public void RegexToNFA_ConvertsValidRegexCorrectly(string regex, char[] alphabet, string[] validWords, string[] invalidWords, int expectedStates)
{
var automaton = AutomatonUtil.RegexToAutomaton(regex, alphabet);
Assert.Equal(expectedStates, automaton.States.Count);
foreach (var validWord in validWords)
{
Assert.True(automaton.IsValidWord(validWord));
}
foreach (var invalidWord in invalidWords)
{
Assert.False(automaton.IsValidWord(invalidWord));
}
}
internal static Automaton RegularLanguageRegexToCanonicalWordAcceptingDfa(string regex, string regexName = "")
{
// It isn't required to convert to DFA, but as it is an inexpensive operation for the automaton created via the thompson construction,
// it probably will end up saving time overall by reducing the time on the other steps.
var sw = Stopwatch.StartNew();
var automaton = TimedFunction(() => AutomatonUtil.RegexToAutomaton(regex, Constants.RegularLanguage.Symbols),
$"Regex to Thompson NFA for regex {regexName}");
automaton = TimedFunction(() => automaton.ToDFA(),
$"Thompson NFA to DFA for regex {regexName}");
TimedFunction(() => automaton.UpdateAutomatonToAcceptCanonicalWords(),
$"Canonicalisation of DFA for regex {regexName}");
automaton = TimedFunction(() => automaton.ToDFA(),
$"Canonicalised NFA to DFA for regex {regexName}");
return(automaton);
}
public void RegexToPostFix_ConvertsCorrectly(string infixRegex, string expectedPostfix)
{
Assert.Equal(expectedPostfix, String.Join("", AutomatonUtil.RegexToPostfix(infixRegex)));
}
