Unfortunately the combination of the two iterator_adaptors
binary_from_base64
and transform_width
is not a complete base64 encoder/decoder. Base64 represents groups of 24 bits (3 bytes) as 4 characters, each of which encodes 6 bits. If the input data is not an integer multiple of such 3 byte groups it has to be padded with one or two zero bytes. To indicate how many padding bytes were added, one or two =
characters are appended to the encoded string.
transform_width
, which is responsible for the 8bit binary to 6bit integer conversion does not apply this padding automatically, it has do be done by the user. A simple example:
#include <boost/archive/iterators/base64_from_binary.hpp>
#include <boost/archive/iterators/binary_from_base64.hpp>
#include <boost/archive/iterators/transform_width.hpp>
#include <boost/archive/iterators/insert_linebreaks.hpp>
#include <boost/archive/iterators/remove_whitespace.hpp>
#include <iostream>
#include <string>
using namespace boost::archive::iterators;
using namespace std;
int main(int argc, char **argv) {
typedef transform_width< binary_from_base64<remove_whitespace<string::const_iterator> >, 8, 6 > it_binary_t;
typedef insert_linebreaks<base64_from_binary<transform_width<string::const_iterator,6,8> >, 72 > it_base64_t;
string s;
getline(cin, s, '
');
cout << "Your string is: '"<<s<<"'"<<endl;
// Encode
unsigned int writePaddChars = (3-s.length()%3)%3;
string base64(it_base64_t(s.begin()),it_base64_t(s.end()));
base64.append(writePaddChars,'=');
cout << "Base64 representation: " << base64 << endl;
// Decode
unsigned int paddChars = count(base64.begin(), base64.end(), '=');
std::replace(base64.begin(),base64.end(),'=','A'); // replace '=' by base64 encoding of ''
string result(it_binary_t(base64.begin()), it_binary_t(base64.end())); // decode
result.erase(result.end()-paddChars,result.end()); // erase padding '' characters
cout << "Decoded: " << result << endl;
return 0;
}
Note that I added the insert_linebreaks
and remove_whitespace
iterators, so that the base64 output is nicely formatted and base64 input with line breaks can be decoded. These are optional though.
Run with different input strings which require different padding:
$ ./base64example
Hello World!
Your string is: 'Hello World!'
Base64 representation: SGVsbG8gV29ybGQh
Decoded: Hello World!
$ ./base64example
Hello World!!
Your string is: 'Hello World!!'
Base64 representation: SGVsbG8gV29ybGQhIQ==
Decoded: Hello World!!
$ ./base64example
Hello World!!!
Your string is: 'Hello World!!!'
Base64 representation: SGVsbG8gV29ybGQhISE=
Decoded: Hello World!!!
You can check the base64 strings with this online-encoder/decoder.
与恶龙缠斗过久,自身亦成为恶龙;凝视深渊过久,深渊将回以凝视…