{

// Work out if we've got enough space to encode the input

// Every 6 bits of input becomes a byte of output

if (aOutputLen < (aInputLen*8)/6)

{

// FIXME Should we return an error here, or just the length

return (aInputLen*8)/6;

}

// If we get here we've got enough space to do the encoding

const char* b64_dictionary = "ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+/";

if (aInputLen == 3)

{

aOutput[0] = b64_dictionary[aInput[0] >> 2];

aOutput[1] = b64_dictionary[(aInput[0] & 0x3)<<4|(aInput[1]>>4)];

aOutput[2] = b64_dictionary[(aInput[1]&0x0F)<<2|(aInput[2]>>6)];

aOutput[3] = b64_dictionary[aInput[2]&0x3F];

}

else if (aInputLen == 2)

{

aOutput[0] = b64_dictionary[aInput[0] >> 2];

aOutput[1] = b64_dictionary[(aInput[0] & 0x3)<<4|(aInput[1]>>4)];

aOutput[2] = b64_dictionary[(aInput[1]&0x0F)<<2];

aOutput[3] = '=';

}

else if (aInputLen == 1)

{

aOutput[0] = b64_dictionary[aInput[0] >> 2];

aOutput[1] = b64_dictionary[(aInput[0] & 0x3)<<4];

aOutput[2] = '=';

aOutput[3] = '=';

}

else

{

// Break the input into 3-byte chunks and process each of them

int i;

for (i = 0; i < aInputLen/3; i++)

{

b64_encode(&aInput[i*3], 3, &aOutput[i*4], 4);

}

if (aInputLen % 3 > 0)

{

// It doesn't fit neatly into a 3-byte chunk, so process what's left

b64_encode(&aInput[i*3], aInputLen % 3, &aOutput[i*4], aOutputLen - (i*4));

}

}