| | |
| | | #ifndef PARSER_HPP |
| | | #define PARSER_HPP |
| | | |
| | | #include <algorithm> |
| | | #include <memory> |
| | | #include <stack> |
| | | #include <stdexcept> |
| | | #include <string> |
| | | #include <vector> |
| | | |
| | | #include "Interpreter/ExpressionBuilder.hpp" |
| | | #include "Interpreter/OperationsFactory.hpp" |
| | | #include "BaseException.hpp" |
| | | #include "Lexer/Token.hpp" |
| | | #include "Lexer/TokenType.hpp" |
| | | #include "Parser/ParsedExpression.hpp" |
| | | #include "Symbols/ParameterContainer.hpp" |
| | | #include "Symbols/SymbolContainer.hpp" |
| | | #include "Symbols/Value.hpp" |
| | | |
| | | namespace Parser { |
| | | |
| | | class SyntaxError : public std::runtime_error { |
| | | public: |
| | | SyntaxError(const std::string & message, const int line, const int col) : |
| | | std::runtime_error(message + " at line " + std::to_string(line) + ", column " + std::to_string(col)) {} |
| | | |
| | | SyntaxError(const std::string & message, const Lexer::Tokens::Token & token) : |
| | | SyntaxError( |
| | | message + " (found token: '" + token.value + "' type: " + Lexer::Tokens::TypeToString(token.type) + ")", |
| | | token.line_number, token.column_number) {} |
| | | }; |
| | | |
| | | class Parser { |
| | | public: |
| | | Parser() {} |
| | | |
| | | void parseScript(const std::vector<Lexer::Tokens::Token> & tokens, std::string_view input_string, |
| | | const std::string & filename) { |
| | | tokens_ = tokens; |
| | | input_str_view_ = input_string; |
| | | current_token_index_ = 0; |
| | | current_filename_ = filename; |
| | | class Exception : public BaseException { |
| | | public: |
| | | using BaseException::BaseException; |
| | | |
| | | try { |
| | | while (!isAtEnd() && currentToken().type != Lexer::Tokens::Type::END_OF_FILE) { |
| | | parseStatement(); |
| | | } |
| | | if (!isAtEnd() && currentToken().type != Lexer::Tokens::Type::END_OF_FILE) { |
| | | reportError("Unexpected tokens after program end"); |
| | | } |
| | | } catch (const SyntaxError & e) { |
| | | std::cerr << "Syntax Error: " << e.what() << '\n'; |
| | | } catch (const std::exception & e) { |
| | | std::cerr << "Error during parsing: " << e.what() << '\n'; |
| | | throw; |
| | | Exception(const std::string & msg, const Lexer::Tokens::Token & token) { |
| | | rawMessage_ = msg + ": " + token.dump(); |
| | | context_ = |
| | | " at line: " + std::to_string(token.line_number) + ", column: " + std::to_string(token.column_number); |
| | | formattedMessage_ = formatMessage(); |
| | | } |
| | | } |
| | | |
| | | Exception(const std::string & msg, int line, int col) { |
| | | rawMessage_ = msg; |
| | | context_ = " at line: " + std::to_string(line) + ", column: " + std::to_string(col); |
| | | formattedMessage_ = formatMessage(); |
| | | } |
| | | |
| | | std::string formatMessage() const override { return "[Syntax ERROR] >>" + context_ + " << : " + rawMessage_; } |
| | | }; |
| | | |
| | | void parseScript(const std::vector<Lexer::Tokens::Token> & tokens, std::string_view input_string, |
| | | const std::string & filename); |
| | | |
| | | static const std::unordered_map<std::string, Lexer::Tokens::Type> keywords; |
| | | static const std::unordered_map<Lexer::Tokens::Type, Symbols::Variables::Type> variable_types; |
| | |
| | | (current_token_index_ == tokens_.size() - 1 && tokens_.back().type == Lexer::Tokens::Type::END_OF_FILE); |
| | | } |
| | | |
| | | // --- Hibakezelés --- |
| | | // Hiba jelentése és kivétel dobása |
| | | [[noreturn]] void reportError(const std::string & message) { |
