采集Google的URL方式多重多样，最简单的方式莫过于js直接获取节点了。比如：

var h3 = document.getElementsByTagName('h3');
for(var i=0;i<h3.length;i++){
    var a = h3[i]. getElementsByTagName('a');
    console.log(a[0].href);
}

在java里面用jsoup也可以非常简单的获取到搜索结果的URL：

public static void main(String[] args) throws IOException {
		Document doc = Jsoup.connect("https://www.google.ws/search?num=100&site=&source=hp&q=filetype%3Ajsp&oq=filetype%3Ajsp&gs_l=hp.3...8115.14780.0.15194.22.21.1.0.0.0.523.5187.3j3j3j5j4j1.19.0....0...1c.1.36.hp..14.8.1440.P_2EQhc7Pz0").userAgent("Googlebot/2.1 (+http://www.googlebot.com/bot.html)").timeout(5000).get();
		Elements element = doc.getElementsByTag("h3");
		for (Element e : element) {
			Matcher m= Pattern.compile("/url\\?q=(.*)&sa").matcher(e.getElementsByTag("a").get(0).attr("href"));
			if(m.find()){
				System.out.println(URLDecoder.decode(m.group(1),"UTF-8"));
			}
		}
	}

正则：

package org.javaweb.test;
import java.util.regex.Matcher;
import java.util.regex.Pattern;
public class TestReg {
    public static void main(String[] args) {
        String source="<h3 class=\"r\"><a href=\"http://baidu.com\">百度</a></h3><h3 class=\"r\"><a href=\"http://google.com\">谷歌</a></h3> ";
        StringBuilder resultComment=new StringBuilder();
        StringBuilder resultName=new StringBuilder();
        System.out.println("=======开始匹配========");
        String patternStrs="(<h3 class=\"r\"><a.+?)href=\"(.+?)\">(.+?)(</a></h3>)";
        Pattern pattern=Pattern.compile(patternStrs);
        Matcher matcher=pattern.matcher(source);
        while(matcher.find()){
            resultName.append(matcher.group(2)+"\n");
            resultComment.append(matcher.group(3)+"\n");
        }
        System.out.println("=======标签内内容=======");
        System.out.println(resultComment.toString());
        System.out.println("=======name属性值=======");
        System.out.println(resultName.toString());
    }
}