c++ multiple regex extractions to an array from dbc entry

Hello I would like to extract the parameters from the following string:
VAL_ 234 State1 123 "Description 1" 0 "Description 2 with \n new line" 90903489 "Big value and special characters &$§())!" ;

The desired matches are

• 234
• State1
• and then an array with unsigned integer and string combination
  • 123 "Description 1"
  • 0 "Description 2 with \n new line"
  • 90903489 "Big value and special characters &$§())!"

The array shall be split in a second step if it is not possible to do it directly.
With the following regex I just get always the last match of the array 90903489 "Big value and special characters &$§())!"

^VAL_ ([0-9]+) ([A-Za-z_][A-Za-z_0-9]*) ([0-9]*\\s\"[^\"]*\"\\s)+

Is there a possibility to extract the values?
I found already

auto blah = std::string{"5001 | 5002 | 5003"};
auto values = std::vector<std::string>{
    std::sregex_token_iterator{blah.begin(), blah.end(), std::regex{R"(\d+)"}},
    std::sregex_token_iterator{}};

from this post but it returns me just the complete string. Is there a possibility to iterate over the submatches?

Not sure if you have any specific requirements on how the matches need to be separated, but you can match either of the patterns with the following regular expression:

(?:^VAL_\s(\d+)\s(\w+)|\s(\d+\s".+?"))

Sample code:

const std::string input{ R"(VAL_ 234 State1 123 "Description 1" 0 "Description 2 with \n new line" 90903489 "Big value and special
characters &$§())!")" };
const std::regex regex{ R"((?:^VAL_\s(\d+)\s(\w+)|\s(\d+\s".+?")))" };
const std::sregex_iterator end{};
for(auto it = std::sregex_iterator{ std::cbegin(input), std::cend(input), regex };
    it != end; ++it) {
    auto match = *it;
    if (match.empty()) {
        std::cerr << "Nothing matched" << '\n';
        continue;
    } else {
        if (match[1].matched) {
            std::cout << "Val match: " << match[1].str() << '\n';
        }
        if (match[2].matched) {
            std::cout << "State match: " << match[2].str() << '\n';
        }
        if (match[3].matched) {
            std::cout << "Etc match: " << match[3].str() << '\n';
        }
    }
}

Based on @rustyx link I created my own parser

enum VALToken {
    Identifier = 0,
    CANId,
    SignalName,
    Value,
    Description
};

struct ValueDescription{
    std::string value;
    std::string description;
};


int main(int argc, char *argv[])
{
    const std::string s = R"(VAL_ 234 State1 123 "Description 1" 0 "Description 2 with \n new line" 90903489 "Big value and special characters &$§())!" ;)";

    auto state = Identifier;
    const char* a = s.data();
    std::string can_id;
    std::string signal_name;
    std::vector<ValueDescription> vds;
    ValueDescription vd;
    for (;;) {
        switch (state) {
        case Identifier: {
            if (*a != 'V')
                return 0;
            a++;
            if (*a != 'A')
                return 0;
            a++;
            if (*a != 'L')
                return 0;
            a++;
            if (*a != '_')
                return 0;
            a++;
            if (*a != ' ')
                return 0;
            a++; // skip whitespace
            state = CANId;
            break;
        }
        case CANId: {
            while(*a >= '0' && *a <= '9') {
                can_id += *a;
                a++;
            }
            if (can_id.empty())
                return 0;
            if (*a != ' ')
                return 0;
            a++; // skip whitespace
            state = SignalName;
            break;
        }
        case SignalName: {
            if ((*a >= 'a' && *a <= 'z') || (*a >= 'A' && *a <= 'Z') || *a == '_')
                signal_name += *a;
            else
                return 0;
            a++;
            while ((*a >= 'a' && *a <= 'z') || (*a >= 'A' && *a <= 'Z') || *a == '_' || (*a >= '0' && *a <= '9')) {
                signal_name += *a;
                a++;
            }
            if (*a != ' ')
                return 0;
            a++; // skip whitespace
            state = Value;
            break;
        }
        case Value: {
            std::string value_str;
            while (*a >= '0' && *a <= '9') {
                value_str += *a;
                a++;
            }
            if (value_str.empty())
                return 0;

            if (*a != ' ')
                return 0;
            a++; // skip whitespace
            vd.value = value_str;
            state = Description;
            break;
        }
        case Description: {
            std::string desc;
            if (*a != '"')
                return 0;
            a++;
            while (*a != '"' && *a != 0) {
                desc += *a;
                a++;
            }
            if (*a == 0)
                return 0;
            a++;
            if (*a != ' ')
                return 0;
            a++; // skip whitespace

            vd.description = desc;
            vds.push_back(vd);

            state = Value;
            break;
        }
        }
    }

    return 0;
}

I would do a regex_match followed by a loop using an sregex_iterator.

[Demo]

#include <fmt/core.h>
#include <regex>
#include <string>

int main() {
    const std::string text{ "VAL_ 234 State1"
        " 123 \"Description 1\""
        " 0 \"Description 2 with \\n new line\""
        " 90903489 \"Big value and special characters &$§())!\""
    };
    const std::regex pattern{ R"(VAL_ (\d+) \w+(\d+)(.*))" };
    std::smatch matches{};
    if (std::regex_match(text, matches, pattern)) {
        fmt::print("{}\n{}\n", matches[1].str(), matches[2].str());

        std::regex array_pattern{ R"(\s+(\d+)\s+\"([^"]+)\")" };
        auto array_text{ matches[3].str() };
        for (std::sregex_iterator it{ array_text.begin(), array_text.end(), array_pattern };
            it != std::sregex_iterator{};
            ++it) {

            std::smatch array_matches{ *it };
            fmt::print("\t'{}', '{}'\n", array_matches[1].str(), array_matches[2].str());
        }
    }
}

// Outputs:
//
// 234
// 1
//    '123', 'Description 1'
//    '0', 'Description 2 with \n new line'
//    '90903489', 'Big value and special characters &$§())!'