| | | // Használjuk az aktuális token pozícióját, ha még nem értünk a végére |
| | | if (current_token_index_ < tokens_.size()) { |
| | | throw SyntaxError(message, tokens_[current_token_index_]); |
| | | } // Ha már a végén vagyunk, az utolsó ismert pozíciót használjuk |
| | | throw Exception(message, tokens_[current_token_index_]); |
| | | } |
| | | int line = tokens_.empty() ? 0 : tokens_.back().line_number; |
| | | int col = tokens_.empty() ? 0 : tokens_.back().column_number; |
| | | throw SyntaxError(message, line, col); |
| | | throw Exception(message, line, col); |
| | | } |
| | | |
| | | // --- Elemzési Módszerek (Moduláris részek) --- |
| | | |
| | | // parseStatement (változatlan) |
| | | void parseStatement() { |
| | |
| | | reportError("Unexpected token at beginning of statement"); |
| | | } |
| | | |
| | | void parseVariableDefinition() { |
| | | Symbols::Variables::Type var_type = parseType(); |
| | | |
| | | Lexer::Tokens::Token id_token = expect(Lexer::Tokens::Type::VARIABLE_IDENTIFIER); |
| | | std::string var_name = id_token.value; |
| | | |
| | | if (!var_name.empty() && var_name[0] == '$') { |
| | | var_name = var_name.substr(1); |
| | | } |
| | | const auto ns = Symbols::SymbolContainer::instance()->currentScopeName(); |
| | | |
| | | expect(Lexer::Tokens::Type::OPERATOR_ASSIGNMENT, "="); |
| | | /* |
| | | Symbols::Value initial_value = parseValue(var_type); |
| | | |
| | | Interpreter::OperationsFactory::defineSimpleVariable(var_name, initial_value, ns, this->current_filename_, |
| | | id_token.line_number, id_token.column_number); |
| | | */ |
| | | |
| | | auto expr = parseParsedExpression(var_type); |
| | | Interpreter::OperationsFactory::defineVariableWithExpression( |
| | | var_name, var_type, std::move(expr), ns, current_filename_, id_token.line_number, id_token.column_number); |
| | | expect(Lexer::Tokens::Type::PUNCTUATION, ";"); |
| | | } |
| | | |
| | | void parseFunctionDefinition() { |
| | | expect(Lexer::Tokens::Type::KEYWORD_FUNCTION_DECLARATION); |
| | | Lexer::Tokens::Token id_token = expect(Lexer::Tokens::Type::IDENTIFIER); |
| | | std::string func_name = id_token.value; |
| | | Symbols::Variables::Type func_return_type = Symbols::Variables::Type::NULL_TYPE; |
| | | expect(Lexer::Tokens::Type::OPERATOR_ASSIGNMENT, "="); |
| | | expect(Lexer::Tokens::Type::PUNCTUATION, "("); |
| | | |
| | | Symbols::FunctionParameterInfo param_infos; |
| | | |
| | | if (currentToken().type != Lexer::Tokens::Type::PUNCTUATION || currentToken().value != ")") { |
| | | while (true) { |
| | | // Paraméter típusa |
| | | Symbols::Variables::Type param_type = parseType(); // Ez elfogyasztja a type tokent |
| | | |
| | | // Paraméter név ($variable) |
| | | Lexer::Tokens::Token param_id_token = expect(Lexer::Tokens::Type::VARIABLE_IDENTIFIER); |
| | | std::string param_name = param_id_token.value; |
| | | if (!param_name.empty() && param_name[0] == '$') { // '$' eltávolítása |
| | | param_name = param_name.substr(1); |
| | | } |
| | | |
| | | param_infos.push_back({ param_name, param_type }); |
| | | |
| | | // Vessző vagy zárójel következik? |
| | | if (match(Lexer::Tokens::Type::PUNCTUATION, ",")) { |
| | | continue; |
| | | } |
| | | if (currentToken().type == Lexer::Tokens::Type::PUNCTUATION && currentToken().value == ")") { |
| | | break; // Lista vége |
| | | } |
| | | reportError("Expected ',' or ')' in parameter list"); |
| | | } |
| | | } |
| | | // Most a ')' következik |
| | | expect(Lexer::Tokens::Type::PUNCTUATION, ")"); |
| | | |
