ToRegExpKleene.h

Go to the documentation of this file.

 
/*
 * This file is part of Algorithms library toolkit.
 * Copyright (C) 2017 Jan Travnicek (jan.travnicek@fit.cvut.cz)
 
 * Algorithms library toolkit is free software: you can redistribute it and/or modify
 * it under the terms of the GNU General Public License as published by
 * the Free Software Foundation, either version 3 of the License, or
 * (at your option) any later version.
 
 * Algorithms library toolkit is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 * GNU General Public License for more details.
 
 * You should have received a copy of the GNU General Public License
 * along with Algorithms library toolkit.  If not, see <http://www.gnu.org/licenses/>.
 */
 
#pragma once
 
#include <regexp/unbounded/UnboundedRegExp.h>
 
#include <automaton/FSM/DFA.h>
#include <automaton/FSM/NFA.h>
#include <automaton/FSM/EpsilonNFA.h>
#include <automaton/FSM/MultiInitialStateNFA.h>
#include <automaton/FSM/MultiInitialStateEpsilonNFA.h>
 
#include <regexp/simplify/RegExpOptimize.h>
#include <regexp/transform/RegExpAlternate.h>
#include <regexp/transform/RegExpConcatenate.h>
#include <regexp/transform/RegExpIterate.h>
 
#include <common/createUnique.hpp>
 
#include <label/FinalStateLabel.h>
 
namespace automaton {
 
namespace convert {
 
class ToRegExpKleene {
public:
    template < class SymbolType, class StateType >
    static regexp::UnboundedRegExp < SymbolType > convert ( const automaton::DFA < SymbolType, StateType > & automaton );
 
    template < class SymbolType, class StateType >
    static regexp::UnboundedRegExp < SymbolType > convert ( const automaton::NFA < SymbolType, StateType > & automaton );
 
    template < class SymbolType, class StateType >
    static regexp::UnboundedRegExp < SymbolType > convert ( const automaton::MultiInitialStateNFA < SymbolType, StateType > & automaton );
 
    template < class SymbolType, class StateType >
    static regexp::UnboundedRegExp < SymbolType > convert ( const automaton::MultiInitialStateEpsilonNFA < SymbolType, StateType > & automaton );
 
    template < class SymbolType, class StateType >
    static regexp::UnboundedRegExp < SymbolType > convert ( const automaton::EpsilonNFA < SymbolType, StateType > & automaton );
 
private:
    template < class SymbolType, class StateType >
    static regexp::UnboundedRegExpStructure < SymbolType > transitionsToRegExp ( const automaton::MultiInitialStateEpsilonNFA < SymbolType, StateType > & automaton, const StateType & from, const StateType & to );
 
    template < class SymbolType, class StateType >
    static regexp::UnboundedRegExpStructure < SymbolType > RGetDefault ( const std::map < std::pair < StateType, StateType >, regexp::UnboundedRegExpStructure < SymbolType > > & R, const std::pair < StateType, StateType > & key );
};
 
template < class SymbolType, class StateType >
regexp::UnboundedRegExpStructure < SymbolType > ToRegExpKleene::RGetDefault ( const std::map < std::pair < StateType, StateType >, regexp::UnboundedRegExpStructure < SymbolType > > & R, const std::pair < StateType, StateType > & key ) {
    auto it = R.find ( key );
    return it != R.end ( ) ? it -> second : regexp::UnboundedRegExpStructure < SymbolType > ( regexp::UnboundedRegExpEmpty < SymbolType > ( ) );
}
 
template < class SymbolType, class StateType >
regexp::UnboundedRegExp < SymbolType > ToRegExpKleene::convert ( const automaton::DFA < SymbolType, StateType > & automaton ) {
    return convert ( automaton::MultiInitialStateEpsilonNFA < SymbolType, StateType > ( automaton ) );
}
 
template < class SymbolType, class StateType >
regexp::UnboundedRegExp < SymbolType > ToRegExpKleene::convert ( const automaton::NFA < SymbolType, StateType > & automaton ) {
    return convert ( automaton::MultiInitialStateEpsilonNFA < SymbolType, StateType > ( automaton ) );
}
 
template < class SymbolType, class StateType >
regexp::UnboundedRegExp < SymbolType > ToRegExpKleene::convert ( const automaton::EpsilonNFA < SymbolType, StateType > & automaton ) {
    return convert ( automaton::MultiInitialStateEpsilonNFA < SymbolType, StateType > ( automaton ) );
}
 
