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()
{
State startState = dfaStates[content.InitialStates.DFA];
dfa.DFA dfa = new dfa.DFA(dfaStates, startState);
return(new StringTokenizer(dfa));
}
/// <summary>
/// Goto the next state depending on an input character.
/// </summary>
/// <param name="ch">The character that determines what state to go to next.</param>
/// <returns>The new current state.</returns>
public State GotoNext(char ch)
{
Transition transition = currentState.Transitions.Find(ch);
if (transition != null)
{
currentState = transition.Target;
return currentState;
}
else
return null;
}
/// <summary>
/// Creates a new transition by specifying the target state and the criteria for
/// taking a transition to another state. The source state does not need to be
/// specified, because the state itself knows its transition.
/// </summary>
/// <param name="target">The target state.</param>
/// <param name="characters">The character set criteria.</param>
public Transition(State target, string characters)
{
this.target = target;
if (characters.Length > 10)
this.charset = new HashSet();
else
this.charset = new ArraySet();
char[] ca = characters.ToCharArray();
foreach (Char ch in ca)
{
this.charset.Add(ch);
}
}
public void Add(State state)
{
list.Add(state);
}
/// <summary>
/// Sets the DFA back to the starting state, so it can be used to get a new token.
/// </summary>
public void Reset()
{
this.currentState = startState;
}
/// <summary>
/// Creates a new DFA.
/// </summary>
/// <param name="states">The states that are part of the DFA.</param>
/// <param name="startState">The starting state</param>
public DFA(StateCollection states, State startState)
{
this.states = states;
this.startState = startState;
this.currentState = startState;
}
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;
}