Convert string in '\\x00\\x00\\x00' format to unsigned char array - c++

Question

Say I have a string as so:
std::string sc = "\\xfc\\xe8\\x82";
how could I convert the sc string into the equivalent of
unsigned char buf[] = "\xfc\xe8\x82";
I'm trying to convert a string containing shellcode into a unsigned char array.
I have tried the following:
char buf[5120];
strncpy(buf, sc.c_str(), sizeof(buf));
buf[sizeof(buf) - 1] = 0;
This seems to store strings into the char array I need the char array to store/represent bytes.
When I print:
//example 1
unsigned char buf[] = "\xfc\xe8\x82";
printf("%s", buf);
The console outputs:
ⁿΦé
When I print:
//example 2
char buf[5120];
strncpy(buf, sc.c_str(), sizeof(buf));
buf[sizeof(buf) - 1] = 0;
The Console outputs:
\xfc\xe8\x82
How do I convert the sc string into a unsigned char array so that when sc is printed sc will produce the same output of example 1.

Answer 1

The literal "\\xfc\\xe8\\x82" as a string uses "\" as an escape character. "\\" will be reduced to "\". As you would expect. So, if you print your given std::string, then the result will be:
\xfc\xe8\x82.
So, what you want to do now is: Create a char array containing those hex values, given in the original std::string.
Please note: Your statement char s[] = "\xfc\xe8\x82"; will create a C-Style array of char, with the size 4 and containing:
s[0]=fc, s[1]=e8, s[2]=82, s[3]=0
In the example below I show 2 proposals for conversion.
1. Straight forward conversion
2. Using C++ standard algorithms
#include <string>
#include <iostream>
#include <iomanip>
#include <regex>
#include <vector>
#include <iterator>
#include <algorithm>
// Hex digit String
std::regex hexValue{R"(\\[xX]([0-9a-fA-F][0-9a-fA-F]))"};
int main ()
{
// Source string
std::string s1 = "\\xfc\\xe8\\x82";
std::cout << "s 1: " << s1 << "\n";
// Proposal 1 ------------------------------------------------------
// Target array
unsigned char s2[3];
// Convert bytes from strings
for (int i=0; i<s1.size()/4; ++i ) {
// Do conversion. Isolate substring, the convert
s2[i] = std::strtoul(s1.substr(i*4+2,2).c_str(), nullptr,16);
// Result is now in s2
// Output value as tring and decimal value
std::cout << s1.substr(i*4+2,2) << " -> " << std::hex << static_cast <unsigned short>(s2[i])
<< " -> " << std::dec << static_cast <unsigned short>(s2[i]) << "\n";
}
// Proposal 2 ------------------------------------------------------
// Get the tokens
std::vector<std::string> vstr(std::sregex_token_iterator(s1.begin(),s1.end(),hexValue, 1), {});
// Convert to unsigned int
std::vector<unsigned int> vals{};
std::transform(vstr.begin(), vstr.end(), std::back_inserter(vals),
[](std::string &s){ return static_cast<unsigned>(std::strtoul(s.c_str(), nullptr,16)); } );
// Print output on std::cout
std::copy(vals.begin(), vals.end(), std::ostream_iterator<unsigned>(std::cout,"\n"));
return 0;
}
The second solution will eat any number of hex numbers given in a string