template < class SymbolType, class StateType >
regexp::UnboundedRegExp < SymbolType > ToRegExpKleene::convert ( const automaton::MultiInitialStateNFA < SymbolType, StateType > & automaton ) {
    return convert ( automaton::MultiInitialStateEpsilonNFA < SymbolType, StateType > ( automaton ) );
}
 
template < class SymbolType, class StateType >
regexp::UnboundedRegExp < SymbolType > ToRegExpKleene::convert ( const automaton::MultiInitialStateEpsilonNFA < SymbolType, StateType > & automaton ) {
    std::vector < std::map < std::pair < StateType, StateType >, regexp::UnboundedRegExpStructure < SymbolType > > > R ( 1 );
    size_t k = 0;
 
    // initialize R [ 0 ]
    for ( const StateType & a : automaton.getStates ( ) )
        for ( const StateType & b : automaton.getStates ( ) )
            R [ k ] [ std::make_pair ( a, b ) ] = transitionsToRegExp ( automaton, a, b );
    k += 1;
 
    // initialize R [ 1 ] ... R [ k ]
    for ( const StateType & kState : automaton.getStates ( ) ) {
        R.push_back ( std::map < std::pair < StateType, StateType >, regexp::UnboundedRegExpStructure < SymbolType > > ( ) );
 
        for ( const StateType & a : automaton.getStates ( ) ) {
            for ( const StateType & b : automaton.getStates ( ) ) {
                // TODO regexp::RegExpConcatenate on unbounded - variadic parameter count
 
                const regexp::UnboundedRegExpStructure < SymbolType > re = regexp::transform::RegExpConcatenate::concatenate (
                            RGetDefault ( R [ k - 1 ], std::make_pair ( a, kState ) ),
                            regexp::transform::RegExpConcatenate::concatenate (
                                regexp::transform::RegExpIterate::iterate ( RGetDefault ( R [ k - 1 ], std::make_pair ( kState, kState ) ) ),
                                RGetDefault ( R [ k - 1 ], std::make_pair ( kState, b ) ) ) );
 
                R [ k ] [ std::make_pair ( a, b ) ] = regexp::simplify::RegExpOptimize::optimize ( regexp::transform::RegExpAlternate::alternate ( RGetDefault ( R [ k - 1 ], std::make_pair ( a, b ) ), re ) );
            }
        }
 
        k += 1;
    }
 
    regexp::UnboundedRegExpStructure < SymbolType > ret ( regexp::UnboundedRegExpEmpty < SymbolType > { } );
    for ( const auto & i : automaton.getInitialStates ( ) ) {
        for ( const auto & f : automaton.getFinalStates ( ) ) {
            ret = regexp::transform::RegExpAlternate::alternate ( ret, RGetDefault ( R [ k - 1 ], std::make_pair ( i, f ) ) );
        }
    }
 
    return regexp::UnboundedRegExp < SymbolType > ( ret );
}
 
 
template < class SymbolType, class StateType >
regexp::UnboundedRegExpStructure < SymbolType > ToRegExpKleene::transitionsToRegExp ( const automaton::MultiInitialStateEpsilonNFA < SymbolType, StateType > & automaton, const StateType & from, const StateType & to ) {
    regexp::UnboundedRegExpStructure < SymbolType > ret ( regexp::UnboundedRegExpEmpty < SymbolType > { } );
 
    if ( from == to )
        ret.setStructure ( regexp::UnboundedRegExpEpsilon < SymbolType > ( ) );
 
    for ( const auto & transition: automaton.getTransitionsFromState ( from ) ) {
        if ( transition.second == to ) {
            if ( transition.first.second.is_epsilon ( ) )
                ret = regexp::transform::RegExpAlternate::alternate ( ret, regexp::UnboundedRegExpStructure < SymbolType > ( regexp::UnboundedRegExpEpsilon < SymbolType > ( ) ) );
            else
                ret = regexp::transform::RegExpAlternate::alternate ( ret, regexp::UnboundedRegExpStructure < SymbolType > ( regexp::UnboundedRegExpSymbol < SymbolType > ( transition.first.second.getSymbol ( ) ) ) );
        }
    }
 
    return regexp::simplify::RegExpOptimize::optimize ( ret );
}
 
} /* namespace convert */
 
} /* namespace automaton */