/// <summary>
/// Returns a new (deterministic) automaton that accepts a single char whose value is in the
/// given interval (including both end points).
/// </summary>
/// <param name="min">The min.</param>
/// <param name="max">The max.</param>
/// <returns>
/// A new (deterministic) automaton that accepts a single char whose value is in the
/// given interval (including both end points).
/// </returns>
public static Automaton MakeCharRange(char min, char max)
{
if (min == max)
{
return(BasicAutomata.MakeChar(min));
}
var a = new Automaton();
var s1 = new State();
var s2 = new State();
a.Initial = s1;
s2.Accept = true;
if (min <= max)
{
s1.Transitions.Add(new Transition(min, max, s2));
}
a.IsDeterministic = true;
return(a);
}
public static Automaton Concatenate(Automaton a1, Automaton a2)
{
if (a1.IsSingleton && a2.IsSingleton)
{
return(BasicAutomata.MakeString(a1.Singleton + a2.Singleton));
}
if (BasicOperations.IsEmpty(a1) || BasicOperations.IsEmpty(a2))
{
return(BasicAutomata.MakeEmpty());
}
bool deterministic = a1.IsSingleton && a2.IsDeterministic;
if (a1 == a2)
{
a1 = a1.CloneExpanded();
a2 = a2.CloneExpanded();
}
else
{
a1 = a1.CloneExpandedIfRequired();
a2 = a2.CloneExpandedIfRequired();
}
foreach (State s in a1.GetAcceptStates())
{
s.Accept = false;
s.AddEpsilon(a2.Initial);
}
a1.IsDeterministic = deterministic;
a1.ClearHashCode();
a1.CheckMinimizeAlways();
return(a1);
}
private Automaton ToAutomaton(
IDictionary <string, Automaton> automata,
IAutomatonProvider automatonProvider,
bool minimize)
{
IList <Automaton> list;
Automaton a = null;
switch (kind)
{
case Kind.RegexpUnion:
list = new List <Automaton>();
this.FindLeaves(exp1, Kind.RegexpUnion, list, automata, automatonProvider, minimize);
this.FindLeaves(exp2, Kind.RegexpUnion, list, automata, automatonProvider, minimize);
a = BasicOperations.Union(list);
a.Minimize();
break;
case Kind.RegexpConcatenation:
list = new List <Automaton>();
this.FindLeaves(exp1, Kind.RegexpConcatenation, list, automata, automatonProvider, minimize);
this.FindLeaves(exp2, Kind.RegexpConcatenation, list, automata, automatonProvider, minimize);
a = BasicOperations.Concatenate(list);
a.Minimize();
break;
case Kind.RegexpIntersection:
a = exp1.ToAutomaton(automata, automatonProvider, minimize)
.Intersection(exp2.ToAutomaton(automata, automatonProvider, minimize));
a.Minimize();
break;
case Kind.RegexpOptional:
a = exp1.ToAutomaton(automata, automatonProvider, minimize).Optional();
a.Minimize();
break;
case Kind.RegexpRepeat:
a = exp1.ToAutomaton(automata, automatonProvider, minimize).Repeat();
a.Minimize();
break;
case Kind.RegexpRepeatMin:
a = exp1.ToAutomaton(automata, automatonProvider, minimize).Repeat(min);
a.Minimize();
break;
case Kind.RegexpRepeatMinMax:
a = exp1.ToAutomaton(automata, automatonProvider, minimize).Repeat(min, max);
a.Minimize();
break;
case Kind.RegexpComplement:
a = exp1.ToAutomaton(automata, automatonProvider, minimize).Complement();
a.Minimize();
break;
case Kind.RegexpChar:
a = BasicAutomata.MakeChar(c);
break;
case Kind.RegexpCharRange:
a = BasicAutomata.MakeCharRange(from, to);
break;
case Kind.RegexpAnyChar:
{
a = this.anyCharAlphabet != null?BasicAutomata.MakeCharSet(this.anyCharAlphabet) : BasicAutomata.MakeAnyChar();
break;
}
case Kind.RegexpEmpty:
a = BasicAutomata.MakeEmpty();
break;
case Kind.RegexpString:
a = BasicAutomata.MakeString(s);
break;
case Kind.RegexpAnyString:
a = BasicAutomata.MakeAnyString();
break;
case Kind.RegexpAutomaton:
Automaton aa = null;
if (automata != null)
{
automata.TryGetValue(s, out aa);
}
if (aa == null && automatonProvider != null)
{
try
{
aa = automatonProvider.GetAutomaton(s);
}
catch (IOException e)
{
throw new ArgumentException(string.Empty, e);
}
}
if (aa == null)
{
throw new ArgumentException("'" + s + "' not found");
}
a = aa.Clone(); // Always clone here (ignore allowMutate).
break;
case Kind.RegexpInterval:
a = BasicAutomata.MakeInterval(min, max, digits);
break;
}
return(a);
}
public static Automaton MakeString(string s)
{
return(BasicAutomata.MakeString(s));
}
public static Automaton MakeCharSet(string set)
{
return(BasicAutomata.MakeCharSet(set));
}
public static Automaton MakeChar(char c)
{
return(BasicAutomata.MakeChar(c));
}
/// <summary>
/// Returns an automaton that accepts the intersection of the languages of the given automata.
