private void RunButton_Click(object sender, RoutedEventArgs e)
{
txtOutput.Text = ""; //reset output text
Input = txtInput.Text; //reset input text to new text
ListOfTokens.Clear(); //reset table
getTokens(Input);
if (Flag) //check if there is an error
{
foreach (TokenReg token in ListOfTokens)
{
txtOutput.Text += token.ToString();
}
//Parser program
try
{
removeCommentTokens(); //remove comment tokens from list as we dont need them in the parsing process
Parser p = new Parser(ListOfTokens);
SyntaxTree s = p.program();
TreeViewItem treeViewItem = makeTreeView(s);
Window1 secondWindow = new Window1();
secondWindow.mainTreeView.Items.Add(treeViewItem);
secondWindow.Show();
}
catch (ParsingErrorException ex)
{
MessageBox.Show(ex.Message, ex.Source,
MessageBoxButton.OK, MessageBoxImage.Error);
}
}
}
private void removeCommentTokens()
{
for (int i = 0; i < ListOfTokens.Count; i++)
{
if (ListOfTokens.ElementAt(i).Token == TokenReg.TOKEN_TYPE.T_COMMENT)
{
ListOfTokens.Remove(ListOfTokens.ElementAt(i));
}
}
}
/// <summary>
/// takes a string as input and divides it into lexemes and tokens adding them into the listOfTokens variable
/// </summary>
/// <param name="input">input string to compute</param>
private void getTokens(string input)
{
Flag = true; //this variable indicates that the function finished succesfully with no errors
int j = 0; //integer to keep track of the first index of the lexeme
int i; //integer to keep track of the last index of the lexeme
STATE currentState = STATE.Start;
for (i = 0; i < input.Length; i++)
{
try
{
switch (currentState)
{
case STATE.Start:
j = i; //to keep track of the first element's index
if (isLetter(input.ElementAt(i)))
{
if (i == input.Length - 1) //we do this check to not neglect the last letter
{
i--;
}
currentState = STATE.Identifier;
}
else if (isDigit(input.ElementAt(i)))
{
if (i == input.Length - 1) //we do this check to not neglect the last digit
{
i--;
}
currentState = STATE.Num1;
}
else if (isWhiteSpace(input.ElementAt(i)))
{
//do nothing
}
else if (input.ElementAt(i) == '/')
{
if (i == input.Length - 1)
{
i--;
}
currentState = STATE.Comment1;
}
else if (input.ElementAt(i) == '"')
{
currentState = STATE.String;
}
else if (input.ElementAt(i) == '(')
{
ListOfTokens.AddLast(new TokenReg(TokenReg.TOKEN_TYPE.T_LEFTBRACKET, "("));
}
else if (input.ElementAt(i) == ')')
{
ListOfTokens.AddLast(new TokenReg(TokenReg.TOKEN_TYPE.T_RIGHTBRACKET, ")"));
}
else if (input.ElementAt(i) == ';')
{
ListOfTokens.AddLast(new TokenReg(TokenReg.TOKEN_TYPE.T_SEMICOLON, ";"));
}
else if (input.ElementAt(i) == ',')
{
ListOfTokens.AddLast(new TokenReg(TokenReg.TOKEN_TYPE.T_COMMA, ","));
}
else if (input.ElementAt(i) == '+')
{
ListOfTokens.AddLast(new TokenReg(TokenReg.TOKEN_TYPE.T_PLUS, "+"));
}
else if (input.ElementAt(i) == '-')
{
ListOfTokens.AddLast(new TokenReg(TokenReg.TOKEN_TYPE.T_MINUS, "-"));
}
else if (input.ElementAt(i) == '*')
{
ListOfTokens.AddLast(new TokenReg(TokenReg.TOKEN_TYPE.T_MULTIPLICATION, "*"));
}
else if (input.ElementAt(i) == ':')
{
if (input.ElementAt(i + 1) == '=')
{
ListOfTokens.AddLast(new TokenReg(TokenReg.TOKEN_TYPE.T_ASSIGN, ":="));
i++; //we increment i to skip the equal sign evaluation
}
else
{
i--; //to catch the symbol that did the error
currentState = STATE.error; //since the colon char is never used alone it must always accompany the equal sign
}
}
else if (input.ElementAt(i) == '=')
{
ListOfTokens.AddLast(new TokenReg(TokenReg.TOKEN_TYPE.T_EQUAL, "="));
}
else if (input.ElementAt(i) == '<')
{
if (input.ElementAt(i + 1) == '>')
{
listOfTokens.AddLast(new TokenReg(TokenReg.TOKEN_TYPE.T_NOTEQUAL, "<>"));
i++; //we increment i to skip the > evaluation
}
else
{
ListOfTokens.AddLast(new TokenReg(TokenReg.TOKEN_TYPE.T_LESSTHAN, "<"));
}
}
else if (input.ElementAt(i) == '>')
{
ListOfTokens.AddLast(new TokenReg(TokenReg.TOKEN_TYPE.T_GREATERTHAN, ">"));
}
else if (input.ElementAt(i) == '&')
{
if (input.ElementAt(i + 1) == '&')
{
ListOfTokens.AddLast(new TokenReg(TokenReg.TOKEN_TYPE.T_AND, "&&"));
i++; //we increment i to skip the & evaluation
}
else
{
currentState = STATE.error; //same as the assign lexeme
}
}
else if (input.ElementAt(i) == '|')
{
if (input.ElementAt(i + 1) == '|')
{
ListOfTokens.AddLast(new TokenReg(TokenReg.TOKEN_TYPE.T_OR, "||"));
i++; //we increment i to skip the | evaluation
}
else
{
currentState = STATE.error; //same as the assign lexeme
}
}
else if (input.ElementAt(i) == '{')
{
ListOfTokens.AddLast(new TokenReg(TokenReg.TOKEN_TYPE.T_LEFTBRACE, "{"));
}
else if (input.ElementAt(i) == '}')
{
ListOfTokens.AddLast(new TokenReg(TokenReg.TOKEN_TYPE.T_RIGHTBRACE, "}"));
}
else
{
i--; //we decrement it as if it was the last index in the string we wouldnt cycle back to indicate an error occured
currentState = STATE.error; //anything else is out of the alphabet and therefore an error
}
break;
case STATE.Num1:
if (isDigit(input.ElementAt(i)) && i != input.Length - 1)
{
//do nothing
}
else if (input.ElementAt(i) == '.')
