Java 和 JavaScript Base64编码

在开发Java  Web应用的时候，可能会在服务器端用Java做Base64编码，而在客户端用JavaScript进行解码。这样就要求两边的Base64编码机制保持一致。

使用Base64编码，可能会碰到各种奇怪情况，甚至怀疑编码有bug。但实际上不是这样的。Base64理论上操作的对象不是字符串而是字节数组。它的原理就是把ASCII码的255个字符缩小到用64个来表示。具体就是原来三个字节用四个字节表示，编码后长度有一定的增长。

1) 最好一次编码，避免分段编码，确实要分段编码，每一段字节数应该是3的倍数。

长字节流，如果要边读取边编码，每一段必须是3的倍数，否则就可能在还原的时候出乱。一般人喜欢用2的乘方来定义数组，例如 byte[1024]，因为不是3的倍数，可能还原时出错。正确的例子是：

byte[] bs=new byte[3*100] ….inputStream.read(bs)……encode(bs )….

对于字符串，一般要整个一次编码，以避免分段编码出错。

当然，如果你分段编码，还原的时候也是一段一段地还原，那是没有问题的。

2)确保字符串还原的时候按照原来的编码还原。

因为它操作的是字节数组，所以对于GBK编码的汉字和UTF-8编码汉字，经过 Base64编码后结果是不一样的。例如“我们”这两个字如果是GBK编码，转成Base64后就是ztLDxw== ；如果是UTF-8编码，转成Base64后就是5oiR5Lus。

也就是 “我们” ==》  getBytes(“GBK”) ==> Base64

所以Java这边用什么编码转换，在JavaScript那边就要用什么编码还原。要保证Java和JavaScript通用，我们采用Unicode的编码（JavaScript转成UTF-8、GBK不方便，所以就采用了其本身的Unicode编码），具体如下：

