Request feature: Exemple Usage + possible bug: strings above 55 bytes not handle properly about rlpvalue HOT 7 OPEN

Did you solve that problem ? I have same problem in the latest version including your fix.

from rlpvalue.

Did you solve that problem ? I have same problem in the latest version including your fix.

What input did you try to serialize ? What result do you expect from it ?

from rlpvalue.

Same input like you. below:

RLPValue lorem("Lorem ipsum dolor sit amet, consectetur adipisicing elit");
std::string serialized_lorem = lorem.write();

for (size_t i = 0; i < serialized_lorem.size(); i++) {
    printf("\\x%.2x ", serialized_lorem.c_str()[i]);
}

But it prints out \xffffffb8 \x38 \x4c ... but should print \xb8 \x38 \x4c ...

from rlpvalue.

My code is here.

/* g++ -o rlp rlp.cpp -I. -L/usr/local/lib -lrlpvalue -static */

#include <stdio.h>
#include "rlpvalue.h"

using namespace std;

int main(void)
{
    RLPValue lorem("Lorem ipsum dolor sit amet, consectetur adipisicing elit");
    std::string serialized_lorem = lorem.write();

    for (size_t i = 0; i < serialized_lorem.size(); i++) {
        printf("\\x%.2x ", serialized_lorem.c_str()[i]);
    }
    printf("\n");

    return 0;
}

from rlpvalue.

I think the data is placed in "serialized_lorem" correctly.
Is this simply a matter of output?

from rlpvalue.

Recommend using ParseHex and HexStr for going to/from hexidecimal. Example:
https://github.com/bloq/rlpvalue/blob/master/test/unitester.cpp#L129

			std::string genHex = HexStr(genOutput.begin(),
						    genOutput.end());
			fprintf(stderr, "INS :%s\nGENS:%s\n",
				outs.c_str(),
				genHex.c_str());

from rlpvalue.

Pretty late to the party, this got me also. It isn't a bug in the serialized_lorem. Instead its an issue with how you print to hex.

In the code the length is prepended to the data using this:

static std::string encodeLength(size_t n, unsigned char offset)
{
	std::string rs;

	if (n < 56) {
		unsigned char ch = n + offset;
		rs.assign((const char *) &ch, 1);
	}

	else {
		// assert(n too big);
		std::string binlen = encodeBinary(n);

		unsigned char ch = binlen.size() + offset + 55;
		rs.assign((const char *) &ch, 1);
		rs.append(binlen);
	}

	return rs;
}

which represents the length as a single char, if the offset (0xC0 for example) pushes this above what can be represented as a single signed byte then it will be represented as a negative in std::string.

When printing if this gets interpreted as a signed char and converted to an unsigned char, then your output will be incorrect. Hence why HexStr is recommended going to hexidecimal

from rlpvalue.