コード例 #1
0
ファイル: IOModule.cs プロジェクト: jarveson/und-projects
 private void VerifyAddToken(int lineNum, Token token)
 {
     if (token != null) {
         if (token.Type == TokenType.Illegal) {
             throw new LexicalException(lineNum, token.ErrorMsg);
         }
         else tokens.Add(token);
     }
 }
コード例 #2
0
ファイル: Tokenizer.cs プロジェクト: jarveson/und-projects
 private bool ProcessSingleQuoteState(Char character)
 {
     //this state is easy
     if (character == '\'') {
         //cool end of our 'string'
         //check it for length of 1
         if (currentStream.Length > 1) {
             lastFoundToken = new Token(TokenType.Illegal, currentStream);
             lastFoundToken.ErrorMsg = "Invalid character token";
         }
         else {
             lastFoundToken = new Token(TokenType.QuoteString, currentStream);
             currentState = State.InitialState;
         }
         return true;
     }
     else {
         //keep taking chars for our string
         currentStream += character;
         return false;
     }
 }
コード例 #3
0
ファイル: Tokenizer.cs プロジェクト: jarveson/und-projects
        private bool ProcessOperatorState(Char character)
        {
            //lets figure out what symbol we should report back!
            if (Char.IsWhiteSpace(character)) {
                //lets just verify what we have so far
                if (ExpectedOperatorTokens.ContainsKey(currentStream)) {
                    lastFoundToken = new Token(ExpectedOperatorTokens[currentStream], currentStream);
                    //not much more to see here
                    currentState = State.InitialState;
                    return true;
                }
                else {
                    //ok, u want to give me an invalid symbol
                    lastFoundToken = new Token(TokenType.Illegal, currentStream);
                    lastFoundToken.ErrorMsg = "Invalid Symbol/Operator!";
                    return true;
                }
            }
            else if (ExpectedOperatorTokens.ContainsKey(character.ToString())){
                //your not a comment are you? i dont want those
                if (currentStream+character == "(*"){
                    currentState = State.Comment;
                    return false;
                }
                if (currentStream + character == "*)") {
                    //ugh...end comment without a beginning? thats a paddlin
                    lastFoundToken = new Token(TokenType.Illegal, currentStream);
                    lastFoundToken.ErrorMsg = "Unmatched end of comment";
                    return true;
                }
                //another symbol eh? check if still valid
                if (ExpectedOperatorTokens.ContainsKey(currentStream+character)){
                    lastFoundToken = new Token(ExpectedOperatorTokens[currentStream+character], currentStream+character);
                    currentState = State.InitialState;
                    return true;
                }
                else {
                    if (currentStream.Length == 2){
                        //somethings fubared
                        lastFoundToken = new Token(TokenType.Illegal, currentStream);
                        lastFoundToken.ErrorMsg = "Illegal Symbol/Operator";
                        return true;
                    }
                    // just return the last token and spin up initial to delay parse of next symbol
                    // on paper this should always be found
                    lastFoundToken = new Token(ExpectedOperatorTokens[currentStream], currentStream);
                    if ( ProcessInitialState(character))
                        return true; //ugh, invalid
                    else return true;
                }
            }
            else if (character == '\'' || character == '\"') {
                //verify what we have, then send the char to initial state to take care of it all
                if (ExpectedOperatorTokens.ContainsKey(currentStream)) {
                    lastFoundToken = new Token(ExpectedOperatorTokens[currentStream], currentStream);
                    if (ProcessInitialState(character))
                        return true; //invalid yey
                    else return true;
                }
                else {
                    //invalid? report it
                    lastFoundToken = new Token(TokenType.Illegal, currentStream);
                    lastFoundToken.ErrorMsg = "Invalid Symbol/Operator";
                    return true;
                }
            }

            else if (Char.IsLetterOrDigit(character)) {
                //no longer a symbol, on paper, what we have should always be valid
                lastFoundToken = new Token(ExpectedOperatorTokens[currentStream], currentStream);
                //spin up initial to check next char
                if (ProcessInitialState(character))
                    return true;
                else return true;
            }
            else {
                //ugh what?
                lastFoundToken = new Token(TokenType.Illegal, currentStream);
                lastFoundToken.ErrorMsg = "State Machine broke processing symbol";
                return true;
            }
        }
コード例 #4
0
ファイル: Tokenizer.cs プロジェクト: jarveson/und-projects
 private bool ProcessNumberState(Char character)
 {
     //simple state, only take in more numbers until space or symbol
     if (Char.IsNumber(character)) {
         currentStream += character;
         return false;
     }
     else if (Char.IsWhiteSpace(character)) {
         //got a token!
         lastFoundToken = new Token(TokenType.Number, currentStream);
         currentState = State.InitialState;
         return true;
     }
     else if (ExpectedOperatorTokens.ContainsKey(character.ToString())) {
         //still a token
         lastFoundToken = new Token(TokenType.Number, currentStream);
         //spin up initial state
         if (ProcessInitialState(character))
             return true; //ugh, invalid character is next up
         else return true;
     }
     else {
         //should be everything else
         lastFoundToken = new Token(TokenType.Illegal, currentStream);
         lastFoundToken.ErrorMsg = "Invalid Number expression";
         return true;
     }
 }
コード例 #5
0
ファイル: Tokenizer.cs プロジェクト: jarveson/und-projects
        //initialstate function
        //true if invalid token, false for no token to report
        private bool ProcessInitialState(Char character)
        {
            currentStream = "";
            //lets do this one first, not sure if any of the other cheatchecks will steal this from us
            if (character == '\'') {
                //eat the quote, go to state
                currentState = State.SingleQuote;
            }
            else if (character == '\"') {
                //must be a string
                currentState = State.DoubleQuote;
            }
            else if (Char.IsLetter(character)) {
                //parse this 'string'
                currentStream += character;
                currentState = State.Character;
            }

