fork download
copy 
  1. #include <iostream>
  2. #include <string>
  3. #include <vector> 
  4.  
  5. std::u32string utf8_to_utf32(const std::string& str) {
  6.     std::u32string result;
  7.     size_t i = 0;
  8.  
  9.     while (i < str.size()) {
  10.         uint32_t codepoint = 0;
  11.         unsigned char c = str[i];
  12.  
  13.         if ((c & 0x80) == 0) {  // 1-byte (ASCII)
  14.             codepoint = c;
  15.             i += 1;
  16.         } else if ((c & 0xE0) == 0xC0) {  // 2-byte
  17.             codepoint = ((c & 0x1F) << 6) | (str[i + 1] & 0x3F);
  18.             i += 2;
  19.         } else if ((c & 0xF0) == 0xE0) {  // 3-byte
  20.             codepoint = ((c & 0x0F) << 12) | ((str[i + 1] & 0x3F) << 6) | (str[i + 2] & 0x3F);
  21.             i += 3;
  22.         } else if ((c & 0xF8) == 0xF0) {  // 4-byte
  23.             codepoint = ((c & 0x07) << 18) | ((str[i + 1] & 0x3F) << 12) |
  24.                         ((str[i + 2] & 0x3F) << 6) | (str[i + 3] & 0x3F);
  25.             i += 4;
  26.         } else {
  27.             throw std::runtime_error("Invalid UTF-8 sequence");
  28.         }
  29.  
  30.         result.push_back(codepoint);
  31.     }
  32.  
  33.     return result;
  34. }
  35.  
  36. std::wstring utf32_to_utf16(const std::u32string& str) {
  37.     std::wstring result;
  38.     for (char32_t codepoint : str) {
  39.         if (codepoint <= 0xFFFF) {  // BMP (Basic Multilingual Plane)
  40.             result.push_back(static_cast<wchar_t>(codepoint));
  41.         } else {  // Surrogate pair needed
  42.             codepoint -= 0x10000;
  43.             wchar_t high_surrogate = static_cast<wchar_t>((codepoint >> 10) + 0xD800);
  44.             wchar_t low_surrogate = static_cast<wchar_t>((codepoint & 0x3FF) + 0xDC00);
  45.             result.push_back(high_surrogate);
  46.             result.push_back(low_surrogate);
  47.         }
  48.     }
  49.     return result;
  50. }
  51.  
  52. int main() {
  53.     std::string utf8_text = "A – B";
  54.     std::u32string unicode_text = utf8_to_utf32(utf8_text);  // UTF-8 → UTF-32
  55.     std::wstring wide_text = utf32_to_utf16(unicode_text);   // UTF-32 → UTF-16
  56.  
  57.     std::wcout << L"UTF-16 wstring: " << wide_text << std::endl;
  58.     return 0;
  59. }
Success #stdin #stdout 0.01s 5292KB
comments (?)
stdin
copy 
Standard input is empty
stdout
copy 
UTF-16 wstring: A - B
https://ideone.com/YHxyrT
language:
C++ (gcc 8.3)
created:
1 day ago
visibility:
public

Share or Embed source code

Discover > Sphere Engine API

The brand new service which powers Ideone!

Discover > IDE Widget

Widget for compiling and running the source code in a web browser!