服务器端：

1）用getBytes(“Unicode”)转成Unicode字节数组。

2） 编码成Base64字符串

3）传送到客户端

客户端：

1）Base64 解码成字节数组

2）按Unicode还原

代码如下（相关的函数看附件）：

Base64.encode(data,”Unicode”); //java 端编码

decode64(data);   //javascript解码

附一：Java中Base64编码

package websharp.util; 
public class Base64 { 
  private static final byte[] encodingTable = { 
      (byte) 'A',(byte) 'B',(byte) 'C',(byte) 'D',(byte) 'E',(byte) 'F',(byte) 'G',(byte) 'H',(byte) 'I',(byte) 'J',(byte) 'K',(byte) 'L',(byte) 'M',(byte) 'N',(byte) 'O',(byte) 'P',(byte) 'Q',(byte) 'R',(byte) 'S',(byte) 'T',(byte) 'U',(byte) 'V',(byte) 'W',(byte) 'X',(byte) 'Y',(byte) 'Z',(byte) 'a',(byte) 'b',(byte) 'c',(byte) 'd',(byte) 'e',(byte) 'f',(byte) 'g',(byte) 'h',(byte) 'i',(byte) 'j',(byte) 'k',(byte) 'l',(byte) 'm',(byte) 'n',(byte) 'o',(byte) 'p',(byte) 'q',(byte) 'r',(byte) 's',(byte) 't',(byte) 'u',(byte) 'v',(byte) 'w',(byte) 'x',(byte) 'y',(byte) 'z',(byte) '0',(byte) '1',(byte) '2',(byte) '3',(byte) '4',(byte) '5',(byte) '6',(byte) '7',(byte) '8',(byte) '9',(byte) '+',(byte) '/' 
    }; 
  private static final byte[] decodingTable; 
  static { 
    decodingTable = new byte[128]; 
    for (int i = 0; i < 128; i++) { 
      decodingTable[i] = (byte) -1; 
    } 
    for (int i = 'A'; i <= 'Z'; i++) { 
      decodingTable[i] = (byte) (i - 'A'); 
    } 
    for (int i = 'a'; i <= 'z'; i++) { 
      decodingTable[i] = (byte) (i - 'a' + 26); 
    } 
    for (int i = '0'; i <= '9'; i++) { 
      decodingTable[i] = (byte) (i - '0' + 52); 
    } 
    decodingTable['+'] = 62; 
    decodingTable['/'] = 63; 
  } 
  public static byte[] encode(byte[] data,int offset) { 
    byte[] bytes; 
    int realCount=data.length-offset; 
    int modulus = realCount % 3; 
    if (modulus == 0) { 
      bytes = new byte[(4 * realCount) / 3]; 
    } else { 
      bytes = new byte[4 * ((realCount / 3) + 1)]; 
    } 
    int dataLength = (data.length - modulus); 
    int a1; 
    int a2; 
    int a3; 
    for (int i = offset,j = 0; i < dataLength; i += 3,j += 4) { 
      a1 = data[i] & 0xff; 
      a2 = data[i + 1] & 0xff; 
      a3 = data[i + 2] & 0xff; 
      bytes[j] = encodingTable[(a1 >>> 2) & 0x3f]; 
      bytes[j + 1] = encodingTable[((a1 << 4) | (a2 >>> 4)) & 0x3f]; 
      bytes[j + 2] = encodingTable[((a2 << 2) | (a3 >>> 6)) & 0x3f]; 
      bytes[j + 3] = encodingTable[a3 & 0x3f]; 
    } 
    int b1; 
    int b2; 
    int b3; 
    int d1; 
    int d2; 
    switch (modulus) { 
    case 0: /* nothing left to do */ 
      break; 
    case 1: 
      d1 = data[data.length - 1] & 0xff; 
      b1 = (d1 >>> 2) & 0x3f; 
      b2 = (d1 << 4) & 0x3f; 
      bytes[bytes.length - 4] = encodingTable[b1]; 
      bytes[bytes.length - 3] = encodingTable[b2]; 
      bytes[bytes.length - 2] = (byte) '='; 
      bytes[bytes.length - 1] = (byte) '='; 
      break; 
    case 2: 
      d1 = data[data.length - 2] & 0xff; 
      d2 = data[data.length - 1] & 0xff; 
      b1 = (d1 >>> 2) & 0x3f; 
      b2 = ((d1 << 4) | (d2 >>> 4)) & 0x3f; 
      b3 = (d2 << 2) & 0x3f; 
      bytes[bytes.length - 4] = encodingTable[b1]; 
      bytes[bytes.length - 3] = encodingTable[b2]; 
      bytes[bytes.length - 2] = encodingTable[b3]; 
      bytes[bytes.length - 1] = (byte) '='; 
      break; 
    } 
    return bytes; 
  } 
  public static byte[] decode(byte[] data) { 
    byte[] bytes; 
    byte b1; 
    byte b2; 
    byte b3; 
    byte b4; 
    data = discardNonBase64Bytes(data); 
    if (data[data.length - 2] == '=') { 
      bytes = new byte[(((data.length / 4) - 1) * 3) + 1]; 
    } else if (data[data.length - 1] == '=') { 
      bytes = new byte[(((data.length / 4) - 1) * 3) + 2]; 
    } else { 
      bytes = new byte[((data.length / 4) * 3)]; 
    } 
    for (int i = 0,j = 0; i < (data.length - 4); i += 4,j += 3) { 
      b1 = decodingTable[data[i]]; 
      b2 = decodingTable[data[i + 1]]; 
      b3 = decodingTable[data[i + 2]]; 
      b4 = decodingTable[data[i + 3]]; 
      bytes[j] = (byte) ((b1 << 2) | (b2 >> 4)); 
      bytes[j + 1] = (byte) ((b2 << 4) | (b3 >> 2)); 
      bytes[j + 2] = (byte) ((b3 << 6) | b4); 
    } 
    if (data[data.length - 2] == '=') { 
      b1 = decodingTable[data[data.length - 4]]; 
      b2 = decodingTable[data[data.length - 3]]; 
      bytes[bytes.length - 1] = (byte) ((b1 << 2) | (b2 >> 4)); 
    } else if (data[data.length - 1] == '=') { 
      b1 = decodingTable[data[data.length - 4]]; 
      b2 = decodingTable[data[data.length - 3]]; 
      b3 = decodingTable[data[data.length - 2]]; 