{
if (i == input.Length - 1)
{
i--; //decrement to not go out of loop
currentState = STATE.error; //because you cant end string with dot
}
else
{
currentState = STATE.Num2;
}
}
else if (isLetter(input.ElementAt(i)))
{
if (i == input.Length - 1)
{
i--;
}
currentState = STATE.error; //if a letter follows a number without another delimeter that means that this was supposed to be an invalid identifier and therefore we move it to the error state.
}
else
{
if (!isDigit(input.ElementAt(i))) //if last character in string is digit then we dont decrement the i
{
i--; //we decrement i as we do not want to include the [other] character in our number
}
ListOfTokens.AddLast(new TokenReg(TokenReg.TOKEN_TYPE.T_NUMBER, input.Substring(j, i - j + 1)));
currentState = STATE.Start; //go back to start state to accept new token
}
break;
case STATE.Num2:
if (isDigit(input.ElementAt(i)))
{
if (i == input.Length - 1) //we do this check to not neglect the last number after the dot
{
i--;
}
currentState = STATE.Num3;
}
else
{
if (i == input.Length - 1) //we do this check to not neglect the last symbol after the dot and advance to the error state
{
i--;
}
currentState = STATE.error; //as we dont expect anything other than a digit to follow
}
break;
case STATE.Num3:
if (isDigit(input.ElementAt(i)) && i != input.Length - 1)
{
//do nothing
}
else if (isLetter(input.ElementAt(i)))
{
if (i == input.Length - 1)
{
i--;
}
currentState = STATE.error; //if a letter follows a number without another delimeter that means that this was supposed to be an invalid identifier and therefore we move it to the error state.
}
else
{
if (!isDigit(input.ElementAt(i)))
{
i--; //we decrement i as we do not want to include the [other] character in our number
}
ListOfTokens.AddLast(new TokenReg(TokenReg.TOKEN_TYPE.T_NUMBER, input.Substring(j, i - j + 1)));
currentState = STATE.Start; //go back to start state to accept new token
}
break;
case STATE.Identifier:
if ((isLetter(input.ElementAt(i)) || isDigit(input.ElementAt(i))) && i != input.Length - 1) // i != input.Length - 1 to make sure that the last lexeme is taken into consideration without doing nothing
{
//do nothing
}
else
{
if (!(isLetter(input.ElementAt(i)) || isDigit(input.ElementAt(i)))) //we dont want to decrement i if it is the last character of a lexeme so we make that check
{
i--; //we decrement i as we do not want to include the [other] character in our identifier
}
TokenReg.TOKEN_TYPE tokenVal = isReservedWord(input.Substring(j, i - j + 1)); //we check if it is one of the reserved keywords first and register it in the tokenVal variable, if its not token Val will have a value of T_ID
ListOfTokens.AddLast(new TokenReg(tokenVal, input.Substring(j, i - j + 1)));
currentState = STATE.Start; //go back to start state to accept new token
}
break;
case STATE.String:
if (input.ElementAt(i) == '"')
{
ListOfTokens.AddLast(new TokenReg(TokenReg.TOKEN_TYPE.T_STRINGVAL, input.Substring(j + 1, i - j - 1))); //we put the start of substring as j + 1 to ignore the quotation marks; same thing for i - j - 1 instead of i - j + 1
currentState = STATE.Start; //go back to start state to accept new token
}
break;
case STATE.Comment1:
if (input.ElementAt(i) == '*')
{
currentState = STATE.Comment2;
}
else
{
if (input.ElementAt(i) != '/')
{
i--; //as we just want the division lexeme so we decrement and go back to start state
}
ListOfTokens.AddLast(new TokenReg(TokenReg.TOKEN_TYPE.T_DIVISION, "/")); //as there was a / character and another character other than * appeared then we consider it as a division symbol rather than the start of a comment
currentState = STATE.Start; //go back to start state to accept new token
}
break;
case STATE.Comment2:
if (input.ElementAt(i) == '*')
{
currentState = STATE.Comment3;
}
break;
case STATE.Comment3:
if (input.ElementAt(i) == '/')
{
ListOfTokens.AddLast(new TokenReg(TokenReg.TOKEN_TYPE.T_COMMENT, input.Substring(j, i - j + 1)));
currentState = STATE.Start; //go back to start state to accept new token
}
else if (input.ElementAt(i) == '*')
{
//do nothing
}
else
{
currentState = STATE.Comment2;
}
break;
case STATE.error:
string errorString = input.Substring(0, i + 1); //split the string from start to where the error occured
int numLines = errorString.Split('\n').Length; //get the number of lines of where the error occured
txtOutput.Text = "error at line " + numLines + ".";
Flag = false; //indicating that an error has occured
return;
}
}
catch (ArgumentOutOfRangeException)
{
if (input.ElementAt(i) == '<')
{
ListOfTokens.AddLast(new TokenReg(TokenReg.TOKEN_TYPE.T_LESSTHAN, "<"));
}
else
{
if (i == input.Length - 1)
{
i--; //to not get out of loop before going to error state
}
currentState = STATE.error;
}
}
}
}