dbc_statement_handler.cpp

// SPDX-License-Identifier: MIT
#include "dbc_statement_handler.h"
#include "j1939.h"
#include "text_utils.h"
#include "signal_layout.h"
#include <charconv>
#include <cmath>
#include <limits>
#include <regex>
#include <sstream>
#include <stdexcept>
 
namespace {
    std::uint32_t ParseDecimal(const std::string &text,
                               std::uint32_t maximum,
                               const char *field) {
        std::uint32_t value = 0U;
        const std::from_chars_result parsed = std::from_chars(text.data(), text.data() + text.size(), value);
        if (text.empty() || parsed.ec != std::errc() ||
            parsed.ptr != text.data() + text.size() || value > maximum)
            throw std::runtime_error(std::string("Invalid ") + field + ": " + text);
        return value;
    }
 
    std::uint32_t ParseCanId(const std::string &text) {
        const std::uint32_t rawId = ParseDecimal(
                text, std::numeric_limits<std::uint32_t>::max(), "CAN ID");
        // Only bit 31 may accompany the actual 29-bit identifier.
        if ((rawId & 0x60000000U) != 0U)
            throw std::runtime_error("CAN ID exceeds the 29-bit identifier range.");
        return rawId;
    }
 
    double ParsePhysicalValue(const std::string &text, const char *field) {
        static const std::regex number(
                R"(^[+-]?(?:[0-9]+(?:\.[0-9]*)?|\.[0-9]+)(?:[eE][+-]?[0-9]+)?$)");
        if (!std::regex_match(text, number))
            throw std::runtime_error(std::string("Invalid ") + field + ": " + text);
        // Floating-point from_chars is locale-independent and does not accept
        // a leading '+', which the DBC numeric grammar permits.
        const char *begin = text.data();
        const char *end = begin + text.size();
        if (*begin == '+')
            ++begin;
        double value = 0.0;
        const std::from_chars_result parsed =
                std::from_chars(begin, end, value, std::chars_format::general);
        // Keep the existing rejection of conversion underflow into subnormal values.
        if (parsed.ec != std::errc() || parsed.ptr != end || !std::isfinite(value) ||
            std::fpclassify(value) == FP_SUBNORMAL)
            throw std::runtime_error(std::string("Invalid ") + field + ": " + text);
        return value;
    }
 
    Multiplexing ParseMultiplexingMarker(const std::string &marker) {
        Multiplexing result;
        if (marker == "M")
            result.role = MultiplexRole::Selector;
        else if (!marker.empty()) {
            if (marker[0] != 'm')
                throw std::runtime_error("Unsupported multiplexing marker.");
            const std::string value = marker.substr(1U);
            const std::from_chars_result parsed =
                std::from_chars(value.data(), value.data() + value.size(), result.value);
            if (value.empty() || parsed.ec != std::errc() || parsed.ptr != value.data() + value.size())
                throw std::runtime_error("Invalid or unsupported multiplexing marker.");
            result.role = MultiplexRole::Branch;
        }
 
        return result;
    }
 
    RawSignalValue ParseValueKey(const std::string &number) {
        RawSignalValue key;
        if (!number.empty() && number[0] == '-') {
            std::int64_t value = 0;
            const std::from_chars_result parsed =
                std::from_chars(number.data(), number.data() + number.size(), value);
            if (parsed.ec != std::errc() || parsed.ptr != number.data() + number.size())
                throw std::runtime_error("Invalid VAL_ integer: " + number);
            key = NormalizeValueKey(RawSignalValue(value));
        } else {
            std::uint64_t value = 0;
            const std::from_chars_result parsed =
                std::from_chars(number.data(), number.data() + number.size(), value);
            if (number.empty() || parsed.ec != std::errc() || parsed.ptr != number.data() + number.size())
                throw std::runtime_error("Invalid VAL_ integer: " + number);
            key = value;
        }
        return key;
    }
 
    std::string UnescapeDescription(const std::string &quoted) {
        std::string description;
        for (std::size_t i = 0; i < quoted.size(); ++i) {
            if (quoted[i] == '\\')
                ++i;
            description += quoted[i];
        }
        return description;
    }
 
