遍历整个HTML文档的链接 : HTML文档 « Swing


import java.io.BufferedReader;

import java.io.InputStream;

import java.io.InputStreamReader;

import java.net.URL;

import java.net.URLConnection;



import javax.swing.text.AttributeSet;

import javax.swing.text.html.HTML;

import javax.swing.text.html.HTMLDocument;

import javax.swing.text.html.HTMLEditorKit;

import javax.swing.text.html.parser.ParserDelegator;



public class MainClass {



  public static void main(String args[]) throws Exception {

    URL url = new URL("http://www.google.com");

    URLConnection connection = url.openConnection();

    InputStream is = connection.getInputStream();

    InputStreamReader isr = new InputStreamReader(is);

    BufferedReader br = new BufferedReader(isr);



    HTMLEditorKit htmlKit = new HTMLEditorKit();

    HTMLDocument htmlDoc = (HTMLDocument) htmlKit.createDefaultDocument();

    HTMLEditorKit.Parser parser = new ParserDelegator();

    HTMLEditorKit.ParserCallback callback = htmlDoc.getReader(0);

    parser.parse(br, callback, true);



    for (HTMLDocument.Iterator iterator = htmlDoc.getIterator(HTML.Tag.A); iterator.isValid(); iterator

        .next()) {



      AttributeSet attributes = iterator.getAttributes();

      String srcString = (String) attributes.getAttribute(HTML.Attribute.HREF);

      System.out.print(srcString);

      int startOffset = iterator.getStartOffset();

      int endOffset = iterator.getEndOffset();

      int length = endOffset - startOffset;

      String text = htmlDoc.getText(startOffset, length);

      System.out.println(" - " + text);

    }

  }

}

url?sa=p&pref=ig&pval=3&q=http://www.google.ca/ig%3Fhl%3Den&usg=__o-KrRDBI3nbRElKzYEMqfOl3_t0= – Personalize this page https://www.google.com/accounts/Login?continue=http://www.google.ca/&hl=en – Sign in http://images.google.ca/imghp?ie=ISO-8859-1&oe=ISO-8859-1&hl=en&tab=wi – Images http://groups.google.ca/grphp?ie=ISO-8859-1&oe=ISO-8859-1&hl=en&tab=wg – Groups http://news.google.ca/nwshp?ie=ISO-8859-1&oe=ISO-8859-1&hl=en&tab=wn – News /maps?ie=ISO-8859-1&oe=ISO-8859-1&hl=en&tab=wl – Maps http://scholar.google.com/schhp?ie=ISO-8859-1&oe=ISO-8859-1&hl=en&tab=ws – Scholar /intl/en/options/ – more » /advanced_search?hl=en – Advanced Search /preferences?hl=en – Preferences /language_tools?hl=en – Language Tools

遍历整个HTML文档的链接 : HTML文档 « Swing « Java 教程