Operations for minimizing automata. @lucene.experimental
Esempio n. 1
0
 internal virtual void CheckMinimizeAlways()
 {
     if (minimize_always)
     {
         MinimizationOperations.Minimize(this);
     }
 }
Esempio n. 2
0
        /// <summary>
        /// Return an automaton that accepts all transpositions of s
        /// (transposing two adjacent characters)
        /// </summary>
        private Automaton TranspositionsOf(string s)
        {
            if (s.Length < 2)
            {
                return(BasicAutomata.MakeEmpty());
            }
            IList <Automaton> list = new List <Automaton>();

            for (int i = 0; i < s.Length - 1; i++)
            {
                StringBuilder sb = new StringBuilder();
                sb.Append(s.Substring(0, i));
                sb.Append(s[i + 1]);
                sb.Append(s[i]);
                sb.Append(s.Substring(i + 2, s.Length - (i + 2)));
                string st = sb.ToString();
                if (!st.Equals(s))
                {
                    list.Add(BasicAutomata.MakeString(st));
                }
            }
            Automaton a = BasicOperations.Union(list);

            MinimizationOperations.Minimize(a);
            return(a);
        }
Esempio n. 3
0
        /// <summary>
        /// Return an automaton that accepts all 1-character insertions, deletions,
        /// substitutions, and transpositions of s.
        /// </summary>
        private Automaton NaiveLev1T(string s)
        {
            Automaton a = NaiveLev1(s);

            a = BasicOperations.Union(a, TranspositionsOf(s));
            MinimizationOperations.Minimize(a);
            return(a);
        }
Esempio n. 4
0
        public virtual void Test()
        {
            int num = AtLeast(200);

            for (int i = 0; i < num; i++)
            {
                Automaton a = AutomatonTestUtil.RandomAutomaton(Random);
                Automaton b = (Automaton)a.Clone();
                MinimizationOperations.Minimize(b);
                Assert.IsTrue(BasicOperations.SameLanguage(a, b));
            }
        }
Esempio n. 5
0
        /// <summary>
        /// Return an automaton that accepts all 1-character insertions, deletions, and
        /// substitutions of s.
        /// </summary>
        private Automaton NaiveLev1(string s)
        {
            Automaton a = BasicAutomata.MakeString(s);

            a = BasicOperations.Union(a, InsertionsOf(s));
            MinimizationOperations.Minimize(a);
            a = BasicOperations.Union(a, DeletionsOf(s));
            MinimizationOperations.Minimize(a);
            a = BasicOperations.Union(a, SubstitutionsOf(s));
            MinimizationOperations.Minimize(a);

            return(a);
        }
Esempio n. 6
0
        public virtual void TestAgainstBrzozowski()
        {
            int num = AtLeast(200);

            for (int i = 0; i < num; i++)
            {
                Automaton a = AutomatonTestUtil.RandomAutomaton(Random);
                AutomatonTestUtil.MinimizeSimple(a);
                Automaton b = (Automaton)a.Clone();
                MinimizationOperations.Minimize(b);
                Assert.IsTrue(BasicOperations.SameLanguage(a, b));
                Assert.AreEqual(a.GetNumberOfStates(), b.GetNumberOfStates());
                Assert.AreEqual(a.GetNumberOfTransitions(), b.GetNumberOfTransitions());
            }
        }
Esempio n. 7
0
        public virtual void TestFiniteStrings()
        {
            Automaton a = BasicOperations.Union(BasicAutomata.MakeString("dog"), BasicAutomata.MakeString("duck"));

            MinimizationOperations.Minimize(a);
            ISet <Int32sRef> strings = SpecialOperations.GetFiniteStrings(a, -1);

            Assert.AreEqual(2, strings.Count);
            Int32sRef dog = new Int32sRef();

            Util.ToInt32sRef(new BytesRef("dog"), dog);
            Assert.IsTrue(strings.Contains(dog));
            Int32sRef duck = new Int32sRef();

            Util.ToInt32sRef(new BytesRef("duck"), duck);
            Assert.IsTrue(strings.Contains(duck));
        }
Esempio n. 8
0
        /// <summary>
        /// Return an automaton that accepts all 1-character substitutions of s
        /// (replacing one character)
        /// </summary>
        private Automaton SubstitutionsOf(string s)
        {
            IList <Automaton> list = new List <Automaton>();

            for (int i = 0; i < s.Length; i++)
            {
                Automaton au = BasicAutomata.MakeString(s.Substring(0, i));
                au = BasicOperations.Concatenate(au, BasicAutomata.MakeAnyChar());
                au = BasicOperations.Concatenate(au, BasicAutomata.MakeString(s.Substring(i + 1)));
                list.Add(au);
            }

