Convert a char array of escaped UTF-8 octets to a string in C++

Question

I have a char array containing some UTF-8-encoded Turkish characters - in the form of escaped octets. Thus if I run this code in C++11:

void foo(char* utf8_encoded) { 

    cout << utf8_encoded << endl;

}

it prints \\xc4\\xb0-\\xc3\\x87-\\xc3\\x9c-\\xc4\\x9e . I want to convert this char[] to an std::string so that it contains UTF-8 decoded values İ-Ç-Ü-Ğ . I have converted that char[] to wstring but it still prints as \\xc4\\xb0-\\xc3\\x87-\\xc3\\x9c-\\xc4\\x9e . How can I do that?

EDIT: I'm not the one who constructs this char[]. It is one of the static-length parameter of a callback function called by a private library. So the callback function is as follows:

void some_callback_function (INFO *info) { 

    cout << info->some_char_array << endl;
    cout << "*****" << endl;

    for(int i=0; i<64; i++) {
        cout << "-" << info->some_char_array[i];
    }
    cout << "*****" << endl;

    char bar[65] = "\xc4\xb0-\xc3\x87-\xc3\x9c-\xc4\x9e";
    cout << bar << endl;
}

Where the INFO struct is:

typedef struct {
    char some_char_array[65];
} INFO;

So when my callback function is called, the output is as follows:

\xc4\xb0-\xc3\x87-\xc3\x9c-\xc4\x9e
*****
-\-x-c-4-\-x-b-0---\-x-c-3-\-x-8-7---\-x-c-3-\-x-9-c---\-x-c-4-\-x-9-e-----------------------------
*****
İ-Ç-Ü-Ğ

So my current question is, I didn't get the difference between info->some_char_array and bar char arrays. What I want is to edit info->some_char_array such that, it prints the output as İ-Ç-Ü-Ğ .

Answer 1

OK, this is a bit of a handful, ripped out of a larger parser I am using. But "a bit of a handful" is the nature of Boost.Spirit . ;-)

The parser will not only parse hexadecimal escapes, but octals ( \\123 ) and "standard" escapes ( \\n ) as well. Provided under CC0, so you can do with it whatever you like. ;-)

Boost.Spirit is a "header only" part of Boost, so you don't need to link in any library code. The rather involved "magic" done by the Spirit headers to allow grammars expressed in C++ source this way is a bit hard on the compile time, though.

But it works, and works well.

#define BOOST_SPIRIT_USE_PHOENIX_V3

#include "boost/spirit/include/qi.hpp"
#include "boost/spirit/include/phoenix.hpp"

#include <string>
#include <cstring>
#include <sstream>
#include <stdexcept>

namespace
{

// Helper function: Turn on_error positional parameters into error message.
template< typename Iterator >
std::string make_error_message( boost::spirit::info const & info, Iterator first, Iterator last )
{
    std::ostringstream oss;
    oss << "Invalid sequence. Expecting " << info << " here: \"" << std::string( first, last ) << "\"";
    return oss.str();
}

}

// Wrap helper function with Boost.Phoenix boilerplate, so the function
// can be called from within a parser's [].
BOOST_PHOENIX_ADAPT_FUNCTION( std::string, make_error_message_, make_error_message, 3 )

// Supports various escape sequences:
// - Character escapes ( \a \b \f \n \r \t \v \" \\ )
// - Octal escapes ( \n \nn \nnn )
// - Hexadecimal escapes ( \xnn ) (*)
//
// (*): In C/C++, a hexadecimal escape runs until the first non-hexdigit
//      is encountered, which is not very helpful. This one takes exactly
//      two hexdigits.

// Declaring a grammer that works given any kind of iterator,
// and results in a std::string object.
template < typename Iterator >
class EscapedString : public boost::spirit::qi::grammar< Iterator, std::string() >
{
    public:
        // Constructor
        EscapedString() : EscapedString::base_type( escaped_string )
        {
            // An escaped string is a sequence of
            // characters that are not '\', or
            // an escape sequence
            escaped_string = *( +( boost::spirit::ascii::char_ - '\\' ) | escapes );

            // An escape sequence begins with '\', followed by
            // an escaped character (e.g. "\n"), or
            // an 'x' and 2..2 hexadecimal digits, or
            // 1..3 octal digits.
            escapes = '\\' > ( escaped_character
                               | ( "x" > boost::spirit::qi::uint_parser< char, 16, 2, 2 >() )
                               | boost::spirit::qi::uint_parser< char, 8, 1, 3 >() );

            // The list of special "escape" characters
            escaped_character.add
            ( "a", 0x07 )  // alert
            ( "b", 0x08 )  // backspace
            ( "f", 0x0c )  // form feed
            ( "n", 0x0a )  // new line
            ( "r", 0x0d )  // carriage return
            ( "t", 0x09 )  // horizontal tab
            ( "v", 0x0b )  // vertical tab
            ( "\"", 0x22 ) // literal quotation mark
            ( "\\", 0x5c ) // literal backslash
            ;

            // Error handling
            boost::spirit::qi::on_error< boost::spirit::qi::fail >
            (
                escapes,
                // backslash not followed by a valid sequence
                boost::phoenix::throw_(
                    boost::phoenix::construct< std::runtime_error >( make_error_message_( boost::spirit::_4, boost::spirit::_3, boost::spirit::_2 ) )
                )
            );
        }

    private:
        // Qi Rule member
        boost::spirit::qi::rule< Iterator, std::string() > escaped_string;

        // Helpers
        boost::spirit::qi::rule< Iterator, std::string() > escapes;
        boost::spirit::qi::symbols< char const, char > escaped_character;
};


int main()
{
    // Need to escape the backslashes, or "\xc4" would give *one*
    // byte of output (0xc4, decimal 196). I understood the input
    // to be the FOUR character hex char literal,
    // backslash, x, c, 4 in this case,
    // which is what this string literal does.
    char * some_char_array = "\\xc4\\xb0-\\xc3\\x87-\\xc3\\x9c-\\xc4\\x9e";

    std::cout << "Input: '" << some_char_array << "'\n";

    // result object    
    std::string s;

    // Create an instance of the grammar with "char *"
    // as the iterator type.
    EscapedString< char * > es;

    // start, end, parsing grammar, result object
    boost::spirit::qi::parse( some_char_array,
                              some_char_array + std::strlen( some_char_array ),
                              es,
                              s );

    std::cout << "Output: '" << s << "'\n";

    return 0;
}

This gives:

Input: '\xc4\xb0-\xc3\x87-\xc3\x9c-\xc4\x9e'
Output: 'İ-Ç-Ü-Ğ'