| | | // check if we have a option return type: function name() type { ... } |
| | | for (const auto & _type : Parser::variable_types) { |
| | | if (match(_type.first)) { |
| | | func_return_type = _type.second; |
| | | break; |
| | | } |
| | | } |
| | | |
| | | Lexer::Tokens::Token opening_brace = expect(Lexer::Tokens::Type::PUNCTUATION, "{"); |
| | | |
| | | // only parse the body if we checked out if not exists the function and created the symbol |
| | | parseFunctionBody(opening_brace, func_name, func_return_type, param_infos); |
| | | } |
| | | void parseVariableDefinition(); |
| | | void parseFunctionDefinition(); |
| | | |
| | | // --- Elemzési Segédfüggvények --- |
| | | |
| | |
| | | return Symbols::Value(); // compiler happy |
| | | } |
| | | |
| | | Symbols::Value parseNumericLiteral(const std::string & value, bool is_negative, Symbols::Variables::Type type) { |
| | | try { |
| | | switch (type) { |
| | | case Symbols::Variables::Type::INTEGER: |
| | | { |
| | | if (value.find('.') != std::string::npos) { |
| | | throw std::invalid_argument("Floating point value in integer context: " + value); |
| | | } |
| | | int v = std::stoi(value); |
| | | return Symbols::Value(is_negative ? -v : v); |
| | | } |
| | | case Symbols::Variables::Type::DOUBLE: |
| | | { |
| | | double v = std::stod(value); |
| | | return Symbols::Value(is_negative ? -v : v); |
| | | } |
| | | case Symbols::Variables::Type::FLOAT: |
| | | { |
| | | float v = std::stof(value); |
| | | return Symbols::Value(is_negative ? -v : v); |
| | | } |
| | | default: |
| | | throw std::invalid_argument("Unsupported numeric type"); |
| | | } |
| | | } catch (const std::invalid_argument & e) { |
| | | reportError("Invalid numeric literal: " + value + " (" + e.what() + ")"); |
| | | } catch (const std::out_of_range & e) { |
| | | reportError("Numeric literal out of range: " + value + " (" + e.what() + ")"); |
| | | } |
| | | |
| | | return Symbols::Value(); // unreachable |
| | | } |
| | | Symbols::Value parseNumericLiteral(const std::string & value, bool is_negative, Symbols::Variables::Type type); |
| | | |
| | | void parseFunctionBody(const Lexer::Tokens::Token & opening_brace, const std::string & function_name, |
| | | Symbols::Variables::Type return_type, const Symbols::FunctionParameterInfo & params) { |
| | | size_t braceDepth = 0; |
| | | int peek = 0; |
| | | int tokenIndex = current_token_index_; |
| | | Lexer::Tokens::Token currentToken_; |
| | | Lexer::Tokens::Token closing_brace; |
| | | Symbols::Variables::Type return_type, const Symbols::FunctionParameterInfo & params); |
| | | |
| | | while (tokenIndex < tokens_.size()) { |
| | | currentToken_ = peekToken(peek); |
| | | if (currentToken_.type == Lexer::Tokens::Type::PUNCTUATION) { |
| | | if (currentToken_.value == "{") { |
| | | ++braceDepth; |
| | | } else if (currentToken_.value == "}") { |
| | | if (braceDepth == 0) { |
| | | closing_brace = currentToken_; |
| | | break; |
| | | } |
| | | --braceDepth; |
| | | } |
| | | } |
| | | tokenIndex++; |
| | | peek++; |
| | | } |
| | | if (braceDepth != 0) { |
| | | reportError("Unmatched braces in function body"); |
| | | } |
| | | std::vector<Lexer::Tokens::Token> filtered_tokens; |
| | | auto startIt = std::find(tokens_.begin(), tokens_.end(), opening_brace); |
| | | auto endIt = std::find(tokens_.begin(), tokens_.end(), closing_brace); |
| | | |
| | | if (startIt != tokens_.end() && endIt != tokens_.end() && startIt < endIt) { |
| | | filtered_tokens = std::vector<Lexer::Tokens::Token>(startIt + 1, endIt); |
| | | } |