    FrameInfo *FindUniqueFrame(DbcDatabase &database, std::uint32_t canId, bool isExtended) {
        FrameInfo *frame = nullptr;
        for (FrameInfo &candidate : database.frames) {
            if (candidate.canId == canId && candidate.isExtended == isExtended) {
                if (frame)
                    throw std::runtime_error("Ambiguous VAL_ frame reference.");
                frame = &candidate;
            }
        }
        if (!frame)
            throw std::runtime_error("Unresolved VAL_ frame reference.");
        return frame;
    }
 
    SignalInfo *FindUniqueSignal(FrameInfo &frame, const std::string &signalName) {
        SignalInfo *signal = nullptr;
        for (SignalInfo &candidate : frame.signals) {
            if (candidate.name == signalName) {
                if (signal)
                    throw std::runtime_error("Ambiguous VAL_ signal reference: " + signalName);
                signal = &candidate;
            }
        }
        if (!signal)
            throw std::runtime_error("Unresolved VAL_ signal reference: " + signalName);
        return signal;
    }
 
    std::uint64_t NextSignalBit(std::uint64_t bit, bool isLittleEndian) {
        if (isLittleEndian)
            return bit + 1U;
        // From bit 0, Motorola continues at bit 7 of the next byte (0 -> 15).
        return bit % 8U == 0U ? bit + 15U : bit - 1U;
    }
 
    FrameInfo *FindFrameById(DbcDatabase &, std::uint32_t, bool);
    SignalInfo *FindSignalByName(FrameInfo &, const std::string &);
} // namespace
 
namespace {
    std::vector<std::string> SplitReceivers(const std::string &text) {
        std::vector<std::string> receivers;
        std::string token;
        std::istringstream stream(text);
 
        while (std::getline(stream, token, ',')) {
            token = dbc_detail::TrimText(token);
            if (!token.empty())
                receivers.push_back(token);
        }
 
        return receivers;
    }
 
    void NormalizeCanId(std::uint32_t rawCanId, std::uint32_t &normalizedCanId, bool &isExtended) {
        /*
         * DBC commonly stores extended identifiers with bit 31 set.
         * Example:
         * raw id = 0x80000000 | actual_29_bit_id
         */
        isExtended = (rawCanId & 0x80000000U) != 0U || rawCanId > 0x7FFU;
        normalizedCanId = rawCanId & 0x1FFFFFFFU;
    }
} // namespace
 
FrameInfo DbcFrameHandler::ParseFrameLine(const std::string &line) {
    /*
   * Example:
   * BO_ 256 EngineData: 8 EEC1
   */
 
    std::istringstream stream(line);
    std::string keyword;
    std::string idText;
    if (!(stream >> keyword >> idText) || keyword != "BO_")
        throw std::runtime_error("Missing CAN ID in frame definition.");
 
    const std::uint32_t rawCanId = ParseCanId(idText);
 
    std::string nameText;
    if (!std::getline(stream, nameText, ':') || stream.eof())
        throw std::runtime_error("Missing ':' in frame definition.");
    std::istringstream nameStream(nameText);
    FrameInfo frame;
    std::string extra;
    if (!(nameStream >> frame.name) || (nameStream >> extra))
        throw std::runtime_error("Missing or invalid frame name.");
 
    std::string lengthText;
    if (!(stream >> lengthText))
        throw std::runtime_error("Missing payload length in frame definition.");
    frame.dlc = static_cast<std::uint8_t>(ParseDecimal(
            lengthText, std::numeric_limits<std::uint8_t>::max(), "payload length"));
    if (!(stream >> frame.transmitter))
        throw std::runtime_error("Missing transmitter in frame definition.");
    if (stream >> extra)
        throw std::runtime_error("Unexpected trailing field in frame definition: " + extra);
 