            else if (Char.IsWhiteSpace(character)) {
                //eat dat whitespace, nom nom
            }

            else if (Char.IsNumber(character)) {
                //k we got a number
                currentStream += character;
                currentState = State.Number;
            }
            //this one gets hairy, IsSymbol wont catch all our cases,
            //the idea is the piggyback off a single char in dict, then figure out if theres more in the state
            else if (ExpectedOperatorTokens.ContainsKey(character.ToString())) {
                currentStream += character;
                currentState = State.Operator;
            }

            else { //well, i got nothin, illegal token it is
                lastFoundToken = new Token(TokenType.Illegal, character.ToString());
                return true;
            }
            return false; //only return true for an error
        }
コード例 #6
0
ファイル: Tokenizer.cs プロジェクト: jarveson/und-projects
 private bool ProcessDoubleQuoteState(Char character)
 {
     //this state is easy
     if (character == '\"') {
         //cool end of our 'string'
         lastFoundToken = new Token(TokenType.DoubleQuoteString, currentStream);
         currentState = State.InitialState;
         return true;
     }
     else {
         //keep taking chars for our string
         currentStream += character;
         return false;
     }
 }
コード例 #7
0
ファイル: Tokenizer.cs プロジェクト: jarveson/und-projects
        private bool ProcessCharacterState(Char character)
        {
            bool updateToken = false;
            string tempStream = "";
            if (Char.IsWhiteSpace(character)) {
                //easy, eat the whitespace and check the token, hold in our buffer
                updateToken = true;
                currentState = State.InitialState;
                tempStream = currentStream;
            }
            else if (ExpectedOperatorTokens.ContainsKey(character.ToString())) {
                //little bit more tough, check token but also have to
                //spin up initialstate otherwise we will lose a char
                updateToken = true;
                tempStream = currentStream;
                currentStream += character;
                if (ProcessInitialState(character))
                    return true; //ugh..invalid next token, kill me now
            }
            else if (Char.IsLetterOrDigit(character)) {
                //easy, lets keep gettin more characters
                currentStream += character;
                return false;
            }
            else {
                //bad symbol
                lastFoundToken = new Token(TokenType.Illegal, character.ToString());
                lastFoundToken.ErrorMsg = "Expected Letter or Digit";
                return true;
            }
            if (updateToken) {
                //We have a token! Lets find a type for it,
                //dont touch state here
                if (ExpectedReserveTokens.ContainsKey(tempStream)) {
                    lastFoundToken = new Token(ExpectedReserveTokens[tempStream], tempStream);
                }
                else {
                    //must just be an identifier
                    lastFoundToken = new Token(TokenType.Identifier, tempStream);
                }
                return true;
            }
            else return false; // no token found so idk

            /*
            //this shouldnt happen, i guess make it known
            lastFoundToken = new Token(TokenType.Illegal, character.ToString());
            lastFoundToken.ErrorMsg = "Hit unknown condition for Character State! Token: " + character.ToString();
            return true;*/
        }
コード例 #8
0
ファイル: Tokenizer.cs プロジェクト: jarveson/und-projects
 //call at end of line, will return last token, or null if theres none to report
 //or illegal token if theres a token error
 public Token EndOfLine()
 {
     switch (currentState) {
         case State.InitialState:
             //nothing to report, all whitespace
             return null;
         case State.Character:
             //cheat and call the function with a 'space' character;
             ProcessCharacterState(' ');
             break;
         case State.Operator:
             ProcessOperatorState(' ');
             break;
         case State.Number:
             ProcessNumberState(' ');
             break;
         case State.SingleQuote:
             // rut rho!
             lastFoundToken = new Token(TokenType.Illegal, currentStream);
             lastFoundToken.ErrorMsg = "Missing Matching Single Quote";
             break;
         case State.DoubleQuote:
             lastFoundToken = new Token(TokenType.Illegal, currentStream);
             lastFoundToken.ErrorMsg = "Missing Matching double Quote";
             break;
         case State.Comment:
             //rut rho!
             lastFoundToken = new Token(TokenType.Illegal, currentStream);
             lastFoundToken.ErrorMsg = "Missing Matching Comment Symbol!";
             break;
     }
     return lastFoundToken;
 }