/// Never modifies the input automata languages.
/// </summary>
/// <param name="a1">The a1.</param>
/// <param name="a2">The a2.</param>
/// <returns></returns>
public static Automaton Intersection(Automaton a1, Automaton a2)
{
if (a1.IsSingleton)
{
if (a2.Run(a1.Singleton))
{
return(a1.CloneIfRequired());
}
return(BasicAutomata.MakeEmpty());
}
if (a2.IsSingleton)
{
if (a1.Run(a2.Singleton))
{
return(a2.CloneIfRequired());
}
return(BasicAutomata.MakeEmpty());
}
if (a1 == a2)
{
return(a1.CloneIfRequired());
}
Transition[][] transitions1 = Automaton.GetSortedTransitions(a1.GetStates());
Transition[][] transitions2 = Automaton.GetSortedTransitions(a2.GetStates());
var c = new Automaton();
var worklist = new LinkedList <StatePair>();
var newstates = new Dictionary <StatePair, StatePair>();
var p = new StatePair(c.Initial, a1.Initial, a2.Initial);
worklist.AddLast(p);
newstates.Add(p, p);
while (worklist.Count > 0)
{
p = worklist.RemoveAndReturnFirst();
p.S.Accept = p.FirstState.Accept && p.SecondState.Accept;
Transition[] t1 = transitions1[p.FirstState.Number];
Transition[] t2 = transitions2[p.SecondState.Number];
for (int n1 = 0, b2 = 0; n1 < t1.Length; n1++)
{
while (b2 < t2.Length && t2[b2].Max < t1[n1].Min)
{
b2++;
}
for (int n2 = b2; n2 < t2.Length && t1[n1].Max >= t2[n2].Min; n2++)
{
if (t2[n2].Max >= t1[n1].Min)
{
var q = new StatePair(t1[n1].To, t2[n2].To);
StatePair r;
newstates.TryGetValue(q, out r);
if (r == null)
{
q.S = new State();
worklist.AddLast(q);
newstates.Add(q, q);
r = q;
}
char min = t1[n1].Min > t2[n2].Min ? t1[n1].Min : t2[n2].Min;
char max = t1[n1].Max < t2[n2].Max ? t1[n1].Max : t2[n2].Max;
p.S.Transitions.Add(new Transition(min, max, r.S));
}
}
}
}
c.IsDeterministic = a1.IsDeterministic && a2.IsDeterministic;
c.RemoveDeadTransitions();
c.CheckMinimizeAlways();
return(c);
}
public static Automaton Concatenate(IList <Automaton> l)
{
if (l.Count == 0)
{
return(BasicAutomata.MakeEmptyString());
}
bool allSingleton = l.All(a => a.IsSingleton);
if (allSingleton)
{
var b = new StringBuilder();
foreach (Automaton a in l)
{
b.Append(a.Singleton);
}
return(BasicAutomata.MakeString(b.ToString()));
}
else
{
if (l.Any(a => a.IsEmpty))
{
return(BasicAutomata.MakeEmpty());
}
var ids = new HashSet <int>();
foreach (Automaton a in l)
{
ids.Add(RuntimeHelpers.GetHashCode(a));
}
bool hasAliases = ids.Count != l.Count;
Automaton b = l[0];
b = hasAliases ? b.CloneExpanded() : b.CloneExpandedIfRequired();
var ac = b.GetAcceptStates();
bool first = true;
foreach (Automaton a in l)
{
if (first)
{
first = false;
}
else
{
if (a.IsEmptyString())
{
continue;
}
Automaton aa = a;
aa = hasAliases ? aa.CloneExpanded() : aa.CloneExpandedIfRequired();
HashSet <State> ns = aa.GetAcceptStates();
foreach (State s in ac)
{
s.Accept = false;
s.AddEpsilon(aa.Initial);
if (s.Accept)
{
ns.Add(s);
}
}
ac = ns;
}
}
b.IsDeterministic = false;
b.ClearHashCode();
b.CheckMinimizeAlways();
return(b);
}
}
/// <summary>
/// Returns a new (deterministic) automaton that accepts any single character.
/// </summary>
/// <returns>A new (deterministic) automaton that accepts any single character.</returns>
public static Automaton MakeAnyChar()
{
return(BasicAutomata.MakeCharRange(char.MinValue, char.MaxValue));
}
/// <summary>
/// Constructs sub-automaton corresponding to decimal numbers of value between x.Substring(n)
/// and y.Substring(n) and of length x.Substring(n).Length (which must be equal to
/// y.Substring(n).Length).