    NormalizeCanId(rawCanId, frame.canId, frame.isExtended);
 
    return frame;
}
 
SignalInfo DbcSignalHandler::ParseSignalLine(const std::string &line) {
    // Match the complete statement before converting any numeric fields.
    enum Field {
        NameAndMarker = 1, StartBit, BitLength, ByteOrder, Signedness,
        Factor, Offset, Minimum, Maximum, Unit, Receivers
    };
    static const std::regex syntax(
            // Name and optional multiplexing marker.
            R"dbc(^SG_[ \t]+([^:]+):[ \t]*)dbc"
            // Bit position, length, byte order and signedness.
            R"dbc(([^| \t]+)[ \t]*\|[ \t]*([^@ \t]+)[ \t]*@[ \t]*([01])([+-])[ \t]*)dbc"
            // Scaling factor and offset.
            R"dbc(\‍([ \t]*([^, \t]+)[ \t]*,[ \t]*([^ )\t]+)[ \t]*\)[ \t]*)dbc"
            // Physical range.
            R"dbc(\[[ \t]*([^| \t]+)[ \t]*\|[ \t]*([^\] \t]+)[ \t]*\][ \t]*)dbc"
            // Unit and receivers.
            R"dbc("([^"\r\n]*)"[ \t]+([^ \t,]+(?:[ \t]*,[ \t]*[^ \t,]+)*)[ \t]*$)dbc");
    std::smatch fields;
    if (!std::regex_match(line, fields, syntax))
        throw std::runtime_error("Malformed or unsupported SG_ signal definition.");
 
    SignalInfo signal;
    std::istringstream nameFields(fields[NameAndMarker].str());
    std::string marker;
    std::string extra;
    nameFields >> signal.name >> marker;
    if (signal.name.empty() || (nameFields >> extra))
        throw std::runtime_error("Invalid signal name or multiplexing marker.");
    signal.multiplexing = ParseMultiplexingMarker(marker);
 
    signal.startBit = ParseDecimal(fields[StartBit].str(),
                                   std::numeric_limits<std::uint32_t>::max(), "signal start bit");
    signal.length = ParseDecimal(fields[BitLength].str(), 64U, "signal length");
    if (signal.length == 0U)
        throw std::runtime_error("Signal length must be within 1..64 bits.");
    signal.isLittleEndian = fields[ByteOrder].str() == "1";
    signal.isSigned = fields[Signedness].str() == "-";
 
    signal.factor = ParsePhysicalValue(fields[Factor].str(), "signal factor");
    signal.offset = ParsePhysicalValue(fields[Offset].str(), "signal offset");
    signal.minimum = ParsePhysicalValue(fields[Minimum].str(), "signal minimum");
    signal.maximum = ParsePhysicalValue(fields[Maximum].str(), "signal maximum");
    if (signal.minimum > signal.maximum)
        throw std::runtime_error("Signal minimum exceeds maximum.");
    signal.unit = fields[Unit].str();
    signal.receivers = SplitReceivers(fields[Receivers].str());
    return signal;
}
 
 
namespace {
    FrameInfo *FindFrameById(DbcDatabase &database,
                             std::uint32_t canId,
                             bool isExtended) {
        for (FrameInfo &frame : database.frames) {
            if (frame.canId == canId && frame.isExtended == isExtended)
                return &frame;
        }
 
        return nullptr;
    }
 
    SignalInfo *FindSignalByName(FrameInfo &frame, const std::string &signalName) {
        for (SignalInfo &signal : frame.signals) {
            if (signal.name == signalName)
                return &signal;
        }
 
