Base64的java实现

来源:互联网 发布:用js写的注册页面 编辑:程序博客网 时间:2024/05/17 22:48

Base64 使用US-ASCII子集的65个字符, 每个字符用6位表示 

因此"m"的Base64值为38, 二进制形式是 100110. 

 

对于文本串,编码过程如下。例如"men": 

 

先转成US-ASCII值. 

 

"m"十进制 109 

"e"十进制 101 

"n"十进制 110 

二进制 : 

m 01101101 

e 01100101 

n 01101110 

 

三个8位连起来是24位 

011011010110010101101110 

 

然后分成4个6位 

011011 010110 010101 101110 

 

现在得到4个值,十进制为 

27 22 21 46 

 

对应的 Base64 字符是 : 

b W V u 

 

编码总是基于3个字符,从而产生4个Base64字符。 

如果只剩1或2个字符,使用特殊字符"="补齐Base64的4字。 

如,编码"me" 

 

01101101 01100101 

0110110101100101 

011011 010110 0101 

111111 (与,补足6位) 

011011 010110 010100 

b W U 

b W U = ("=" 补足4字符) 

 

于是 "bWU=" 就是"me"的Base64值. 

 

再如编码 "m" 

 

01101101 

011011 01 

111111 

011011 010000 

b Q = = 

于是 "bQ==" 就是"m"的Base64值. 

 

值得注意的是,MIME规定一行最多76个字符.

 

import java.io.*; 

 

public class MIMEBase64 { 

 

static String BaseTable[] = { 

"A","B","C","D","E","F","G","H","I","J","K","L","M","N","O","P", 

"Q","R","S","T","U","V","W","X","Y","Z","a","b","c","d","e","f", 

"g","h","i","j","k","l","m","n","o","p","q","r","s","t","u","v", 

"w","x","y","z","0","1","2","3","4","5","6","7","8","9","+","/" 

}; 

 

 

public static void encode(String filename, BufferedWriter out) { 

try { 

File f = new File(filename); 

FileInputStream fin = new FileInputStream(filename); 

 

// 读文件到BYTE数组 

byte bytes[] = new byte[(int)(f.length())]; 

int n = fin.read(bytes); 

 

if (n < 1) return; // 没有内容 

 

byte buf[] = new byte[4]; // base64 字符数组 

 

int n3byt = n / 3; // 3 bytes 组数 

int nrest = n % 3; // 分组后剩余 bytes 

int k = n3byt * 3; // 

int linelength = 0; // 行长 

int i = 0; // 指针 

 

// 3-bytes 分组 ... 

while ( i < k ) { 

buf[0] = (byte)(( bytes[i] & 0xFC) >> 2); 

buf[1] = (byte)(((bytes[i] & 0x03) << 4) | 

((bytes[i+1] & 0xF0) >> 4)); 

buf[2] = (byte)(((bytes[i+1] & 0x0F) << 2) | 

((bytes[i+2] & 0xC0) >> 6)); 

buf[3] = (byte)( bytes[i+2] & 0x3F); 

send(out, BaseTable[buf[0]]); 

send(out, BaseTable[buf[1]]); 

send(out, BaseTable[buf[2]]); 

send(out, BaseTable[buf[3]]); 

/* 

以上代码可以优化,但会难以理解 

buf[0]= (byte)(b[i] >> 2); 

buf[1]= (byte)(((b[i] & 0x03) << 4)|(b[i+1]>> 4)); 

buf[2]= (byte)(((b[i+1] & 0x0F)<< 2)|(b[i+2]>> 6)); 

buf[3]= (byte)(b[i+2] & 0x3F); 

send(out,BaseTable[buf[0]]+BaseTable[buf[1]]+ 

BaseTable[buf[2]]+BaseTable[buf[3]]); 

*/ 

 

if ((linelength += 4) >= 76) { 

send(out, "/r/n"); 

linelength = 0; 

i += 3; 

 

// 处理尾部 ... 

if (nrest==2) { 

// 2 bytes left 

buf[0] = (byte)(( bytes[k] & 0xFC) >> 2); 

buf[1] = (byte)(((bytes[k] & 0x03) << 4) | 

((bytes[k+1] & 0xF0) >> 4)); 

buf[2] = (byte)(( bytes[k+1] & 0x0F) << 2); 

else if (nrest==1) { 

// 1 byte left 

buf[0] = (byte)((bytes[k] & 0xFC) >> 2); 

buf[1] = (byte)((bytes[k] & 0x03) << 4); 

if (nrest > 0) { 

// 发送尾部 

if ((linelength += 4) >= 76) send(out, "/r/n"); 

send(out, BaseTable[buf[0]]); 

send(out, BaseTable[buf[1]]); 

// 

if (nrest==2) { 

send(out, BaseTable[buf[2]]); 

else { 

send(out, "="); 

send(out, "="); 

out.flush(); 

//这里用到的send方法,请大家根据需要,自己写。可以是把结果输出到控制台,或发送邮件。 

catch (Exception e) { 

e.printStackTrace(); 

}

原创粉丝点击