/* -*- mode: C++; indent-tabs-mode: nil; -*-

 *

 * This file is a part of LEMON, a generic C++ optimization library.

 *

 * Copyright (C) 2003-2009

 * Egervary Jeno Kombinatorikus Optimalizalasi Kutatocsoport

 * (Egervary Research Group on Combinatorial Optimization, EGRES).

 *

 * Permission to use, modify and distribute this software is granted

 * provided that this copyright notice appears in all copies. For

 * precise terms see the accompanying LICENSE file.

 *

 * This software is provided "AS IS" with no warranty of any kind,

 * express or implied, and with no claim as to its suitability for any

 * purpose.

 *

 */

#ifndef LEMON_SUURBALLE_H

#define LEMON_SUURBALLE_H

///\ingroup shortest_path

///\file

///\brief An algorithm for finding arc-disjoint paths between two

/// nodes having minimum total length.

#include <vector>

#include <limits>

#include <lemon/bin_heap.h>

#include <lemon/path.h>

#include <lemon/list_graph.h>

#include <lemon/dijkstra.h>

#include <lemon/maps.h>

namespace lemon {

  /// \addtogroup shortest_path

  /// @{

  /// \brief Algorithm for finding arc-disjoint paths between two nodes

  /// having minimum total length.

  ///

  /// \ref lemon::Suurballe "Suurballe" implements an algorithm for

  /// finding arc-disjoint paths having minimum total length (cost)

  /// from a given source node to a given target node in a digraph.

  ///

  /// Note that this problem is a special case of the \ref min_cost_flow

  /// "minimum cost flow problem". This implementation is actually an

  /// efficient specialized version of the \ref CapacityScaling

  /// "successive shortest path" algorithm directly for this problem.

  /// Therefore this class provides query functions for flow values and

  /// node potentials (the dual solution) just like the minimum cost flow

  /// algorithms.

  ///

  /// \tparam GR The digraph type the algorithm runs on.

  /// \tparam LEN The type of the length map.

  /// The default value is <tt>GR::ArcMap<int></tt>.

  ///

  /// \warning Length values should be \e non-negative.

  ///

  /// \note For finding \e node-disjoint paths, this algorithm can be used

  /// along with the \ref SplitNodes adaptor.

#ifdef DOXYGEN

  template <typename GR, typename LEN>

#else

  template < typename GR,

             typename LEN = typename GR::template ArcMap<int> >

#endif

  class Suurballe

  {

    TEMPLATE_DIGRAPH_TYPEDEFS(GR);

    typedef ConstMap<Arc, int> ConstArcMap;

    typedef typename GR::template NodeMap<Arc> PredMap;

  public:

    /// The type of the digraph the algorithm runs on.

    typedef GR Digraph;

    /// The type of the length map.

    typedef LEN LengthMap;

    /// The type of the lengths.

    typedef typename LengthMap::Value Length;

#ifdef DOXYGEN

    /// The type of the flow map.

    typedef GR::ArcMap<int> FlowMap;

    /// The type of the potential map.

    typedef GR::NodeMap<Length> PotentialMap;

#else

    /// The type of the flow map.

    typedef typename Digraph::template ArcMap<int> FlowMap;

    /// The type of the potential map.

    typedef typename Digraph::template NodeMap<Length> PotentialMap;

#endif

    /// The type of the path structures.

    typedef SimplePath<GR> Path;

  private:

    typedef typename Digraph::template NodeMap<int> HeapCrossRef;

    typedef BinHeap<Length, HeapCrossRef> Heap;

    // ResidualDijkstra is a special implementation of the

    // Dijkstra algorithm for finding shortest paths in the

    // residual network with respect to the reduced arc lengths

    // and modifying the node potentials according to the

    // distance of the nodes.

    class ResidualDijkstra

    {

    private:

      const Digraph &_graph;

      const LengthMap &_length;

      const FlowMap &_flow;

      PotentialMap &_pi;

      PredMap &_pred;

      Node _s;

      Node _t;

      PotentialMap _dist;

      std::vector<Node> _proc_nodes;

    public:

      // Constructor

      ResidualDijkstra(Suurballe &srb) :

        _graph(srb._graph), _length(srb._length),

        _flow(*srb._flow), _pi(*srb._potential), _pred(srb._pred), 

        _s(srb._s), _t(srb._t), _dist(_graph) {}

      // Run the algorithm and return true if a path is found

      // from the source node to the target node.

      bool run(int cnt) {

        return cnt == 0 ? startFirst() : start();