      bytes[bytes.length - 2] = (byte) ((b1 << 2) | (b2 >> 4)); 
      bytes[bytes.length - 1] = (byte) ((b2 << 4) | (b3 >> 2)); 
    } else { 
      b1 = decodingTable[data[data.length - 4]]; 
      b2 = decodingTable[data[data.length - 3]]; 
      b3 = decodingTable[data[data.length - 2]]; 
      b4 = decodingTable[data[data.length - 1]]; 
      bytes[bytes.length - 3] = (byte) ((b1 << 2) | (b2 >> 4)); 
      bytes[bytes.length - 2] = (byte) ((b2 << 4) | (b3 >> 2)); 
      bytes[bytes.length - 1] = (byte) ((b3 << 6) | b4); 
    } 
    return bytes; 
  } 
  public static byte[] decode(String data) { 
    byte[] bytes; 
    byte b1; 
    byte b2; 
    byte b3; 
    byte b4; 
    data = discardNonBase64Chars(data); 
    if (data.charAt(data.length() - 2) == '=') { 
      bytes = new byte[(((data.length() / 4) - 1) * 3) + 1]; 
    } else if (data.charAt(data.length() - 1) == '=') { 
      bytes = new byte[(((data.length() / 4) - 1) * 3) + 2]; 
    } else { 
      bytes = new byte[((data.length() / 4) * 3)]; 
    } 
    for (int i = 0,j = 0; i < (data.length() - 4); i += 4,j += 3) { 
      b1 = decodingTable[data.charAt(i)]; 
      b2 = decodingTable[data.charAt(i + 1)]; 
      b3 = decodingTable[data.charAt(i + 2)]; 
      b4 = decodingTable[data.charAt(i + 3)]; 
      bytes[j] = (byte) ((b1 << 2) | (b2 >> 4)); 
      bytes[j + 1] = (byte) ((b2 << 4) | (b3 >> 2)); 
      bytes[j + 2] = (byte) ((b3 << 6) | b4); 
    } 
    if (data.charAt(data.length() - 2) == '=') { 
      b1 = decodingTable[data.charAt(data.length() - 4)]; 
      b2 = decodingTable[data.charAt(data.length() - 3)]; 
      bytes[bytes.length - 1] = (byte) ((b1 << 2) | (b2 >> 4)); 
    } else if (data.charAt(data.length() - 1) == '=') { 
      b1 = decodingTable[data.charAt(data.length() - 4)]; 
      b2 = decodingTable[data.charAt(data.length() - 3)]; 
      b3 = decodingTable[data.charAt(data.length() - 2)]; 
      bytes[bytes.length - 2] = (byte) ((b1 << 2) | (b2 >> 4)); 
      bytes[bytes.length - 1] = (byte) ((b2 << 4) | (b3 >> 2)); 
    } else { 
      b1 = decodingTable[data.charAt(data.length() - 4)]; 
      b2 = decodingTable[data.charAt(data.length() - 3)]; 
      b3 = decodingTable[data.charAt(data.length() - 2)]; 
      b4 = decodingTable[data.charAt(data.length() - 1)]; 
      bytes[bytes.length - 3] = (byte) ((b1 << 2) | (b2 >> 4)); 
      bytes[bytes.length - 2] = (byte) ((b2 << 4) | (b3 >> 2)); 
      bytes[bytes.length - 1] = (byte) ((b3 << 6) | b4); 
    } 
    for(int i=0;i<bytes.length;i++) System.out.println(","+bytes[i]); 
    return bytes; 
  } 
  private static byte[] discardNonBase64Bytes(byte[] data) { 
    byte[] temp = new byte[data.length]; 
    int bytesCopied = 0; 
    for (int i = 0; i < data.length; i++) { 
      if (isValidBase64Byte(data[i])) { 
        temp[bytesCopied++] = data[i]; 
      } 
    } 
    byte[] newData = new byte[bytesCopied]; 
    System.arraycopy(temp,newData,bytesCopied); 
    return newData; 
  } 
  private static String discardNonBase64Chars(String data) { 
    StringBuffer sb = new StringBuffer(); 
    int length = data.length(); 
    for (int i = 0; i < length; i++) { 
      if (isValidBase64Byte((byte) (data.charAt(i)))) { 
        sb.append(data.charAt(i)); 
      } 
    } 
    return sb.toString(); 
  } 
  private static boolean isValidBase64Byte(byte b) { 
    if (b == '=') { 
      return true; 
    } else if ((b < 0) || (b >= 128)) { 
      return false; 
    } else if (decodingTable[b] == -1) { 
      return false; 
    } 
    return true; 
  } 
  public static String encode(String data,String charset)throws Exception 
  { 
    // byte[] result = (data.getBytes("Unicode")); 
     if(data==null || data.length()==0) return data; 
     int offset=0; 
     // getBytes("unicode")转完后会在前头加上两字节”FE“ 
     byte[] result=encode (data.getBytes(charset),offset); 
     StringBuffer sb=new StringBuffer(result.length); 
     for (int i=0;i<result.length;i++)  sb.append((char)result[i]); 
     return sb.toString(); 
  } 
  public static String decode(String data,String charset)throws Exception 
  { 
    if(data==null || data.length()==0) return data; 
    return new String(Base64.decode(data),charset); 
  } 
  public static void main(String[] args) throws Exception { 
    String data = "我们"; 
    String data1=encode(data,"Unicode"); 
    String data2=decode(data1,"Unicode"); 
    System.out.println(data); 
    System.out.println(data1); 
    System.out.println(data2); 
  } 
}