#include "escape.h"
#include <stdexcept>

using namespace ts;
using namespace std;

std::string query::escape(std::string in) {
	size_t index = 0;
	while(index < in.length()) {
		if(in[index] == '\\')
			in.replace(index, 1, "\\\\", 2);
		else if(in[index] == ' ')
			in.replace(index, 1, "\\s", 2);
		else if(in[index] == '/')
			in.replace(index, 1, "\\/", 2);
		else if(in[index] == '|')
			in.replace(index, 1, "\\p", 2);
		else if(in[index] == '\b')
			in.replace(index, 1, "\\b", 2);
		else if(in[index] == '\f')
			in.replace(index, 1, "\\f", 2);
		else if(in[index] == '\n')
			in.replace(index, 1, "\\n", 2);
		else if(in[index] == '\r')
			in.replace(index, 1, "\\r", 2);
		else if(in[index] == '\t')
			in.replace(index, 1, "\\t", 2);
		else if(in[index] == '\x07')
			in.replace(index, 1, "\\a", 2);
		else if(in[index] == '\x0B')
			in.replace(index, 1, "\\v", 2);
		else {
			index += 1;
			continue;
		}
		index += 2;
	}

	return in;
}

/*
if (input[i] = $ED) and   // Prüfe auf High and Low Surrogate
(input[i+3] = $ED) and
(input[i+1] or $F = $AF) and
(input[i+4] or $F = $BF) then
begin
output := output +
		AnsiChar((input[i+1] and $F + 1) shr 2 and 7 or $F0) + // and 7 kann man auch weglassen
		AnsiChar((input[i+2] shr 2 and $F + (input[i+1] and $F + 1) shl 4 and $30) or $80) +
		AnsiChar((input[i+4] and $F + input[i+2] shl 4 and $30) or $80) +
		AnsiChar(input[i+5]);
end
else
output := output + AnsiChar(input);
 */
std::string query::unescape(std::string in, bool throw_error) {
	size_t index = 0;
	while(index < in.length()){
		if(in[index] == '\\'){
			if(in.length() <= index + 1) break;
			char replace = 0;

			switch (in[index + 1]){
				case 's': replace = ' '; break;
				case '/': replace = '/'; break;
				case 'p': replace = '|'; break;
				case 'b': replace = '\b'; break;
				case 'f': replace = '\f'; break;
				case 'n': replace = '\n'; break;
				case 'r': replace = '\r'; break;
				case 't': replace = '\t'; break;
				case 'a': replace = '\x07'; break;
				case 'v': replace = '\x0B'; break;
				case '\\': replace = '\\'; break;
				default:
					replace = '\x00'; break;
			}
			if(replace != 0x00)
				in.replace(index, 2, string(&replace, 1));
		}

		uint8_t current = (uint8_t) in[index];
		if(in.length() - index >= 6) { //Check for CESU-8
			if((current == 0xED) && ((uint8_t) in[index + 3] == 0xED) && ((((uint8_t) in[index + 1] | 0xF) == 0xAF)) && ((((uint8_t) in[index + 4] | 0xF) == 0xBF))){
				char replaced[4];

				replaced[0] = (((((uint8_t) in[index + 1] & 0xF) + 1) >> 2) & 7) | 0xF0;
				replaced[1] = ((((uint8_t) in[index + 2] >> 2) & 0xF) + (((((uint8_t) in[index + 1] & 0xF)) + (1 << 4)) | 0x80));
				replaced[2] = (((uint8_t) in[index + 4] & 0xF) + (((uint8_t) in[index + 2] << 4) & 0x30)) | 0x80;
				replaced[3] = ((uint8_t) in[index + 5]);

				in.replace(index, 6, string(replaced, 4));
				index -= 2; //Substract the replace difference

				index += 5; //Skip 5 bytes (CESU-8 length for emoty)
			}
		}
		//"\221\210\200\200\200\210\221\221\221\221\204\210\221\221\221\221\221\221\221\221\221\204\210\221\221\221\221\221\204\210\221\221"
		if(current >= 128) { //UTF8 check
			if(current >= 192 && (current <= 193 || current >= 245)) {
				in.replace(index, 1, "", 0); //Cut the character out
				index--;
			} else if(current >= 194 && current <= 223) {
				if(in.length() - index <= 1)
					in.replace(index, in.length() - index, "", 0);
				else if((uint8_t) in[index + 1] >= 128 && (uint8_t) in[index + 1] <= 191) index += 1; //Valid
				else {
					if(throw_error) throw invalid_argument("Invalid UTF-8 character at index " + to_string(index));
					in.replace(index, 2, "", 0); //Cut the two characters out
					index--;
				}
			} else if(current >= 224 && current <= 239) {
				if(in.length() - index <= 2)
					in.replace(index, in.length() - index, "", 0);
				else if((uint8_t) in[index + 1] >= 128 && (uint8_t) in[index + 1] <= 191 &&
				        (uint8_t) in[index + 2] >= 128 && (uint8_t) in[index + 2] <= 191) index += 2; //Valid
				else {
					if(throw_error) throw invalid_argument("Invalid UTF-8 character at index " + to_string(index));
					in.replace(index, 3, "", 0); //Cut the three characters out
					index--;
				}
			} else if(current >= 240 && current <= 244) {
				if(in.length() - index <= 3)
					in.replace(index, in.length() - index, "", 0);
				else if((uint8_t) in[index + 1] >= 128 && (uint8_t) in[index + 1] <= 191 &&
				        (uint8_t) in[index + 2] >= 128 && (uint8_t) in[index + 2] <= 191 &&
				        (uint8_t) in[index + 3] >= 128 && (uint8_t) in[index + 3] <= 191) index += 3; //Valid
				else {
					if(throw_error) throw invalid_argument("Invalid UTF-8 character at index " + to_string(index));
					in.replace(index, 4, "", 0); //Cut the three characters out
					index--;
				}
			} else {
				if(throw_error) throw invalid_argument("Invalid UTF-8 character at index " + to_string(index));
				in.replace(index, 1, "", 0); //Cut the character out
				index--;
			}
		}
		index++;
	}
	return in;
}