private DFA.StateCollection CreateDFAStates(CGTContent content)
{
symbols = CreateSymbols(content);
DFA.StateCollection states = new DFA.StateCollection();
foreach (DFAStateRecord stateRecord in content.DFAStateTable)
{
DFA.State state;
if (stateRecord.AcceptState)
{
Symbol symbol = symbols[stateRecord.AcceptIndex];
state = new DFA.EndState(stateRecord.Index, (SymbolTerminal)symbol);
//todo: type checking (exception?)
}
else
{
state = new DFA.State(stateRecord.Index);
}
states.Add(state);
}
foreach (DFAStateRecord stateRecord in content.DFAStateTable)
{
foreach (EdgeSubRecord edgeRecord in stateRecord.EdgeSubRecords)
{
DFA.State source = states[stateRecord.Index];
DFA.State target = states[edgeRecord.TargetIndex];
CharacterSetRecord charsetRec = content.CharacterSetTable[edgeRecord.CharacterSetIndex];
DFA.Transition transition = new DFA.Transition(target, charsetRec.Characters);
source.Transitions.Add(transition);
}
}
return(states);
}
/// <summary>
/// Creates a new tokenizer. Useful if for some reason
/// you don't want a full LALR parser, but are just interested in a tokenizer.
/// </summary>
/// <returns></returns>
public StringTokenizer CreateNewTokenizer()
{
DFA.State startState = dfaStates[content.InitialStates.DFA];
DFA.DFA dfa = new DFA.DFA(dfaStates, startState);
return(new StringTokenizer(dfa));
}