最近研究了jsp中作HTTP认证的问题,它的工作方式如下:
1、server发送一个要求认证代码401和一个头信息WWW-authenticate,激发browser弹出一个认证窗口
(本文来源于图老师网站,更多请访问http://m.tulaoshi.com/java/)2、server取得browser送来的认证头"Authorization",它是加密的了,要用Base64方法解密,取得明文的用户名和密码
3、检查用户名和密码,根据结果传送不同的页面
(本文来源于图老师网站,更多请访问http://m.tulaoshi.com/java/)
以下是jsp的片断,你也可以把它做成include文件。和Base64的加解密的class源码。
如有兴趣可与我联系:unixboy@yeah.net
<jsp:useBean id="base64"scope="page"class="Base64"/
<%
if(request.getHeader("Authorization")==null){
response.setStatus(401);
response.setHeader("WWW-authenticate","Basic realm="unixboy.com"");
}else{
String encoded=(request.getHeader("Authorization"));
String tmp=encoded.substring(6);
String up=Base64.decode(tmp);
String user="";
String password="";
if(up!=null){
user=up.substring(0,up.indexOf(":"));
password=up.substring(up.indexOf(":")+1);
}
if(user.equals("unixboy")&&password.equals("123456")){
//认证成功
}else{
//认证失败
}
}
%
//消息加解密class
public class Base64
{
/** decode a Base 64 encoded String.
*<p<h4String to byte conversion</h4
* This method uses a naive String to byte interpretation, it simply gets each
* char of the String and calls it a byte.</p
*<pSince we should be dealing with Base64 encoded Strings that is a reasonable
* assumption.</p
*<p<h4End of data</h4
* We don''t try to stop the converion when we find the"="end of data padding char.
* We simply add zero bytes to the unencode buffer.</p
*/
public static String decode(String encoded)
{
StringBuffer sb=new StringBuffer();
int maxturns;
//work out how long to loop for.
if(encoded.length()%3==0)
maxturns=encoded.length();
else
maxturns=encoded.length()+(3-(encoded.length()%3));
//tells us whether to include the char in the unencode
boolean skip;
//the unencode buffer
byte[] unenc=new byte[4];
&n