        return nullptr;
    }
}// namespace
 
void DbcFrameHandler::Handle(const std::string &line, DbcParseContext &context) const {
    FrameInfo frame = ParseFrameLine(line);
    if (context.enableJ1939 && frame.isExtended) {
        frame.hasPgn = true;
        frame.pgn = ExtractJ1939Pgn(frame.canId);
    }
    if (FindFrameById(context.database, frame.canId, frame.isExtended) != nullptr)
        throw std::runtime_error("Duplicate " + std::string(frame.isExtended ? "extended" : "standard") +
                                 " frame CAN ID: " + std::to_string(frame.canId));
    context.database.frames.push_back(std::move(frame));
    context.currentFrame = &context.database.frames.back();
    context.signalLines.emplace_back();
}
 
void DbcSignalHandler::Handle(const std::string &line, DbcParseContext &context) const {
    if (context.currentFrame == nullptr)
        throw std::runtime_error("Signal found before any frame definition.");
    SignalInfo signal = ParseSignalLine(line);
    if (FindSignalByName(*context.currentFrame, signal.name) != nullptr)
        throw std::runtime_error("Duplicate signal name '" + signal.name +
                                 "' in frame '" + context.currentFrame->name + "'.");
    if (!IsValidSignalLayout(signal.startBit, signal.length, signal.isLittleEndian,
                            context.currentFrame->dlc))
        throw std::runtime_error("Signal exceeds the frame payload length.");
    context.currentFrame->signals.push_back(std::move(signal));
    context.signalLines.back().push_back(context.lineNumber);
}
 
void DbcCommentHandler::Handle(const std::string &line, DbcParseContext &context) const {
    std::istringstream stream(line);
    std::string keyword;
    std::string target;
    stream >> keyword >> target;
    if (keyword != "CM_" || (target != "BO_" && target != "SG_"))
        return;
 
    PendingComment pending;
    pending.isSignal = target == "SG_";
    pending.lineNumber = context.lineNumber;
    std::string idText;
    if (!(stream >> idText))
        throw std::runtime_error("Missing CAN ID in CM_ statement.");
    NormalizeCanId(ParseCanId(idText), pending.canId, pending.isExtended);
    if (pending.isSignal && (!(stream >> pending.signalName) ||
                             pending.signalName.find_first_of("\";") != std::string::npos))
        throw std::runtime_error("Missing or invalid signal name in CM_ statement.");
 
    stream >> std::ws;
    if (stream.get() != '"')
        throw std::runtime_error("Missing opening quote in CM_ statement.");
    bool closed = false;
    char character;
    while (stream.get(character)) {
        if (character == '"') {
            closed = true;
            break;
        }
        if (character == '\\') {
            if (!stream.get(character) || (character != '"' && character != '\\'))
                throw std::runtime_error("Invalid escape in CM_ statement.");
        }
        pending.comment += character;
    }
    if (!closed)
        throw std::runtime_error("Unterminated CM_ comment.");
    stream >> std::ws;
    if (stream.get() != ';')
        throw std::runtime_error("Missing semicolon in CM_ statement.");
    stream >> std::ws;
    if (!stream.eof())
        throw std::runtime_error("Unexpected trailing text in CM_ statement.");
    context.pendingComments.push_back(std::move(pending));
}
 
void DbcParseContext::ResolveComments(const std::string &filePath) {
    for (const PendingComment &pending: pendingComments) {
        FrameInfo *frame = FindFrameById(database, pending.canId, pending.isExtended);
        if (!frame)
            throw DbcDiagnosticError(DbcDiagnosticCode::InvalidStatement, filePath,
                                     pending.lineNumber, "Unresolved CM_ frame reference.");
        if (pending.isSignal) {
            SignalInfo *signal = FindSignalByName(*frame, pending.signalName);
            if (!signal)
                throw DbcDiagnosticError(DbcDiagnosticCode::InvalidStatement, filePath,
                                         pending.lineNumber, "Unresolved CM_ signal reference: " + pending.signalName);
            signal->comment = pending.comment;
        } else {
            frame->comment = pending.comment;
        }
    }
}
 
