public ParsingTableGenerationTests()
{
E = new NonTerminal("E");
T = new NonTerminal("T");
F = new NonTerminal("F");
tokenizer = new RegExpTokenizer();
tokenizer.SetTransitionFunction(new TableDrivenTransitionFunction());
id = tokenizer.UseTerminal(RegExp.GetNumberRegExp());
plus = tokenizer.UseTerminal(RegExp.Literal('+'));
mult = tokenizer.UseTerminal(RegExp.Literal('*'));
leftBrace = tokenizer.UseTerminal(RegExp.Literal('('));
rightBrace = tokenizer.UseTerminal(RegExp.Literal(')'));
grammar = new AugmentedGrammar()
{
E --> E & plus & T,
E --> T,
T --> T & mult & F,
T --> F,
F --> leftBrace & E & rightBrace,
F --> id
};
}
/// <summary>
/// Gets value of last terminal symbol provided.
/// </summary>
/// <param name="term">Terminal symbol used in the production rule.</param>
/// <exception cref="ArgumentOutOfRangeException">Thrown in case there is no such token value.</exception>
public Token this[Terminal term]
{
// TODO: Reverse order of popping tokens
get
{
if (term == null) throw new ArgumentNullException("term");
if (mPoppedTokensCount < mPopableTokensCount)
{
int stackPeek = mTokens.Count - 1;
// TODO: Check whether it's enough to reverse this loop
for (int i = stackPeek; i >= 0; i--)
{
if (stackPeek - mPopableTokensCount == i)
break;
if (mTokens[i].Class == term.TokenClassID)
{
Token token = mTokens[i];
mTokens.RemoveAt(i);
mPoppedTokensCount++;
return token;
}
}
}
throw new ArgumentOutOfRangeException("term",
"The terminal symbol " + term +
" is not a part of the production.");
}
}
//TODO: Func<Token, bool> => Predicate<Token>
public Terminal UseTerminal(RegExp regExp, Func<Token, bool> lexicalAction = null)
{
FiniteAutomata tokenNfa = GetTokenNfa(regExp);
if (lexicalAction != null)
{
mLexicalActions[tokenNfa.Terminator.TokenClass] = lexicalAction;
}
AddTokenToNfa(tokenNfa);
var terminal = new Terminal(regExp.ToString(), mClassIdProvider.GetCurrent());
return terminal;
}
private void SetUpLexer()
{
// Ignore white space
lexer.IgnoreTerminal(RegExp.AtLeastOneOf(RegExp.Choice(RegExp.Literal(' '), RegExp.Literal('\t'))));
// Ignore comments (![a-zA-Z0-9...]\n)
lexer.IgnoreTerminal(RegExp.Sequence(RegExp.Literal('!'),
RegExp.AnyNumberOf((
RegExp.Range(' ', (char)255, encoding)))));
// New line is delimiter between code lines
mNewLine =
lexer.UseTerminal(RegExp.Choice(RegExp.Literal('\n'), RegExp.Literal(0x0D), RegExp.Literal(0x0A)));
mIdentifier =
lexer.UseTerminal(
RegExp.Sequence(
RegExp.Choice(RegExp.Range('a', 'z', encoding), RegExp.Range('A', 'Z', encoding),
RegExp.Literal('_')),
RegExp.AnyNumberOf(RegExp.Choice(RegExp.Range('a', 'z', encoding), RegExp.Range('A', 'Z', encoding),
RegExp.Literal('_'), RegExp.Range('0', '9', encoding)))));
mRegister = lexer.UseTerminal(RegExp.Sequence(RegExp.Literal('r'), RegExp.Range('0', '9', encoding), RegExp.Optional(RegExp.Range('0', '9', encoding))));
mNumber = lexer.UseTerminal(RegExp.Sequence(RegExp.Optional(RegExp.Literal('-')), RegExp.AtLeastOneOf(RegExp.Range('0', '9', encoding))));
mHexNumber =
lexer.UseTerminal(RegExp.Sequence(RegExp.Literal("0x"),
RegExp.AtLeastOneOf(RegExp.Choice(
RegExp.Range('0', '9', encoding),
RegExp.Range('a', 'f', encoding)))));
mComma = lexer.UseTerminal(RegExp.Literal(','));
mDot = lexer.UseTerminal(RegExp.Literal('.'));
mSemicolon = lexer.UseTerminal(RegExp.Literal(';'));
mColon = lexer.UseTerminal(RegExp.Literal(':'));
mSlash = lexer.UseTerminal(RegExp.Literal('/'));
mSquareBraceOpen = lexer.UseTerminal(RegExp.Literal('['));
mSquareBraceClose = lexer.UseTerminal(RegExp.Literal(']'));
mPlus = lexer.UseTerminal(RegExp.Literal('+'));
mDirective = lexer.UseTerminal(RegExp.Sequence(RegExp.Literal('.'), RegExp.Choice(RegExp.Literal("include"), RegExp.Literal("text"), RegExp.Literal("if"), RegExp.Literal("else"), RegExp.Literal("endif")),
RegExp.AnyNumberOf(
RegExp.Range(' ', (char)255, encoding))));
mDotC = lexer.UseTerminal(RegExp.Literal(".c"));
}
private void SetUpLexer()
{
mIdentifier = lexer.UseTerminal(RegExp.Sequence( // ([a-z]+::)?[a-z]+
RegExp.Optional( // ([a-z]+::)?
RegExp.Sequence( //[a-z]+::
RegExp.AtLeastOneOf(RegExp.Range('a', 'z', encoding)),
RegExp.Literal("::"))
),
RegExp.AtLeastOneOf(RegExp.Range('a', 'z', encoding))
));
mEqualSign = lexer.UseTerminal(RegExp.Literal('='));
// Ignore white space
lexer.IgnoreTerminal(RegExp.AtLeastOneOf(RegExp.Choice(RegExp.Literal(' '), RegExp.Literal('\t'), RegExp.Literal('\n'), RegExp.Literal(0x0D), RegExp.Literal(0x0A))));
mPlusSign = lexer.UseTerminal(RegExp.Literal('+'));
mMinusSign = lexer.UseTerminal(RegExp.Literal('-'));
mAsterisk = lexer.UseTerminal(RegExp.Literal('*'));
mDevideSign = lexer.UseTerminal(RegExp.Literal('/'));
mNumber = lexer.UseTerminal(RegExp.Sequence(RegExp.Optional(RegExp.Literal('-')), RegExp.AtLeastOneOf(RegExp.Range('0', '9', encoding))));
mSemicolon = lexer.UseTerminal(RegExp.Literal(';'));
mLeftBrace = lexer.UseTerminal(RegExp.Literal('('));
mRightBrace = lexer.UseTerminal(RegExp.Literal(')'));
mComma = lexer.UseTerminal(RegExp.Literal(','));
}