      }

    private:

      // Execute the algorithm for the first time (the flow and potential

      // functions have to be identically zero).

      bool startFirst() {

        HeapCrossRef heap_cross_ref(_graph, Heap::PRE_HEAP);

        Heap heap(heap_cross_ref);

        heap.push(_s, 0);

        _pred[_s] = INVALID;

        _proc_nodes.clear();

        // Process nodes

        while (!heap.empty() && heap.top() != _t) {

          Node u = heap.top(), v;

          Length d = heap.prio(), dn;

          _dist[u] = heap.prio();

          _proc_nodes.push_back(u);

          heap.pop();

          // Traverse outgoing arcs

          for (OutArcIt e(_graph, u); e != INVALID; ++e) {

            v = _graph.target(e);

            switch(heap.state(v)) {

              case Heap::PRE_HEAP:

                heap.push(v, d + _length[e]);

                _pred[v] = e;

                break;

              case Heap::IN_HEAP:

                dn = d + _length[e];

                if (dn < heap[v]) {

                  heap.decrease(v, dn);

                  _pred[v] = e;

                }

                break;

              case Heap::POST_HEAP:

                break;

            }

          }

        }

        if (heap.empty()) return false;

        // Update potentials of processed nodes

        Length t_dist = heap.prio();

        for (int i = 0; i < int(_proc_nodes.size()); ++i)

          _pi[_proc_nodes[i]] = _dist[_proc_nodes[i]] - t_dist;

        return true;

      }

      // Execute the algorithm.

      bool start() {

        HeapCrossRef heap_cross_ref(_graph, Heap::PRE_HEAP);

        Heap heap(heap_cross_ref);

        heap.push(_s, 0);

        _pred[_s] = INVALID;

        _proc_nodes.clear();

        // Process nodes

        while (!heap.empty() && heap.top() != _t) {

          Node u = heap.top(), v;

          Length d = heap.prio() + _pi[u], dn;

          _dist[u] = heap.prio();

          _proc_nodes.push_back(u);

          heap.pop();

          // Traverse outgoing arcs

          for (OutArcIt e(_graph, u); e != INVALID; ++e) {

            if (_flow[e] == 0) {

              v = _graph.target(e);

              switch(heap.state(v)) {

                case Heap::PRE_HEAP:

                  heap.push(v, d + _length[e] - _pi[v]);

                  _pred[v] = e;

                  break;

                case Heap::IN_HEAP:

                  dn = d + _length[e] - _pi[v];

                  if (dn < heap[v]) {

                    heap.decrease(v, dn);

                    _pred[v] = e;

                  }

                  break;

                case Heap::POST_HEAP:

                  break;

              }

            }

          }

          // Traverse incoming arcs

          for (InArcIt e(_graph, u); e != INVALID; ++e) {

            if (_flow[e] == 1) {

              v = _graph.source(e);

              switch(heap.state(v)) {

                case Heap::PRE_HEAP:

                  heap.push(v, d - _length[e] - _pi[v]);

                  _pred[v] = e;

                  break;

                case Heap::IN_HEAP:

                  dn = d - _length[e] - _pi[v];

                  if (dn < heap[v]) {

                    heap.decrease(v, dn);

                    _pred[v] = e;

                  }

                  break;

                case Heap::POST_HEAP:

                  break;

              }

            }

          }

        }

        if (heap.empty()) return false;

        // Update potentials of processed nodes

        Length t_dist = heap.prio();

        for (int i = 0; i < int(_proc_nodes.size()); ++i)

          _pi[_proc_nodes[i]] += _dist[_proc_nodes[i]] - t_dist;

        return true;

      }

    }; //class ResidualDijkstra

  private:

    // The digraph the algorithm runs on

    const Digraph &_graph;

    // The length map

    const LengthMap &_length;

    // Arc map of the current flow

    FlowMap *_flow;

    bool _local_flow;

    // Node map of the current potentials

    PotentialMap *_potential;

    bool _local_potential;

    // The source node

    Node _s;

    // The target node

    Node _t;

    // Container to store the found paths

    std::vector<Path> _paths;

    int _path_num;

    // The pred arc map

    PredMap _pred;

    // Data for full init

    PotentialMap *_init_dist;

    PredMap *_init_pred;

    bool _full_init;

  public:

    /// \brief Constructor.

    ///

    /// Constructor.

    ///

    /// \param graph The digraph the algorithm runs on.

    /// \param length The length (cost) values of the arcs.