DbcStatementDispatcher::DbcStatementDispatcher() {
    RegisterType<DbcHeaderHandler>("VERSION");
    RegisterType<DbcHeaderHandler>("NS_");
    RegisterType<DbcHeaderHandler>("NS_:");
    RegisterType<DbcHeaderHandler>("BS_");
    RegisterType<DbcHeaderHandler>("BS_:");
    RegisterType<DbcHeaderHandler>("BU_");
    RegisterType<DbcHeaderHandler>("BU_:");
    RegisterType<DbcUnsupportedMultiplexingHandler>("SG_MUL_VAL_");
    RegisterType<DbcFrameHandler>("BO_");
    RegisterType<DbcSignalHandler>("SG_");
    RegisterType<DbcCommentHandler>("CM_");
    RegisterType<DbcValueDescriptionHandler>("VAL_");
    RegisterType<DbcAttributeHandler>("BA_DEF_");
    RegisterType<DbcAttributeHandler>("BA_DEF_DEF_");
    RegisterType<DbcAttributeHandler>("BA_");
}
 
void DbcHeaderHandler::Handle(const std::string &, DbcParseContext &) const {}
 
bool DbcStatementDispatcher::Dispatch(const std::string &line,
                                      DbcParseContext &context) const {
    std::istringstream stream(line);
    std::string keyword;
    stream >> keyword;
    const HandlerMap::const_iterator handler = m_handlers.find(keyword);
    if (handler == m_handlers.end())
        return false;
    handler->second->Handle(line, context);
    return true;
}
 
void DbcValueDescriptionHandler::Handle(const std::string &line, DbcParseContext &context) const {
    static const std::regex header(R"(^VAL_[ \t]+([^ \t]+)[ \t]+([^ \t;]+)[ \t]+(.*)$)");
    static const std::regex pair(R"dbc(^([^ \t]+)[ \t]+"((?:[^"\\]|\\["\\])*)")dbc");
    std::smatch fields;
    if (!std::regex_match(line, fields, header))
        throw std::runtime_error("Malformed or unsupported VAL_ statement.");
    const std::uint32_t rawId = ParseDecimal(fields[1].str(), std::numeric_limits<std::uint32_t>::max(), "VAL_ CAN ID");
    if ((rawId & 0x60000000U) != 0U)
        throw std::runtime_error("VAL_ CAN ID exceeds the 29-bit identifier range.");
    PendingValueDescriptions pending;
    NormalizeCanId(rawId, pending.canId, pending.isExtended);
    pending.signalName = fields[2].str();
    pending.lineNumber = context.lineNumber;
    std::string rest = fields[3].str();
    while (rest != ";") {
        if (!std::regex_search(rest, fields, pair))
            throw std::runtime_error("Malformed VAL_ value/description pair or missing semicolon.");
        const std::string number = fields[1].str();
        const RawSignalValue key = ParseValueKey(number);
        const std::string description = UnescapeDescription(fields[2].str());
        if (!pending.descriptions.emplace(key, description).second)
            throw std::runtime_error("Duplicate VAL_ value key: " + number);
        const std::size_t consumed = static_cast<std::size_t>(fields.length());
        rest.erase(0, consumed);
        if (!rest.empty() && rest[0] != ';' && rest[0] != ' ' && rest[0] != '\t')
            throw std::runtime_error("Missing whitespace between VAL_ pairs.");
        rest = dbc_detail::TrimText(rest);
    }
    if (pending.descriptions.empty())
        throw std::runtime_error("VAL_ requires at least one value/description pair.");
    context.pendingValues.push_back(std::move(pending));
}
 