/// </summary>
/// <param name="x">The x.</param>
/// <param name="y">The y.</param>
/// <param name="n">The n.</param>
/// <param name="initials">The initials.</param>
/// <param name="zeros">if set to <c>true</c> [zeros].</param>
/// <returns></returns>
private static State Between(string x, string y, int n, ICollection <State> initials, bool zeros)
{
var s = new State();
if (x.Length == n)
{
s.Accept = true;
}
else
{
if (zeros)
{
initials.Add(s);
}
char cx = x[n];
char cy = y[n];
if (cx == cy)
{
s.AddTransition(new Transition(cx, Between(x, y, n + 1, initials, zeros && cx == '0')));
}
else
{
// cx < cy
s.AddTransition(new Transition(cx, BasicAutomata.AtLeast(x, n + 1, initials, zeros && cx == '0')));
s.AddTransition(new Transition(cy, BasicAutomata.AtMost(y, n + 1)));
if (cx + 1 < cy)
{
s.AddTransition(new Transition((char)(cx + 1), (char)(cy - 1), BasicAutomata.AnyOfRightLength(x, n + 1)));
}
}
}
return(s);
}
private Automaton ToAutomaton(
IDictionary <string, Automaton> automata,
IAutomatonProvider automatonProvider,
bool minimize)
{
IList <Automaton> list;
Automaton a = null;
switch (this.Kind)
{
case Kind.RegexpUnion:
list = new List <Automaton>();
this.FindLeaves(Expr1, Kind.RegexpUnion, list, automata, automatonProvider, minimize);
this.FindLeaves(Expr2, Kind.RegexpUnion, list, automata, automatonProvider, minimize);
a = BasicOperations.Union(list);
a.Minimize();
break;
case Kind.RegexpConcatenation:
list = new List <Automaton>();
this.FindLeaves(Expr1, Kind.RegexpConcatenation, list, automata, automatonProvider, minimize);
this.FindLeaves(Expr2, Kind.RegexpConcatenation, list, automata, automatonProvider, minimize);
a = BasicOperations.Concatenate(list);
a.Minimize();
break;
case Kind.RegexpIntersection:
a = Expr1.ToAutomaton(automata, automatonProvider, minimize)
.Intersection(Expr2.ToAutomaton(automata, automatonProvider, minimize));
a.Minimize();
break;
case Kind.RegexpOptional:
a = Expr1.ToAutomaton(automata, automatonProvider, minimize).Optional();
a.Minimize();
break;
case Kind.RegexpRepeat:
a = Expr1.ToAutomaton(automata, automatonProvider, minimize).Repeat();
a.Minimize();
break;
case Kind.RegexpRepeatMin:
a = Expr1.ToAutomaton(automata, automatonProvider, minimize).Repeat(Min);
a.Minimize();
break;
case Kind.RegexpRepeatMinMax:
a = Expr1.ToAutomaton(automata, automatonProvider, minimize).Repeat(Min, Max);
a.Minimize();
break;
case Kind.RegexpComplement:
a = Expr1.ToAutomaton(automata, automatonProvider, minimize).Complement();
a.Minimize();
break;
case Kind.RegexpChar:
a = BasicAutomata.MakeChar(this.Char);
break;
case Kind.RegexpCharRange:
a = BasicAutomata.MakeCharRange(FromChar, ToChar);
break;
case Kind.RegexpAnyChar:
a = BasicAutomata.MakeAnyChar();
break;
case Kind.RegexpEmpty:
a = BasicAutomata.MakeEmpty();
break;
case Kind.RegexpString:
a = BasicAutomata.MakeString(SourceRegExpr);
break;
case Kind.RegexpAnyString:
a = BasicAutomata.MakeAnyString();
break;
case Kind.RegexpAutomaton:
Automaton aa = null;
if (automata != null)
{
automata.TryGetValue(SourceRegExpr, out aa);
}
if (aa == null && automatonProvider != null)
{
try
{
aa = automatonProvider.GetAutomaton(SourceRegExpr);
}
catch (IOException e)
{
throw new ArgumentException(string.Empty, e);
}
}
if (aa == null)
{
throw new ArgumentException("'" + SourceRegExpr + "' not found");
}
a = aa.Clone(); // Always clone here (ignore allowMutate).
break;
case Kind.RegexpInterval:
a = BasicAutomata.MakeInterval(Min, Max, Digits);
break;
}
return(a);
}
internal static Automaton MakeString(string s)
{
return(BasicAutomata.MakeString(s));
}
internal static Automaton MakeChar(char c)
{
return(BasicAutomata.MakeChar(c));
}