/// <summary>
/// Creates a new FA that will match a repetition of one or more of the specified FA expression
/// </summary>
/// <param name="expr">The expression to repeat</param>
/// <param name="accept">The symbol to accept</param>
/// <returns>A new FA that matches the specified FA one or more times</returns>
public static FA Repeat(FA expr, string accept = "")
{
var result = expr.Clone();
result.FirstAcceptingState.EpsilonTransitions.Add(result);
result.FirstAcceptingState.AcceptingSymbol = accept;
return(result);
}
public TokenEnumerator(FA lexer, IEnumerable <char> @string)
{
_lexer = lexer;
_input = @string.GetEnumerator();
_buffer = new StringBuilder();
_initialStates = _lexer.FillEpsilonClosure();
Reset(); // Reset is used here to initialize the rest of the values
}
/// <summary>
/// Creates a new FA that matches the specified FA expression or empty
/// </summary>
/// <param name="expr">The expression to make optional</param>
/// <param name="accept">The symbol to accept</param>
/// <returns>A new FA that will match the specified expression or empty</returns>
public static FA Optional(FA expr, string accept = "")
{
var result = expr.Clone();
var f = result.FirstAcceptingState;
f.AcceptingSymbol = accept;
result.EpsilonTransitions.Add(f);
return(result);
}
/// <summary>
/// Creates an FA that will match any one of a set of a characters
/// </summary>
/// <param name="set">The set of characters that will be matched</param>
/// <param name="accept">The symbol to accept</param>
/// <returns>An FA that will match the specified set</returns>
public static FA Set(IEnumerable <char> set, string accept = "")
{
var result = new FA();
var final = new FA();
final.AcceptingSymbol = accept;
foreach (char ch in set)
{
result.Transitions.Add(ch, final);
}
return(result);
}
/// <summary>
/// Creates an FA that matches a literal string
/// </summary>
/// <param name="string">The string to match</param>
/// <param name="accept">The symbol to accept</param>
/// <returns>A new FA machine that will match this literal</returns>
public static FA Literal(IEnumerable <char> @string, string accept = "")
{
var result = new FA();
var current = result;
foreach (char ch in @string)
{
current.AcceptingSymbol = null;
var fa = new FA();
fa.AcceptingSymbol = accept;
current.Transitions.Add(ch, fa);
current = fa;
}
return(result);
}
/// <summary>
/// Creates a new FA that matche any one of the FA expressions passed
/// </summary>
/// <param name="exprs">The expressions to match</param>
/// <param name="accept">The symbol to accept</param>
/// <returns>A new FA that will match the union of the FA expressions passed</returns>
public static FA Or(IEnumerable <FA> exprs, string accept = "")
{
var result = new FA();
var final = new FA();
final.AcceptingSymbol = accept;
foreach (var fa in exprs)
{
fa.EpsilonTransitions.Add(fa);
var nfa = fa.Clone();
var nffa = fa.FirstAcceptingState;
nfa.FirstAcceptingState.EpsilonTransitions.Add(final);
nffa.AcceptingSymbol = null;
}
return(result);
}
static void _RunLesson2()
{
// our regular expression engine does not have its own parser
// therefore we must create the expressions manually by using
// the appropriate construction methods.
// create a new lexer with the following five expressions:
// four self titled literals +, *, (, and )
// one regex [0-9]+ as "int"
// note that the symbols we use here match the terminals used in our
// CFG grammar from lesson 1. This is important.
_lexer = new FA();
_lexer.EpsilonTransitions.Add(FA.Literal("+", "+"));
_lexer.EpsilonTransitions.Add(FA.Literal("*", "*"));
_lexer.EpsilonTransitions.Add(FA.Literal("(", "("));
_lexer.EpsilonTransitions.Add(FA.Literal(")", ")"));
_lexer.EpsilonTransitions.Add(FA.Repeat(FA.Set("0123456789"), "int"));
Console.WriteLine("Lesson 2 - FA Lexer");
// there's no easy way to show the contents of this machine so we'll just show the total states
Console.WriteLine("NFA machine containes {0} total states", _lexer.FillClosure().Count);
Console.WriteLine();
}
/// <summary>
/// Creates a new FA that will match a repetition of zero or more of the specified FA expressions
/// </summary>
/// <param name="expr">The expression to repeat</param>
/// <param name="accept">The symbol to accept</param>
/// <returns>A new FA that matches the specified FA zero or more times</returns>
public static FA Kleene(FA expr, string accept = "")
{
return(Optional(Repeat(expr), accept));
}
public Tokenizer(FA lexer, IEnumerable <char> input)
{
_lexer = lexer;
_input = input;
}
/// <summary>
/// This is where the work happens
/// </summary>
/// <returns>The symbol that was matched. members _state _line,_column,_position,_buffer and _input are also modified.</returns>
string _Lex()
{
string acc;
var states = _initialStates;
_buffer.Clear();
switch (_state)
{
case -1: // initial
if (!_MoveNextInput())
{
_state = -2;
acc = _GetAcceptingSymbol(states);
if (null != acc)
{
return(acc);
}
else
{
return("#ERROR");
}
}
_state = 0; // running
break;
case -2: // end of stream
return("#EOS");
}
// Here's where we run most of the match. FillMove runs one interation of the NFA state machine.
// We match until we can't match anymore (greedy matching) and then report the symbol of the last
// match we found, or an error ("#ERROR") if we couldn't find one.
while (true)
{
var next = FA.FillMove(states, _input.Current);
if (0 == next.Count) // couldn't find any states
{
break;
}
_buffer.Append(_input.Current);
states = next;
if (!_MoveNextInput())
{
// end of stream
_state = -2;
acc = _GetAcceptingSymbol(states);
if (null != acc) // do we accept?
{
return(acc);
}
else
{
return("#ERROR");
}
}
}
acc = _GetAcceptingSymbol(states);
if (null != acc) // do we accept?
{
return(acc);
}
else
{
// handle the error condition
_buffer.Append(_input.Current);
if (!_MoveNextInput())
{
_state = -2;
}
return("#ERROR");
}
}