| | | std::string_view input_string = input_str_view_.substr(opening_brace.end_pos, closing_brace.end_pos); |
| | | |
| | | current_token_index_ = tokenIndex; |
| | | expect(Lexer::Tokens::Type::PUNCTUATION, "}"); |
| | | const std::string newns = Symbols::SymbolContainer::instance()->currentScopeName() + "." + function_name; |
| | | Symbols::SymbolContainer::instance()->create(newns); |
| | | std::shared_ptr<Parser> parser = std::make_shared<Parser>(); |
| | | parser->parseScript(filtered_tokens, input_string, this->current_filename_); |
| | | Symbols::SymbolContainer::instance()->enterPreviousScope(); |
| | | // create function |
| | | Interpreter::OperationsFactory::defineFunction( |
| | | function_name, params, return_type, Symbols::SymbolContainer::instance()->currentScopeName(), |
| | | this->current_filename_, currentToken_.line_number, currentToken_.column_number); |
| | | } |
| | | |
| | | ParsedExpressionPtr parseParsedExpression(const Symbols::Variables::Type & expected_var_type) { |
| | | std::stack<std::string> operator_stack; |
| | | std::vector<ParsedExpressionPtr> output_queue; |
| | | |
| | | auto getPrecedence = [](const std::string & op) -> int { |
| | | if (op == "+" || op == "-") { |
| | | return 1; |
| | | } |
| | | if (op == "*" || op == "/") { |
| | | return 2; |
| | | } |
| | | if (op == "u-" || op == "u+") { |
| | | return 3; |
| | | } |
| | | return 0; |
| | | }; |
| | | |
| | | auto isLeftAssociative = [](const std::string & op) -> bool { |
| | | return !(op == "u-" || op == "u+"); |
| | | }; |
| | | |
| | | auto applyOperator = [](const std::string & op, ParsedExpressionPtr rhs, ParsedExpressionPtr lhs = nullptr) { |
| | | if (op == "u-" || op == "u+") { |
| | | std::string real_op = (op == "u-") ? "-" : "+"; |
| | | return ParsedExpression::makeUnary(real_op, std::move(rhs)); |
| | | } else { |
| | | return ParsedExpression::makeBinary(op, std::move(lhs), std::move(rhs)); |
| | | } |
| | | }; |
| | | |
| | | auto pushOperand = [&](const Lexer::Tokens::Token & token) { |
| | | if (token.type == Lexer::Tokens::Type::NUMBER || token.type == Lexer::Tokens::Type::STRING_LITERAL || |
| | | token.type == Lexer::Tokens::Type::KEYWORD) { |
| | | output_queue.push_back( |
| | | ParsedExpression::makeLiteral(Symbols::Value::fromString(token.value, expected_var_type))); |
| | | } else if (token.type == Lexer::Tokens::Type::VARIABLE_IDENTIFIER) { |
| | | std::string name = token.value; |
| | | if (!name.empty() && name[0] == '$') { |
| | | name = name.substr(1); |
| | | } |
| | | output_queue.push_back(ParsedExpression::makeVariable(name)); |
| | | } else { |
| | | reportError("Expected literal or variable"); |
| | | } |
| | | }; |
| | | |
| | | bool expect_unary = true; |
| | | |
| | | while (true) { |
| | | auto token = currentToken(); |
| | | |
| | | if (token.type == Lexer::Tokens::Type::PUNCTUATION && token.lexeme == "(") { |
| | | operator_stack.push("("); |
| | | consumeToken(); |
| | | expect_unary = true; |
| | | } else if (token.type == Lexer::Tokens::Type::PUNCTUATION && token.lexeme == ")") { |
| | | consumeToken(); |
| | | while (!operator_stack.empty() && operator_stack.top() != "(") { |
| | | std::string op = operator_stack.top(); |
| | | operator_stack.pop(); |
| | | |
| | | if (op == "u-" || op == "u+") { |
| | | if (output_queue.empty()) { |
| | | reportError("Missing operand for unary operator"); |
| | | } |
| | | auto rhs = std::move(output_queue.back()); |
| | | output_queue.pop_back(); |
| | | output_queue.push_back(applyOperator(op, std::move(rhs))); |
| | | } else { |
| | | if (output_queue.size() < 2) { |
