/**
* Used to determine the kind for a specific token. Token kinds are defined by
* GetNextToken. This procedure will return token kinds for any token except for
* a comment. This procedure does not check that all restrictions are met, only
* that a general pattern is matched. For example, if the token is a label, it
* will not check that the label is shorter than the maximum label length.
*
* @param token Holds the token for which the kind is to be determined.
* @param tokenKind Used to store the token kind of the specified token.
*
* @refcode S2.4.2.1, S2.4.2.2, S2.4.2.3, S2.4.2.4
* @errtest
* GetTokenKind was tested indirectly through GetToken. If the correct token
* kind was not returned then there was an issue with this procedure. As with
* GetToken, please see the testing plan for a full list of tests performed.
* @errmsg
* None
* @author Andrew Buelow
* @creation April 7, 2011
* @modlog
* - April 8, 2011 - Mark - Changed Tokenizer to work with enumerated types.
* - April 9, 2011 - Andrew - TokenKind was not case insensitive to literals, fixed.
* - April 10, 2011 - Andrew - Added token kind JumpCond as the jump conditions did
* not fit in the other token kinds.
* - April 10, 2011 - Andrew - Added the NOP flag "=0" to be considered a literal.
* - April 14, 2011 - Andrew - Due to a change in the specifications "=0" no longer
* denotes a NOP, so this has been removed from the literals.
* - April 16, 2011 - Andrew - Added Expression as a possible token kind.
* - April 19, 2011 - Jacob - Redid the regular expression for numbers so it didn't
* catch numbers as well.
* @teststandard Andrew Buelow
* @codestandard Mark Mathis
*/
public static void GetTokenKind(string token, out TokenKinds tokenKind)
{
/* Regular expression used to determine if all characters in the token are
* letters or numbers. */
Regex alphaNumeric = new Regex(@"[^0-9a-zA-Z]");
/* Regular expression used to determine if all characters in the token are
* numbers. */
Regex numeric = new Regex(@"^[\-\+]?\d+$");
/* Regular expression used to determine if all characters in the token match
* those of an expression. Expressions can have alphanumeric characters as
* well as '+', '-' and '*'. Note: '*' is not used for multiplication, it is
* used in start notation (see web documentation). */
Regex expression = new Regex(@"[^0-9A-Za-z\*\+\-]");
// Convert to uppercase to give user flexibility. Token is passed by value so
// this change will not affect the token that is returned to the user.
token = token.ToUpper();
// Check that we weren't given an empty string first so we can look at
// characters in the string later.
if (token == "")
{
tokenKind = TokenKinds.Empty;
}
// Can only be a label or command if it begins with a letter and is all letters or numbers.
else if (!alphaNumeric.IsMatch(token) && Char.IsLetter(token[0]))
{
tokenKind = TokenKinds.Label_Or_Command;
}
// Check if it is all numbers
else if (numeric.IsMatch(token))
{
tokenKind = TokenKinds.Number;
}
// Check if the token contains characters only found in expressions.
else if (!expression.IsMatch(token))
{
tokenKind = TokenKinds.Expression;
}
// Only looks for the four possible Literal flags. Does not check format of following chars.
else if (token.Length > 1 && (token.Substring(0, 2) == "X=" ||
token.Substring(0, 2) == "B=" ||
token.Substring(0, 2) == "I=" ||
token.Substring(0, 2) == "C="))
{
tokenKind = TokenKinds.Literal;
}
// Determines if the token is one of the six jump conditions
else if (token == "=" || token == "^=" || token == "<" ||
token == ">" || token == "<=" || token == ">=")
{
tokenKind = TokenKinds.JumpCond;
}
// If no other pattern is matched something is wrong in the token and we give an error.
else
{
tokenKind = TokenKinds.Error;
}
}
/**
* Returns the next token in the provided string. The next token includes all
* characters in the given string until the next ' ', ',', tab, or when the end of the
* string is reached. TokenKind is determined based on the characters in the token.
* Once the token is found it is removed from the given string along with the
* separator character.
*
* @param line The line of code from which to find the next token.
* @param token Used to store the value of the next token.
* @param tokenKind Used to store the token kind of the next token.
*
* @refcode N/A
* @errtest
* Tokenizer was tested using strings that were empty, contained extra whitespace,
* and mixed case letters. For a full list of tests please look at the testing plan.
* @errmsg
* None
* @author Andrew Buelow
* @creation April 6, 2011
* @modlog
* - April 7, 2011 - Andrew - Tokenizer now returns a token kind.
* - April 7, 2011 - Andrew - Fixed a bug that would result in an infinite loop
* when the end of the line was reached.
* - April 8, 2011 - Mark - Changed Tokenizer to work with enumerated types.
* - April 8, 2011 - Andrew - Fixed an aray out of bounds exception that occured when
* tokenizer was given an empty string.
* - April 16, 2011 - Andrew - Added Expression as a possible token kind.
* - April 18, 2011 - Andrew - Character literals would break when there was a space in the single quotes.
* Now using regular expressions to ensure only correct character literals are returned.
* @teststandard Andrew Buelow
* @codestandard Mark Mathis
*/
public static void GetNextToken(ref string line, out string token, out TokenKinds tokenKind)
{
// Write to the log before attempting to remove the next token.
Logger.Log("Getting new token.", "Tokenizer");
// Remove extra whitespace
line = line.Trim();
// If the next token is a comment, we don't need to split up the line and can return everything
// else as the token
if (line.Length > 0 && line[0] == ':')
{
token = line;
tokenKind = TokenKinds.Comment;
line = "";
}
else
{
// Holds the two pieces of the string after it has been split at the next space, comma or tab.
// The first piece is the next token, the other is the rest of the line.
string[] tokens;
if (line.ToUpper().StartsWith("C='"))
{
Regex charLiteral = new Regex("([Cc]='.{0,2}')");
Match match = charLiteral.Match(line);
if (match.Success)
{
token = match.Groups[1].Value;
tokenKind = TokenKinds.Literal;
// remove the token from the line
line = line.Substring(token.Length);
}
else
{
// Split the string at the next space, comma or tab and store the first part as the token.
tokens = line.Split(new char[] { ' ', ',', '\t' }, 2);
token = tokens[0];
// If we are at the end of a string the second item in the array will be empty so we need to
// set the line to be an empty string.
if (tokens.Length > 1)
{
line = tokens[1];
}
else
{
line = "";
}
tokenKind = TokenKinds.Error;
}
}
else
{
// Split the string at the next space, comma or tab and store the first part as the token.
tokens = line.Split(new char[] { ' ', ',', '\t' }, 2);
token = tokens[0];
// If we are at the end of a string the second item in the array will be empty so we need to
// set the line to be an empty string.
if (tokens.Length > 1)
{
line = tokens[1];
}
else
{
line = "";
}
// Determine the kind for this token.
GetTokenKind(token, out tokenKind);
}
}
// Trim spaces off the beginning of the returned line so the first character is the beginning of the next token.
line = line.TrimStart();
// Write to the log after the token has been successfully retrieved.
Logger.Log("Token acquired, returning.", "Tokenizer");
}