            Automaton a = BasicOperations.Union(list);

            MinimizationOperations.Minimize(a);
            return(a);
        }
Esempio n. 9
0
        private Automaton ToAutomaton(IDictionary <string, Automaton> automata, IAutomatonProvider automaton_provider)
        {
            IList <Automaton> list;
            Automaton         a = null;

            switch (kind)
            {
            case Kind.REGEXP_UNION:
                list = new List <Automaton>();
                FindLeaves(exp1, Kind.REGEXP_UNION, list, automata, automaton_provider);
                FindLeaves(exp2, Kind.REGEXP_UNION, list, automata, automaton_provider);
                a = BasicOperations.Union(list);
                MinimizationOperations.Minimize(a);
                break;

            case Kind.REGEXP_CONCATENATION:
                list = new List <Automaton>();
                FindLeaves(exp1, Kind.REGEXP_CONCATENATION, list, automata, automaton_provider);
                FindLeaves(exp2, Kind.REGEXP_CONCATENATION, list, automata, automaton_provider);
                a = BasicOperations.Concatenate(list);
                MinimizationOperations.Minimize(a);
                break;

            case Kind.REGEXP_INTERSECTION:
                a = exp1.ToAutomaton(automata, automaton_provider).Intersection(exp2.ToAutomaton(automata, automaton_provider));
                MinimizationOperations.Minimize(a);
                break;

            case Kind.REGEXP_OPTIONAL:
                a = exp1.ToAutomaton(automata, automaton_provider).Optional();
                MinimizationOperations.Minimize(a);
                break;

            case Kind.REGEXP_REPEAT:
                a = exp1.ToAutomaton(automata, automaton_provider).Repeat();
                MinimizationOperations.Minimize(a);
                break;

            case Kind.REGEXP_REPEAT_MIN:
                a = exp1.ToAutomaton(automata, automaton_provider).Repeat(min);
                MinimizationOperations.Minimize(a);
                break;

            case Kind.REGEXP_REPEAT_MINMAX:
                a = exp1.ToAutomaton(automata, automaton_provider).Repeat(min, max);
                MinimizationOperations.Minimize(a);
                break;

            case Kind.REGEXP_COMPLEMENT:
                a = exp1.ToAutomaton(automata, automaton_provider).Complement();
                MinimizationOperations.Minimize(a);
                break;

            case Kind.REGEXP_CHAR:
                a = BasicAutomata.MakeChar(c);
                break;

            case Kind.REGEXP_CHAR_RANGE:
                a = BasicAutomata.MakeCharRange(from, to);
                break;

            case Kind.REGEXP_ANYCHAR:
                a = BasicAutomata.MakeAnyChar();
                break;

            case Kind.REGEXP_EMPTY:
                a = BasicAutomata.MakeEmpty();
                break;

            case Kind.REGEXP_STRING:
                a = BasicAutomata.MakeString(s);
                break;

            case Kind.REGEXP_ANYSTRING:
                a = BasicAutomata.MakeAnyString();
                break;

            case Kind.REGEXP_AUTOMATON:
                Automaton aa = null;
                if (automata != null)
                {
                    aa = automata[s];
                }
                if (aa == null && automaton_provider != null)
                {
                    try
                    {
                        aa = automaton_provider.GetAutomaton(s);
                    }
                    catch (Exception e) when(e.IsIOException())
                    {
                        throw new ArgumentException(e.ToString(), e);
                    }
                }
                if (aa == null)
                {
                    throw new ArgumentException("'" + s + "' not found");
                }
                a = (Automaton)aa.Clone();     // always clone here (ignore allow_mutate)
                break;

            case Kind.REGEXP_INTERVAL:
                a = BasicAutomata.MakeInterval(min, max, digits);
                break;
            }
            return(a);
        }
Esempio n. 10
0
 /// <summary>
 /// See <see cref="MinimizationOperations.Minimize(Automaton)"/>. Returns the
 /// automaton being given as argument.
 /// </summary>
 public static Automaton Minimize(Automaton a)
 {
     MinimizationOperations.Minimize(a);
     return(a);
 }