C ++中的URL编码和Node.js中的解码

Question

I am transferring data from a C++ client to a nodejs server.

I compress the string using zlib deflate first, then I use curl_easy_escape to url encode the compressed string.

std::string s = zlib_compress(temp.str());
std::cout << s <<"\n";
CURL *handle = curl_easy_init();
char* o = curl_easy_escape(handle, s.data(), s.size());
std::cout << o <<"\n";

Then I send it using:

std::string bin(o);
curl_easy_setopt(handle, CURLOPT_POSTFIELDSIZE, bin.size());
curl_easy_setopt(handle, CURLOPT_POSTFIELDS, bin.data());
curl_easy_perform(handle);

When I run this, I get the output:

x??с??Ҵ4?
x%DA%D3%D1%81%80%E2%92%D2%B44%1D%03%00%1BW%03%E5

Now, I receive the second encoded string on my nodejs server as it is. I now try to decode it.

var x = req.params;
for (var key in req.body)
{   
console.log(key);
var x = unescape(key);
var buffer = new Buffer(x);
console.log(x);
zlib.inflate(buffer, function(err, buffer) {

    console.log(err+" here");
});
}

Which outputs:

x%DA%D3%D1%81%80%E2%92%D2%B44%1D%03%00%1BW%03%E5
xÚÓÑâÒ´4å
Error: incorrect header check here

What is the problem here? How do I debug it?

Answer 1

You can debug it by printing the decimal value for each byte in the compressed string in C++ and node.js code. For C++ that code would be:

for(int i=0; i<s.size(); i++) {
    std::cout << static_cast<int>(s[i]);
}

In node.js code you would need to print the decimal value for each byte contained in variable buffer .

If the decimal values for each byte are identical in both C++ and node.js parts, then zlib libraries are incompatible or functions do not match: eg zlib_compress in C++ may correspond to something else than zlib.inflate in node.js: maybe there is a function like zlib.decompress() .

The root cause can be in that characters are 1-byte in C++ std::string and 2-byte in node.js . Specifying the encoding when constructing Buffer in node.js may solve the problem if that is it:

var buffer = new Buffer(x, 'binary');

See https://nodejs.org/api/buffer.html#buffer_new_buffer_str_encoding

As the data is zlib compressed here, or in a general compressed case, the encoding should be binary .