Convert UTF8 to UTF16

utf8_utf16.cpp

#include <string>
#include <string_view>
#include <iostream>
#include <tuple>

std::tuple<std::u8string_view, char32_t> decode_u8_charpoint(std::u8string_view input) {
    // XXX: silently accept invalid UTF8
    if ((input.at(0) & 0b10000000) == 0b00000000) {
        return { 
            input.substr(1), 
            input.at(0) & 0b01111111 
        };
    } else if ((input.at(0) & 0b11100000) == 0b11000000) {
        return {
            input.substr(2), 
            (
                ((input.at(0) & 0b00011111) << 6) |
                ((input.at(1) & 0b00111111) << 0)
            )
        };
    } else if ((input.at(0) & 0b11110000) == 0b11100000) {
        return {
            input.substr(3), 
            (
                ((input.at(0) & 0b00001111) << 12) |
                ((input.at(1) & 0b00111111) <<  6) |
                ((input.at(2) & 0b00111111) <<  0)
            )
        };
    } else if ((input[0] & 0b11111000) == 0b11110000) {
        return {
            input.substr(4), 
            (
                ((input.at(0) & 0b00001111) << 18) |
                ((input.at(1) & 0b00111111) << 12) |
                ((input.at(2) & 0b00111111) <<  6) |
                ((input.at(3) & 0b00111111) <<  0)
            )
        };
    }
    return { input.substr(0, 0), -1 };
}

std::u16string from_utf8(std::u8string_view input) {
    int len = 0;
    std::u8string_view tinput = input;
    while (!tinput.empty()) {
        auto [ninput, code] = decode_u8_charpoint(input);
        tinput = ninput;
        if (int(code) < 0) {
            throw new std::runtime_error("Wrong UTF8 Encoding");
        }
        if (code >= 0 && code <= 0xD7FF) [[likely]] {
            len += 1;
            continue;
        }
        if (code >= 0xE000 && code <= 0xFFFF) {
            len += 1;
            continue;
        }
        if ((code >= 0xD800 && code <= 0xDBFF) || (code >= 0xDC00 && code <= 0xDFFF)) [[unlikely]] {
            throw new std::runtime_error("Reserved High-Surrogates and Low-Surrogates");
        }
        len += 2;
    }
    std::u16string res;
    res.reserve(len + 1);
    tinput = input;
    while (!tinput.empty()) {
        auto [ninput, code] = decode_u8_charpoint(tinput);
        tinput = ninput;
        if (code >= 0 && code <= 0xD7FF) [[likely]] {
            res += code;
            continue;
        }
        if (code >= 0xE000 && code <= 0xFFFF) {
            res += code;
            continue;
        }
        auto sg1 = (code - 0x10000) / 0x400 + 0xD800;
        auto sg2 = code % 0x400 + 0xDC00;
        res += sg1;
        res += sg2;
    }
    return res;
}


int main() {
    {
        std::u8string_view u8view = u8"𝄞";
        std::u16string_view u16view = u"𝄞";
        std::string_view view = "𝄞";
        std::cout << "u8  " << u8view.size() << " " << u8view.length() << std::endl;
        std::cout << "u16 " << u16view.size() << " " << u16view.length() << std::endl;
        std::cout << "    " << view.size() << " " << view.length() <<  std::endl;
    }
    {
        std::u8string u8view = u8"𝄞";
        std::u16string u16view = u"𝄞";
        std::string view = "𝄞";
        std::u16string u16_v2 = from_utf8(u8"𝄞");
        std::cout << "u8  " << u8view.size() << " " << u8view.length() << std::endl;
        std::cout << "u16 " << u16view.size() << " " << u16view.length() << std::endl;
        std::cout << "    " << view.size() << " " << view.length() <<  std::endl;
        std::cout << "u16_v2 " << u16_v2.size() << " " << u16_v2.length() <<  std::endl;
    }
}