elna/source/lexer.cpp

#include "elna/lexer.hpp"

namespace elna
{
    source::source(const std::string& buffer)
        : m_buffer(buffer)
    {
    }

    source::const_iterator source::begin() const
    {
        return source::const_iterator(std::cbegin(m_buffer));
    }

    source::const_iterator source::end() const
    {
        Position end_position{ 0, 0 };

        return source::const_iterator(std::cend(m_buffer), end_position);
    }

    source::const_iterator::const_iterator(std::string::const_iterator buffer,
            const Position start_position)
        : m_buffer(buffer), m_position(start_position)
    {
    }

    const Position& source::const_iterator::position() const noexcept
    {
        return this->m_position;
    }

    source::const_iterator::reference source::const_iterator::operator*() const noexcept
    {
        return *m_buffer;
    }

    source::const_iterator::pointer source::const_iterator::operator->() const noexcept
    {
        return m_buffer.base();
    }

    source::const_iterator& source::const_iterator::operator++()
    {
        if (*this->m_buffer == '\n')
        {
            this->m_position.column = 1;
            ++this->m_position.line;
        }
        else
        {
            ++this->m_position.column;
        }
        std::advance(this->m_buffer, 1);

        return *this;
    }

    source::const_iterator& source::const_iterator::operator++(int)
    {
        auto tmp = *this;
        ++(*this);
        return *this;
    }

    bool source::const_iterator::operator==(const source::const_iterator& that) const noexcept
    {
        return this->m_buffer == that.m_buffer;
    }

    bool source::const_iterator::operator!=(const source::const_iterator& that) const noexcept
    {
        return !(*this == that);
    }

    Token::Token(const Type of, const char *value, Position position)
        : m_type(of), m_position(position)
    {
        std::size_t value_length = strlen(value);
        char *buffer = reinterpret_cast<char *>(malloc(value_length + 1));

        std::memcpy(buffer, value, value_length);
        buffer[value_length] = 0;

        m_value.identifier = buffer;
    }

    Token::Token(const Type of, std::int32_t number, Position position)
        : m_type(of), m_position(position)
    {
        m_value.number = number;
    }

    Token::Token(const Type of, Position position)
        : m_type(of), m_position(position)
    {
    }

    Token::Token(const Token& that)
        : m_type(that.of()), m_position(that.position())
    {
        *this = that;
    }

    Token::Token(Token&& that)
        : m_type(that.of()), m_position(that.position())
    {
        *this = std::move(that);
    }

    Token::~Token()
    {
        if (m_type == TOKEN_IDENTIFIER || m_type == TOKEN_OPERATOR)
        {
            std::free(const_cast<char*>(m_value.identifier));
        }
    }

    Token& Token::operator=(const Token& that)
    {
        m_type = that.of();
        m_position = that.position();
        if (that.of() == TOKEN_IDENTIFIER || that.of() == TOKEN_OPERATOR)
        {
            std::size_t value_length = strlen(that.identifier());
            char *buffer = reinterpret_cast<char *>(malloc(value_length + 1));

            std::memcpy(buffer, that.identifier(), value_length);
            buffer[value_length] = 0;

            m_value.identifier = buffer;
        }
        else if (that.of() == TOKEN_NUMBER)
        {
            m_value.number = that.number();
        }
        return *this;
    }

    Token& Token::operator=(Token&& that)
    {
        m_type = that.of();
        m_position = that.position();
        if (that.of() == TOKEN_IDENTIFIER || that.of() == TOKEN_OPERATOR)
        {
            m_value.identifier = that.identifier();
            that.m_value.identifier = nullptr;
        }
        else if (that.of() == TOKEN_NUMBER)
        {
            m_value.number = that.number();
        }
        return *this;
    }

    Token::Type Token::of() const noexcept
    {
        return m_type;
    }

    const char *Token::identifier() const noexcept
    {
        return m_value.identifier;
    }

    std::int32_t Token::number() const noexcept
    {
        return m_value.number;
    }

    const Position& Token::position() const noexcept
    {
        return m_position;
    }

    Token *lex(const char *buffer, CompileError *compile_error, std::size_t *length)
    {
        std::vector<Token> tokens;
        source input{ buffer };

        for (auto iterator = input.begin(); iterator != input.end();)
        {
            if (*iterator == ' ' || *iterator == '\n')
            {
            }
            else if (std::isdigit(*iterator))
            {
                tokens.emplace_back(
                        Token::TOKEN_NUMBER,
                        static_cast<std::int32_t>(*iterator - '0'),
                        iterator.position()
                );
            }
            else if (*iterator == '=')
            {
                tokens.emplace_back(Token::TOKEN_EQUALS, iterator.position());
            }
            else if (*iterator == '(')
            {
                tokens.emplace_back(Token::TOKEN_LEFT_PAREN, iterator.position());
            }
            else if (*iterator == ')')
            {
                tokens.emplace_back(Token::TOKEN_RIGHT_PAREN, iterator.position());
            }
            else if (*iterator == ';')
            {
                tokens.emplace_back(Token::TOKEN_SEMICOLON, iterator.position());
            }
            else if (*iterator == ',')
            {
                tokens.emplace_back(Token::TOKEN_COMMA, iterator.position());
            }
            else if (*iterator == '!')
            {
                tokens.emplace_back(Token::TOKEN_BANG, iterator.position());
            }
            else if (*iterator == '.')
            {
                tokens.emplace_back(Token::TOKEN_DOT, iterator.position());
            }
            else if (std::isalpha(*iterator))
            {
                std::string word;
                auto i = iterator;
                while (i != input.end() && std::isalpha(*i))
                {
                    word.push_back(*i);
                    ++i;
                }
                if (word == "const")
                {
                    tokens.emplace_back(Token::TOKEN_LET, iterator.position());
                }
                else if (word == "var")
                {
                    tokens.emplace_back(Token::TOKEN_VAR, iterator.position());
                }
                else
                {
                    tokens.emplace_back(Token::TOKEN_IDENTIFIER, word.c_str(), iterator.position());
                }
                iterator = i;
                continue;
            }
            else if (*iterator == '+' || *iterator == '-')
            {
                std::string _operator{ *iterator };

                tokens.emplace_back(Token::TOKEN_OPERATOR, _operator.c_str(), iterator.position());
            }
            else
            {
                *compile_error = CompileError("Unexpected next character", iterator.position());
                return nullptr;
            }
            ++iterator;
        }
        Token *target = reinterpret_cast<Token *>(malloc(tokens.size() * sizeof(Token) + sizeof(Token)));
        int i = 0;
        for (auto& token : tokens)
        {
            target[i] = std::move(token);
            ++i;
        }
        *length = i;

        return target;
    }
}