没有借助任何第三方库，完全基于JAVA Socket实现一个最小化的HTTP文件下载客户端。完整的演示如何通过Socket实现下载文件的HTTP请求(request header)发送如何从Socket中接受HTTP响应(Response header, Response body)报文并解析与保存文件内容。如何通过SwingWork实现UI刷新，实时显示下载进度。

首先看一下UI部分：

【添加下载】按钮：

点击弹出URL输入框，用户Copy要下载文件URL到输入框以后，点击[OK]按钮即开始

下载

【清除完成】按钮：

清除所有已经下载完成的文件列表

文件下载状态分为以下几种：

package com.gloomyfish.socket.tutorial.http.download;

public enum DownLoadStatus {

NOT_STARTED,

IN_PROCESS,

COMPLETED,

ERROR

}

UI部分主要是利用Swing组件完成。点击【添加下载】执行的代码如下：

final JDialog dialog = new JDialog(this,"Add File Link",true);

dialog.getContentPane().setLayout(new BorderLayout());

// dialog.setSize(new Dimension(400,200));

final URLFilePanel panel = new URLFilePanel();

panel.setUpListener(new ActionListener(){

@Override

public void actionPerformed(ActionEvent e) {

if("OK".equals(e.getActionCommand())){

if(panel.validateInput()) {

DownloadDetailStatusInfoModel data = new DownloadDetailStatusInfoModel(panel.getValidFileURL());

tableModel.getData().add(data);

startDownlaod();

refreshUI();

}

dialog.setVisible(false);

dialog.dispose();

} else if("Cancel".equals(e.getActionCommand())) {

dialog.setVisible(false);

dialog.dispose();

}

}});

dialog.getContentPane().add(panel, BorderLayout.CENTER);

dialog.pack();

centre(dialog);

dialog.setVisible(true);

【清除完成】按钮执行的代码如下：

private void clearDownloaded() {

List downloadedList = new ArrayList();

for(DownloadDetailStatusInfoModel fileStatus : tableModel.getData()) {

if(fileStatus.getStatus().toString().equals(DownLoadStatus.COMPLETED.toString())) {

downloadedList.add(fileStatus);

}

}

tableModel.getData().removeAll(downloadedList);

refreshUI();

}

让JFrame组件居中显示的代码如下：

public static void centre(Window w) {

Dimension us = w.getSize();

Dimension them = Toolkit.getDefaultToolkit().getScreenSize();

int newX = (them.width - us.width) / 2;

int newY = (them.height - us.height) / 2;

w.setLocation(newX, newY);

}

HTTP协议实现部分：

概述：HTTP请求头与相应头报文基本结构与解释

HTTP请求：一个标准的HTTP请求报文如

其中请求头可以有多个，message-body可以没有，不是必须的。请求行的格式如下：

Request-Line = Method SP Request-URI SPHTTP-Version CRLF 举例说明如下：

Request-Line = GET http://www.w3.org/pub/WWW/TheProject.htmlHTTP/1.1\r\n

其中SP表示空格， CRLF表示回车换行符\r\n

当你想要上传文件时候，使用Post方式来填写数据到message-body中即可。发送一个

简单的HTTP请求报文如下：

GET /pub/WWW/TheProject.html HTTP/1.1\r\n

\r\n

HTTP响应：一个标准的HTTP响应报文如下

最先得到是状态行，其格式如下：

Status-Line = HTTP-Version SP Status-CodeSP Reason-Phrase CRLF, 一个状态行的简单例子如下：Status-Line = HTTP/1.1 200 OK一般大家最喜欢的就是Status-Code会给你很多提示，最常见的就是404，500等状态码。状态码的意思可以参考RFC2616中的解释。下载文件最要紧是的检查HTTP响应头中的Content-Length与Content-Type两

个中分别声明了文件的长度与文件的类型。其它如Accept-Ranges表示接受多少到多少的字节。可能在多线程下载中使用。搞清楚了HTTP请求与响应的报文格式以后，我们就可以通过Socket按照报文格式解析内容，发送与读取HTTP请求与响应。具体步骤

如下：

一、根据用户输入的文件URL建立Socket连接

URL url = new URL(fileInfo.getFileURL());

String host = url.getHost();

int port = (url.getPort() == -1) ? url.getDefaultPort():url.getPort();

System.out.println("Host Name = " + host);

System.out.println("port = " + port);

System.out.println("File URI = " + url.getFile());

