指定文件建立索引的函数
    private void indexFile(IndexWriter writer, File f) throws IOException {
        // 如果文件隐藏、不存在或者不可读取，则退出函数
        if (f.isHidden() || !f.exists() || !f.canRead()) {
            return;
        }

        // 在控制台输出建立索引的文件信息
        System.out.println("Indexing " + f.getCanonicalPath());

        // 建立索引关键代码
        // 建立Document类变量
        Document doc = new Document();

        // 添加contents索引字段，内容为文件内容
        doc.add(Field.Text("contents", new FileReader(f)));

        // 添加url索引字段，内容为文件名称
        String filename = f.getCanonicalPath();
        String url = filename.substring(filename.lastIndexOf("mirror") + 7);
        doc.add(Field.Keyword("url", url));

        // 写入索引
        writer.addDocument(doc);
    }
}

// 搜索器
class Searcher {
    // 根据搜索词和索引文件路径进行搜索的函数
    public String getResults(String queryWords, String indexDestinationStr) {
        // 返回结果
        StringBuffer results = new StringBuffer();

        try {
            // 打开索引
            File indexDir = new File(indexDestinationStr);
            Directory fsDir = FSDirectory.getDirectory(indexDir, false);
            IndexSearcher is = new IndexSearcher(fsDir);

            // 解析查询词语
            Query query = QueryParser.parse(queryWords, "contents",
                    new StandardAnalyzer());
            // Query query = new TermQuery(new Term("contents", "java"));
            // Query query = new TermQuery(new Term("contents", "application"));
            // Query query = new TermQuery(new Term("contents", "java AND NOT
            // application"));

            // 搜索索引
            Hits hits = is.search(query);

            // 输出搜索结果
            for (int i = 0; i < hits.length(); i++) {
                Document doc = hits.doc(i);
                results.append(doc.get("url") + "\n");
            }
        } catch (Exception e) {
            System.out.println(e.getMessage());
        }
        return results.toString();
    }
}