一个简易的网页爬虫,可用于下载在线API文档

一个简易的网页爬虫,可用于下载在线API文档

package wkx;
 
import java.io.File;
import java.io.FileNotFoundException;
import java.io.FileOutputStream;
import java.io.IOException;
import java.util.HashSet;
import java.util.Set;
 
import org.apache.commons.httpclient.HttpClient;
import org.apache.commons.httpclient.HttpMethod;
import org.apache.commons.httpclient.HttpStatus;
import org.apache.commons.httpclient.methods.PostMethod;
 
public class Main {
 
	private static Set<String> filenames = new HashSet<String>();
 
	public static String getSource(String url) {
		String response = null;
		HttpClient client = new HttpClient();
		HttpMethod method = null;
		try {
			method = new PostMethod(url);
			client.executeMethod(method);
			if (method.getStatusCode() == HttpStatus.SC_OK) {
				response = method.getResponseBodyAsString();
			}
		} catch (IOException e) {
			System.out.println("Get Source Error!");
		} finally {
			if (method != null)
				method.releaseConnection();
		}
		return response;
	}
 
	public static void create(String url, String froot, String cur) {
		String curUrl = url + "/" + cur;
		if (filenames.contains(curUrl)) {
			return;
		}
		filenames.add(curUrl);
		String cont = getSource(curUrl);
		if (cont == null) {
			return;
		}
		File f = null;
		FileOutputStream fos = null;
		try {
			f = new File(froot);
			if (!f.exists()) {
				f.mkdirs();
			}
			f = new File(froot + "\\" + cur);
			f.createNewFile();
			fos = new FileOutputStream(f);
			fos.write(cont.getBytes());
			fos.flush();
		} catch (FileNotFoundException e) {
			e.printStackTrace();
		} catch (IOException e) {
			e.printStackTrace();
		} finally {
			if (fos != null) {
				try {
					fos.close();
				} catch (IOException e) {
					e.printStackTrace();
				}
			}
		}
		String[] files = cont.split("\"");
		for (String file : files) {
			if (file.matches(".*\\.html$") && !file.startsWith("http")) {
				if (file.contains(" "))
					continue;
				String turl = url;
				String tfroot = froot;
				file.replaceAll("//", "/");
				int tindex = file.lastIndexOf("?");
				if (tindex != -1) {
					file = file.substring(0, tindex);
				}
				int index = 0;
				while ((index = file.indexOf("/")) != -1) {
					if (file.charAt(index - 1 < 0 ? 0 : index - 1) == '.') {
						turl = turl.substring(0, turl.lastIndexOf("/"));
						tfroot = tfroot.substring(0, tfroot.lastIndexOf("\\"));
					} else {
						turl = turl + "/" + file.substring(0, index);
						tfroot = tfroot + "\\" + file.substring(0, index);
					}
					file = file.substring(index + 1);
				}
				create(turl, tfroot, file);
			}
		}
	}
 
	public static void main(String[] args) {
		String url = "http://localhost";
		String froot = "C:\\Users\\Jack_Wong\\Desktop\\api";
		create(url, froot, "index.html");
	}
}

发表回复

您的电子邮箱地址不会被公开。 必填项已用 * 标注