// create socket and start to construct the request line

Socket socket = new Socket();

SocketAddress address = new InetSocketAddress(host, port);

socket.connect(address);

用了URL类来把用户输入的url string变成容易解析一点的URL。

二、构造HTTP请求

BufferedWriter bufferedWriter = new BufferedWriter(new OutputStreamWriter(socket.getOutputStream(), "UTF8"));

String requestStr = "GET " + url.getFile() + " HTTP/1.1\r\n"; // request line

// construct the request header - 构造HTTP请求头(request header)

String hostHeader = "Host: " + host + "\r\n";

String acceptHeader = "Accept: text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8\r\n";

String charsetHeader = "Accept-Charset: GBK,utf-8;q=0.7,*;q=0.3\r\n";

String languageHeader = "Accept-Language: zh-CN,zh;q=0.8\r\n";

String keepHeader = "Connection: close\r\n";

三、发送HTTP请求

// 发送HTTP请求

bufferedWriter.write(requestStr);

bufferedWriter.write(hostHeader);

bufferedWriter.write(acceptHeader);

bufferedWriter.write(charsetHeader);

bufferedWriter.write(languageHeader);

bufferedWriter.write(keepHeader);

bufferedWriter.write("\r\n"); // 请求头信息发送结束标志

bufferedWriter.flush();

四、接受HTTP响应并解析内容，写入创建好的文件

// 准备接受HTTP响应头并解析

CustomDataInputStream input = new CustomDataInputStream(socket.getInputStream());

File myFile = new File(fileInfo.getStoreLocation() + File.separator + fileInfo.getFileName());

String content = null;

HttpResponseHeaderParser responseHeader = new HttpResponseHeaderParser();

BufferedOutputStream output = new BufferedOutputStream(new FileOutputStream(myFile));

boolean hasData = false;

while((content = input.readHttpResponseHeaderLine()) != null) {

System.out.println("response header contect -->> " + content);

responseHeader.addResponseHeaderLine(content);

if(content.length() == 0) {

hasData = true;

}

if(hasData) {

int totalBytes = responseHeader.getFileLength();

if(totalBytes == 0) break; // no response body and data

int offset = 0;

byte[] myData = null;

if(totalBytes >= 2048) {

myData = new byte[2048];

} else {

myData = new byte[totalBytes];

}

int numOfBytes = 0;

while((numOfBytes = input.read(myData, 0, myData.length)) > 0 && offset < totalBytes) {

offset += numOfBytes;

float p = ((float)offset) / ((float)totalBytes) * 100.0f;

if(offset > totalBytes) {

numOfBytes = numOfBytes + totalBytes - offset;

p = 100.0f;

}

output.write(myData, 0, numOfBytes);

updateStatus(p);

}

hasData = false;

break;

}

}

简单的HTTP响应头解析类HttpResponseHeaderParser代码如下：

package com.gloomyfish.socket.tutorial.http.download;

import java.util.HashMap;

import java.util.Map;

/**

* it can parse entity header, response head

* and response line

* refer to RFC2616，关于HTTP响应头，请看RFC文档，描写的很详细啊！！

*/

public class HttpResponseHeaderParser {

public final static String CONTENT_LENGTH = "Content-Length";

public final static String CONTENT_TYPE = "Content-Type";

public final static String ACCEPT_RANGES = "Accetp-Ranges";

private Map headerMap;

public HttpResponseHeaderParser() {

headerMap = new HashMap();

}

/**

*

get the response header key value pair

* @param responseHeaderLine

*/

public void addResponseHeaderLine(String responseHeaderLine) {

if(responseHeaderLine.contains(":")) {

String[] keyValue = responseHeaderLine.split(": ");

if(keyValue[0].equalsIgnoreCase(CONTENT_LENGTH)) {

headerMap.put(CONTENT_LENGTH, keyValue[1]);

} else if(keyValue[0].equalsIgnoreCase(CONTENT_TYPE)) {

headerMap.put(CONTENT_TYPE, keyValue[1]);

} else {

headerMap.put(keyValue[0], keyValue[1]);

}

}

}

public int getFileLength() {

if(headerMap.get(CONTENT_LENGTH) == null){

return 0;

}

return Integer.parseInt(headerMap.get(CONTENT_LENGTH));

}

public String getFileType() {

return headerMap.get(CONTENT_TYPE);

}

public Map getAllHeaders() {

return headerMap;

}

}