1. #include <iostream>
2. #include <string>
3. #include <codecvt>
4. #include <stdexcept>
5. #include <sstream>
6. #include <iomanip>
7. #include <vector>
8.  
9. std::vector<std::string> extractUTF16HexValues(const std::string& input) {
10.     std::vector<std::string> utf16HexValues;
11.     std::stringstream ss(input);
12.     std::string token;
13.  
14.     // Iterate over tokens separated by "\\u"
15.     while (std::getline(ss, token, '\\')) {
16.         if (token.empty())
17.             continue;
18.  
19.         // Check if the token starts with "u" and has exactly 4 characters following it
20.         if (token[0] == 'u' && token.length() == 5) {
21.             // Extract the hexadecimal part
22.             std::string hexValue = token.substr(1, 4);
23.             utf16HexValues.push_back(hexValue);
24.         }
25.     }
26.  
27.     return utf16HexValues;
28. }
29.  
30. std::string convertUTF16HexToUTF8Hex(const std::vector<std::string>& utf16HexValues) {
31.     try {
32.         std::wstring_convert<std::codecvt_utf8_utf16<char16_t>, char16_t> converter;
33.         std::stringstream utf8HexStream;
34.         std::u16string utf16String;
35.  
36.         // Combine all UTF-16 hexadecimal values into a single u16string
37.         for (const std::string& utf16Hex : utf16HexValues) {
38.             uint16_t utf16Value = std::stoi(utf16Hex, nullptr, 16);
39.             utf16String.push_back(utf16Value);
40.         }
41.  
42.         // Convert the UTF-16 u16string to UTF-8
43.         std::string utf8Encoded = converter.to_bytes(utf16String);
44.  
45.         // Convert each byte to hexadecimal representation and append to the result
46.         for (unsigned char c : utf8Encoded) {
47.             utf8HexStream << std::hex << std::setw(2) << std::setfill('0') << static_cast<int>(c);
48.         }
49.  
50.         return utf8HexStream.str();
51.     } catch (const std::invalid_argument& e) {
52.         std::cerr << "Error: " << e.what() << std::endl;
53.         return "";
54.     } catch (const std::out_of_range& e) {
55.         std::cerr << "Error: Invalid UTF-16 hexadecimal value." << std::endl;
56.         return "";
57.     }
58. }
59.  
60. std::string convertUTF8HexToUTF8Bytes(const std::string& utf8Hex) {
61.     std::stringstream utf8ByteStream;
62.  
63.     // Iterate over pairs of characters representing hexadecimal values
64.     for (size_t i = 0; i < utf8Hex.length(); i += 2) {
65.         // Get the next two characters
66.         std::string byteHex = utf8Hex.substr(i, 2);
67.  
68.         // Convert hexadecimal string to integer
69.         int byteValue;
70.         std::stringstream(byteHex) >> std::hex >> byteValue;
71.  
72.         // Append the byte as a character to the string
73.         utf8ByteStream << static_cast<char>(byteValue);
74.     }
75.  
76.     return utf8ByteStream.str();
77. }
78.  
79. int main() {
80.     std::string input = "\\uD83E\\uDDA0";
81.     std::vector<std::string> utf16HexValues = extractUTF16HexValues(input);
82.    
83.     //std::vector<std::string> utf16HexValues = {"D83E", "DDA0"};
84.     std::string utf8Hex = convertUTF16HexToUTF8Hex(utf16HexValues);
85.     if (!utf8Hex.empty()) {
86.         std::cout << "UTF-8 Hex: " << utf8Hex << std::endl; // Output: F09FA6A0
87.  
88.         std::string utf8Bytes = convertUTF8HexToUTF8Bytes(utf8Hex);
89.         std::cout << "UTF-8 Bytes: " << utf8Bytes << std::endl; // Output: ?
90.     }
91.     return 0;
92. }
93.