void DbcParseContext::ResolveValueDescriptions(const std::string &filePath) {
    for (const PendingValueDescriptions &pending: pendingValues) {
        try {
            FrameInfo *frame = FindUniqueFrame(database, pending.canId, pending.isExtended);
            SignalInfo *signal = FindUniqueSignal(*frame, pending.signalName);
            for (const ValueDescriptions::value_type &entry: pending.descriptions) {
                if (!signal->valueDescriptions.insert(entry).second)
                    throw std::runtime_error("Duplicate VAL_ value key for signal: " + pending.signalName);
            }
        } catch (const std::exception &error) {
            throw DbcDiagnosticError(DbcDiagnosticCode::InvalidValueDescription,
                                     filePath, pending.lineNumber, error.what());
        }
    }
}
 
void DbcUnsupportedMultiplexingHandler::Handle(const std::string &, DbcParseContext &) const {
    throw std::runtime_error("Extended and nested multiplexing are unsupported.");
}
 
namespace {
    std::vector<std::uint64_t> SignalBits(const SignalInfo &signal) {
        std::vector<std::uint64_t> bits;
        std::uint64_t bit = signal.startBit;
        for (std::uint32_t index = 0U; index < signal.length; ++index) {
            bits.push_back(bit);
            bit = NextSignalBit(bit, signal.isLittleEndian);
        }
        return bits;
    }
 
    bool AreMutuallyExclusive(const SignalInfo &first, const SignalInfo &second) {
        return first.multiplexing.role == MultiplexRole::Branch &&
               second.multiplexing.role == MultiplexRole::Branch &&
               first.multiplexing.value != second.multiplexing.value;
    }
 
    std::string ValidateBranch(const SignalInfo &signal, const SignalInfo *selector) {
        if (signal.multiplexing.role != MultiplexRole::Branch)
            return {};
        if (selector == nullptr)
            return "Multiplexed branch has no selector.";
        if (selector->length < 64U && signal.multiplexing.value >= (std::uint64_t(1U) << selector->length))
            return "Branch value exceeds selector raw range.";
        return {};
    }
 
    const SignalInfo *FindSelector(const FrameInfo &frame,
                                   const std::vector<std::size_t> &lines,
                                   const std::string &filePath) {
        const SignalInfo *selector = nullptr;
        for (std::size_t index = 0U; index < frame.signals.size(); ++index) {
            const SignalInfo &signal = frame.signals[index];
            if (signal.multiplexing.role != MultiplexRole::Selector)
                continue;
            if (selector != nullptr || signal.isSigned)
                throw DbcDiagnosticError(DbcDiagnosticCode::InvalidMultiplexing,
                                         filePath, lines[index],
                                         "Multiplexing requires one unsigned selector per frame.");
            selector = &signal;
        }
        return selector;
    }
 
    bool SignalsOverlap(const SignalInfo &first, const SignalInfo &second) {
        const std::vector<std::uint64_t> firstBits = SignalBits(first);
        const std::vector<std::uint64_t> secondBits = SignalBits(second);
        for (const std::uint64_t firstBit : firstBits) {
            for (const std::uint64_t secondBit : secondBits) {
                if (firstBit == secondBit)
                    return true;
            }
        }
        return false;
    }
}
 
void DbcParseContext::ValidateMultiplexing(const std::string &filePath) const {
    for (std::size_t frameIndex = 0U; frameIndex < database.frames.size(); ++frameIndex) {
        const FrameInfo &frame = database.frames[frameIndex];
        const SignalInfo *selector = FindSelector(frame, signalLines[frameIndex], filePath);
        for (std::size_t index = 0U; index < frame.signals.size(); ++index) {
            const SignalInfo &signal = frame.signals[index];
            std::string error = ValidateBranch(signal, selector);
            for (std::size_t previous = 0U; previous < index && error.empty(); ++previous) {
                const SignalInfo &other = frame.signals[previous];
                if (!AreMutuallyExclusive(signal, other) && SignalsOverlap(signal, other))
                    error = "Signals overlap without mutually exclusive multiplexing conditions.";
            }
            if (!error.empty())
                throw DbcDiagnosticError(DbcDiagnosticCode::InvalidMultiplexing,
                                         filePath, signalLines[frameIndex][index], error);
        }
    }
}