| | | reportError("Malformed expression"); |
| | | } |
| | | auto rhs = std::move(output_queue.back()); |
| | | output_queue.pop_back(); |
| | | auto lhs = std::move(output_queue.back()); |
| | | output_queue.pop_back(); |
| | | output_queue.push_back(applyOperator(op, std::move(rhs), std::move(lhs))); |
| | | } |
| | | } |
| | | |
| | | if (operator_stack.empty() || operator_stack.top() != "(") { |
| | | reportError("Mismatched parentheses"); |
| | | } |
| | | operator_stack.pop(); // remove "(" |
| | | expect_unary = false; |
| | | } else if (token.type == Lexer::Tokens::Type::OPERATOR_ARITHMETIC) { |
| | | std::string op = std::string(token.lexeme); |
| | | if (expect_unary && (op == "-" || op == "+")) { |
| | | op = "u" + op; // pl. u- |
| | | } |
| | | |
| | | while (!operator_stack.empty()) { |
| | | const std::string & top = operator_stack.top(); |
| | | if ((isLeftAssociative(op) && getPrecedence(op) <= getPrecedence(top)) || |
| | | (!isLeftAssociative(op) && getPrecedence(op) < getPrecedence(top))) { |
| | | operator_stack.pop(); |
| | | |
| | | if (top == "u-" || top == "u+") { |
| | | if (output_queue.empty()) { |
| | | reportError("Missing operand for unary operator"); |
| | | } |
| | | auto rhs = std::move(output_queue.back()); |
| | | output_queue.pop_back(); |
| | | output_queue.push_back(applyOperator(top, std::move(rhs))); |
| | | } else { |
| | | if (output_queue.size() < 2) { |
| | | reportError("Malformed expression"); |
| | | } |
| | | auto rhs = std::move(output_queue.back()); |
| | | output_queue.pop_back(); |
| | | auto lhs = std::move(output_queue.back()); |
| | | output_queue.pop_back(); |
| | | output_queue.push_back(applyOperator(top, std::move(rhs), std::move(lhs))); |
| | | } |
| | | } else { |
| | | break; |
| | | } |
| | | } |
| | | |
| | | operator_stack.push(op); |
| | | consumeToken(); |
| | | expect_unary = true; |
| | | } else if (token.type == Lexer::Tokens::Type::NUMBER || token.type == Lexer::Tokens::Type::STRING_LITERAL || |
| | | token.type == Lexer::Tokens::Type::KEYWORD || |
| | | token.type == Lexer::Tokens::Type::VARIABLE_IDENTIFIER) { |
| | | pushOperand(token); |
| | | consumeToken(); |
| | | expect_unary = false; |
| | | } else { |
| | | break; |
| | | } |
| | | } |
| | | |
| | | // Kiürítjük az operator stack-et |
| | | while (!operator_stack.empty()) { |
| | | std::string op = operator_stack.top(); |
| | | operator_stack.pop(); |
| | | |
| | | if (op == "(" || op == ")") { |
| | | reportError("Mismatched parentheses"); |
| | | } |
| | | |
| | | if (op == "u-" || op == "u+") { |
| | | if (output_queue.empty()) { |
| | | reportError("Missing operand for unary operator"); |
| | | } |
| | | auto rhs = std::move(output_queue.back()); |
| | | output_queue.pop_back(); |
| | | output_queue.push_back(applyOperator(op, std::move(rhs))); |
| | | } else { |
| | | if (output_queue.size() < 2) { |
| | | reportError("Malformed expression"); |
| | | } |
| | | auto rhs = std::move(output_queue.back()); |
| | | output_queue.pop_back(); |
| | | auto lhs = std::move(output_queue.back()); |
| | | output_queue.pop_back(); |
| | | output_queue.push_back(applyOperator(op, std::move(rhs), std::move(lhs))); |
| | | } |
| | | } |
| | | |
| | | if (output_queue.size() != 1) { |
| | | reportError("Expression could not be parsed cleanly"); |
| | | } |
| | | |
| | | return std::move(output_queue.back()); |
| | | } |
| | | ParsedExpressionPtr parseParsedExpression(const Symbols::Variables::Type & expected_var_type); |
| | | |
| | | }; // class Parser |
| | | |