pzks-labs/src/parser.cpp

#include "expression.hpp"
#include "print.h"
#include <cctype>
#include <cstring>
#include <iostream>
#include <vector>

namespace Parser {

class Parser
{
	enum State {
		I, // idle
		RC, // read constant
		RV, // read variable
		OP, // read operator
	} s;

	const char *state_str(enum State e)
	{
		switch (e)
		{
		case I:
			return "I";
		case RC:
			return "RC";
		case RV:
			return "RV";
		case OP:
			return "OP";
		default:
			return "Not a valid state!";
		}
	}

	int switch_state(enum State e)
	{
		int cumulative_backtrack = 0;

		switch (s)
		{
		case I:
			cumulative_backtrack += leave_i();
			break;
		case RC:
			cumulative_backtrack += leave_rc();
			break;
		case RV:
			cumulative_backtrack += leave_rv();
			break;
		case OP:
			cumulative_backtrack += leave_op();
			break;
		default:
			pr_warn("Warning: broken state machine on leave");
			break;
		}

		switch (e)
		{
		case I:
			enter_i();
			break;
		case RC:
			enter_rc();
			break;
		case RV:
			enter_rv();
			break;
		case OP:
			enter_op();
			break;
		default:
			pr_warn("Warning: broken state machine on enter");
			break;
		}

		s = e;

		return cumulative_backtrack;
	}

public:
	char buffer[256];
	std::vector<Expression::Token>& pv;

	Parser(std::vector<Expression::Token>& pv)
		: s(I), pv(pv)
	{
		buf_clear();
	}

	void buf_clear(void)
	{
		memset(buffer, 0, MAX_TOKEN_STRING_LEN);
	}

	void buf_push(char c)
	{
		if (int i = strlen(buffer); i < MAX_TOKEN_STRING_LEN - 1) {
			buffer[i] = c;
		}
	}

	int dispatch_to_current(char c)
	{
		switch (s)
		{
		case I:
			return parse_i(c);
		case RC:
			return parse_rc(c);
		case RV:
			return parse_rv(c);
		case OP:
			return parse_op(c);
		default:
			pr_warn("State machine is broken, unable to dispatch character " << c << ", returning 0");
			return 0;
		}
	}

	void enter_i(void) {}
	int  parse_i(char c)
	{
		if (isdigit(c) || c == '.')
		{
			return switch_state(RC) + 1;
		}
		else if (isalpha(c))
		{
			return switch_state(RV) + 1;
		}
		else if (c == '+' || c == '-' || c == '*' || c == '/' || c == '(' || c == ')')
		{
			return switch_state(OP) + 1;
		}
		else
		{
			// as per requirements (report all wrong symbols)
			if (c != ' ' && c != '\0')
			{
				pr_warn("Unrecognized character: " << c);
			}

			return 0;
		}
	}
	int leave_i(void)
	{
		return 0;
	}

	void enter_rc(void)
	{
		buf_clear();
	}
	int  parse_rc(char c)
	{
		if (!(isdigit(c) || c == '.'))
		{
			return switch_state(I) + 1;
		}
		else
		{
			buf_push(c);
			return 0;
		}
	}
	int leave_rc(void)
	{
		// additional check as per requirements
		{
			int i, c;
#if (MAX_TOKEN_STRING_LEN % 8) == 0
			for (i = 0, c = 0; i < MAX_TOKEN_STRING_LEN; i += 8)
			{
				c += (buffer[i+0] == '.')
				   + (buffer[i+1] == '.')
				   + (buffer[i+2] == '.')
				   + (buffer[i+3] == '.')
				   + (buffer[i+4] == '.')
				   + (buffer[i+5] == '.')
				   + (buffer[i+6] == '.')
				   + (buffer[i+7] == '.');
			}
#else
			for (int i = 0, c = 0; i < MAX_TOKEN_STRING_LEN; i++)
			{
				c += buffer[i] == '.';
			}
#endif
			if (c > 1)
			{
				pr_warn("Too many dots in CONST: " << buffer);
				return 0;
			}
		}

		double value;

		if (!sscanf(buffer, "%lf", &value))
		{
			pr_warn("Invalid CONST: " << buffer);
		}
		else
		{
			pv.push_back(Expression::Token(value));
		}

		return 0;
	}

	void enter_rv(void)
	{
		buf_clear();
	}
	int  parse_rv(char c)
	{
		if (!isalpha(c))
		{
			return switch_state(I) + 1;
		}
		else
		{
			buf_push(c);
			return 0;
		}
	}
	int leave_rv(void)
	{
		pv.push_back(Expression::Token(buffer));
		return 0;
	}

