C/C++ base64 编解码

一. Base64编码由来

　　为什么会有Base64编码呢？因为有些网络传送渠道并不支持所有的字节，例如传统的邮件只支持可见字符的传送，像ASCII码的控制字符就 不能通过邮件传送。这样用途就受到了很大的限制，比如图片二进制流的每个字节不可能全部是可见字符，所以就传送不了。最好的方法就是在不改变传统协议的情 况下，做一种扩展方案来支持二进制文件的传送。把不可打印的字符也能用可打印字符来表示，问题就解决了。Base64编码应运而生，Base64就是一种 基于64个可打印字符来表示二进制数据的表示方法。

二. Base64编码原理

　　看一下Base64的索引表，字符选用了"A-Z、a-z、0-9、+、/" 64个可打印字符。数值代表字符的索引，这个是标准Base64协议规定的，不能更改。64个字符用6个bit位就可以全部表示，一个字节有8个bit 位，剩下两个bit就浪费掉了，这样就不得不牺牲一部分空间了。这里需要弄明白的就是一个Base64字符是8个bit，但是有效部分只有右边的6个 bit，左边两个永远是0。

三. 索引表

      const char BASE_CODE[]  = {'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','+','/'};

四.编码

    char* Base64::encode(const char *src){
    size_t sLength = strlen(src);
    size_t dLength = ((sLength%3 == 0 ?0:1)+sLength/3)*4;
    char* dest = (char*)malloc(sizeof(char)*(dLength+1));
    int count = 0;
    for (int i=0; i<=sLength-3; i=i+3) {
        unsigned char sc1 = src[i];
        unsigned char sc2 = src[i+1];
        unsigned char sc3 = src[i+2];
        char c1 = sc1>>2;
        char c2 = ((unsigned char)(sc1<<6))>>2 | (sc2>>4);
        char c3 = ((unsigned char)(sc2<<4))>>2| (sc3>>6);
        unsigned char ct4 = sc3<<2;
        char c4 = ct4>>2;
        printf("%c,%c,%c,%d,%d,%d\n",sc1,sc2,sc3,c1,c2,c3);
        dest[count*4] = BASE_CODE[c1];
        dest[count*4+1] = BASE_CODE[c2];
        dest[count*4+2] = BASE_CODE[c3];
        dest[count*4+3] = BASE_CODE[c4];
        count++;
    }
    size_t m = sLength%3;
    if(m == 1){
        unsigned char sc1 = src[sLength-1];
        char c1 = sc1>>2;
        char c2 = ((unsigned char)(sc1<<6))>>2;
        dest[count*4] = BASE_CODE[c1];
        dest[count*4+1] = BASE_CODE[c2];
        dest[count*4+2] = '=';
        dest[count*4+3] = '=';
    }else if(m == 2){
        unsigned char sc1 = src[sLength-2];
        unsigned char sc2 = src[sLength-1];
        char c1 = sc1>>2;
        char c2 = ((unsigned char)(sc1<<6))>>2 | (sc2>>4);
        char c3 = ((unsigned char)(sc2<<4))>>2;
        dest[count*4] = BASE_CODE[c1];
        dest[count*4+1] = BASE_CODE[c2];
        dest[count*4+2] = BASE_CODE[c3];
        dest[count*4+3] = '=';
    }
    dest[dLength] = '\0';
    return dest;
}

五.解码

   char* Base64::decode(const char *src){
    size_t slength = strlen(src);
    size_t dlength = (slength/4)*3;
    char* dest = (char*)malloc(sizeof(char)*(dlength+1));
    int count = 0;
    for (int i=0; i<slength; i = i+4) {
        char sc1 = getDecodeValue(src[i]);
        char sc2 = getDecodeValue(src[i+1]);
        char sc3 = getDecodeValue(src[i+2]);
        char sc4 = getDecodeValue(src[i+3]);
        unsigned char c1 = sc1<<2 | sc2>>4;
        unsigned char c2 = sc2<<4 | sc3>>2;
        unsigned char c3 = sc3<<6 | sc4;
        dest[count*3] = c1;
        dest[count*3+1] = c2;
        dest[count*3+2] = c3;
        count++;
    }
    dest[dlength] = '\0';
    return dest;
}

char Base64::getDecodeValue(const char src){
    char dest = 0;
    for(int i=0;i<64;i++){
        if (BASE_CODE[i] == src) {
            dest = i;
            break;
        }
    }
    return dest;
}

参考：http://www.cnblogs.com/chengxiaohui/articles/3951129.html