	void enter_op(void)
	{
		buf_clear();
	}
	int  parse_op(char c)
	{
		if (!(c == '+' || c == '-' || c == '*' || c == '/' || c == '(' || c == ')'))
		{
			return switch_state(I) + 1;
		}
		else
		{
			buf_push(c);
			return 0;
		}
	}
	int leave_op(void)
	{
		int i, downsize;
		for (i = strlen(buffer), downsize = 0; i > 0; buffer[--i] = '\0', downsize++)
		{
			if      (!strcmp(buffer, "("))
			{
				pv.push_back(Expression::Token(Expression::Token::OPA));
				break;
			}
			else if (!strcmp(buffer, ")"))
			{
				pv.push_back(Expression::Token(Expression::Token::CPA));
				break;
			}
			else if (!strcmp(buffer, "+"))
			{
				pv.push_back(Expression::Token(Expression::Token::ADD));
				break;
			}
			else if (!strcmp(buffer, "-"))
			{
				pv.push_back(Expression::Token(Expression::Token::SUB));
				break;
			}
			else if (!strcmp(buffer, "*"))
			{
				pv.push_back(Expression::Token(Expression::Token::MUL));
				break;
			}
			else if (!strcmp(buffer, "/"))
			{
				pv.push_back(Expression::Token(Expression::Token::DIV));
				break;
			}
			else if (!strcmp(buffer, "**"))
			{
				pv.push_back(Expression::Token(Expression::Token::POW));
				break;
			}
		}

		if (i)
		{
			return downsize;
		}
		else
		{
			pr_warn("Skip unknown operator: " << buffer);
			return 0;
		}
	}
};

std::vector<Expression::Token> tokenize_input(const char * const input_string)
{
	std::vector<Expression::Token> pv;

	Parser p(pv);
	for (int i = 0, l = strlen(input_string) + 1; i < l; i++)
	{
		i -= p.dispatch_to_current(input_string[i]);
	}

	return pv;
}

int assert_same(const std::vector<Expression::Token>& a, const std::vector<Expression::Token>& b)
{
	if (a.size() != b.size())
	{
		std::cout << "Vectors are of different lengths: " << a.size() << " != " << b.size() << "\n";
		return 1;
	}

	for (int i = 0, l = a.size(); i < l; i++)
	{
		if (a[i].t != b[i].t)
		{
			std::cout << "Token " << i << " types differ: " << a[i].type_str() << " != " << b[i].type_str() << "\n";
			return 1;
		}

		switch (a[i].t)
		{
		case Expression::Token::VAR:
			if (strcmp((*(a[i].v.s)).c_str(),
				   (*(b[i].v.s)).c_str()))
			{
				std::cout << "Variable names at " << i << " differ: " << (*(a[i].v.s)).c_str() << " != " << (*(b[i].v.s)).c_str() << "\n";
				return 1;
			}
			break;
		case Expression::Token::CONST:
			if (a[i].v.d != b[i].v.d)
			{
				std::cout << "Constants at " << i << " differ: " << a[i].v.d << " != " << b[i].v.d << "\n";
				return 1;
			}
			break;
		case Expression::Token::OPERATOR:
			if (a[i].v.o != b[i].v.o)
			{
				std::cout << "Operators at " << i << " differ: " << a[i].op_str() << " != " << b[i].op_str() << "\n";
				return 1;
			}
			break;
		default:
			pr_warn("Token skipped");
			break;
		}
	}

	return 0;
}

void print_token_vector(const std::vector<Expression::Token>& tv)
{
	int i, l;
	for (i = 0, l = tv.size() - 1; i < l; i++)
	{
		std::cout << "(" << *tv[i].to_str() << ") ";
	}

	if (tv.size())
	{
		std::cout << "(" << *tv[i].to_str() << ")\n";
	}
}

void print_math_vector(const std::vector<Expression::Token>& tv)
{
	int i, l;
	for (i = 0, l = tv.size() - 1; i < l; i++)
	{
		std::cout << *tv[i].to_math_str(); //<< " ";
	}

	if (tv.size())
	{
		std::cout << *tv[i].to_math_str() << "\n";
	}
}

void print_math_vector(const std::vector<Expression::Token>& tv, const std::vector<int>& highlight_locations)
{
	int i, l, h, hl;
	for (i = 0, l = tv.size() - 1, h = 0, hl = highlight_locations.size(); i < l; i++)
	{
		if (h < hl && highlight_locations[h] == i)
		{
			std::cout << "\033[31m" << *tv[i].to_math_str() << "\033[0m";
			h++;
		}
		else
		{
			std::cout << *tv[i].to_math_str();;
		}
	}

	if (tv.size())
	{
		if (highlight_locations[highlight_locations.size()-1] == (long) (tv.size() - 1))
		{
			std::cout << "\033[31m" << *tv[tv.size() - 1].to_math_str() << "\033[0m\n";
		}
		else
		{
			std::cout << *tv[tv.size() - 1].to_math_str() << "\n";
		}
	}
}

}