    Suurballe( const Digraph &graph,

               const LengthMap &length ) :

      _graph(graph), _length(length), _flow(0), _local_flow(false),

      _potential(0), _local_potential(false), _pred(graph),

      _init_dist(0), _init_pred(0)

    {}

    /// Destructor.

    ~Suurballe() {

      if (_local_flow) delete _flow;

      if (_local_potential) delete _potential;

      delete _init_dist;

      delete _init_pred;

    }

    /// \brief Set the flow map.

    ///

    /// This function sets the flow map.

    /// If it is not used before calling \ref run() or \ref init(),

    /// an instance will be allocated automatically. The destructor

    /// deallocates this automatically allocated map, of course.

    ///

    /// The found flow contains only 0 and 1 values, since it is the

    /// union of the found arc-disjoint paths.

    ///

    /// \return <tt>(*this)</tt>

    Suurballe& flowMap(FlowMap &map) {

      if (_local_flow) {

        delete _flow;

        _local_flow = false;

      }

      _flow = ↦

      return *this;

    }

    /// \brief Set the potential map.

    ///

    /// This function sets the potential map.

    /// If it is not used before calling \ref run() or \ref init(),

    /// an instance will be allocated automatically. The destructor

    /// deallocates this automatically allocated map, of course.

    ///

    /// The node potentials provide the dual solution of the underlying

    /// \ref min_cost_flow "minimum cost flow problem".

    ///

    /// \return <tt>(*this)</tt>

    Suurballe& potentialMap(PotentialMap &map) {

      if (_local_potential) {

        delete _potential;

        _local_potential = false;

      }

      _potential = ↦

      return *this;

    }

    /// \name Execution Control

    /// The simplest way to execute the algorithm is to call the run()

    /// function.\n

    /// If you need to execute the algorithm many times using the same

    /// source node, then you may call fullInit() once and start()

    /// for each target node.\n

    /// If you only need the flow that is the union of the found

    /// arc-disjoint paths, then you may call findFlow() instead of

    /// start().

    /// @{

    /// \brief Run the algorithm.

    ///

    /// This function runs the algorithm.

    ///

    /// \param s The source node.

    /// \param t The target node.

    /// \param k The number of paths to be found.

    ///

    /// \return \c k if there are at least \c k arc-disjoint paths from

    /// \c s to \c t in the digraph. Otherwise it returns the number of

    /// arc-disjoint paths found.

    ///

    /// \note Apart from the return value, <tt>s.run(s, t, k)</tt> is

    /// just a shortcut of the following code.

    /// \code

    ///   s.init(s);

    ///   s.start(t, k);

    /// \endcode

    int run(const Node& s, const Node& t, int k = 2) {

      init(s);

      start(t, k);

      return _path_num;

    }

    /// \brief Initialize the algorithm.

    ///

    /// This function initializes the algorithm with the given source node.

    ///

    /// \param s The source node.

    void init(const Node& s) {

      _s = s;

      // Initialize maps

      if (!_flow) {

        _flow = new FlowMap(_graph);

        _local_flow = true;

      }

      if (!_potential) {

        _potential = new PotentialMap(_graph);

        _local_potential = true;

      }

      _full_init = false;

    }

    /// \brief Initialize the algorithm and perform Dijkstra.

    ///

    /// This function initializes the algorithm and performs a full

    /// Dijkstra search from the given source node. It makes consecutive

    /// executions of \ref start() "start(t, k)" faster, since they

    /// have to perform %Dijkstra only k-1 times.

    ///

    /// This initialization is usually worth using instead of \ref init()

    /// if the algorithm is executed many times using the same source node.

    ///

    /// \param s The source node.

    void fullInit(const Node& s) {

      // Initialize maps

      init(s);

      if (!_init_dist) {

        _init_dist = new PotentialMap(_graph);

      }

      if (!_init_pred) {

        _init_pred = new PredMap(_graph);

      }

      // Run a full Dijkstra

      typename Dijkstra<Digraph, LengthMap>

        ::template SetStandardHeap<Heap>

        ::template SetDistMap<PotentialMap>

        ::template SetPredMap<PredMap>

        ::Create dijk(_graph, _length);

      dijk.distMap(*_init_dist).predMap(*_init_pred);

      dijk.run(s);

      _full_init = true;

    }

    /// \brief Execute the algorithm.

    ///

    /// This function executes the algorithm.

    ///

    /// \param t The target node.

    /// \param k The number of paths to be found.

    ///

    /// \return \c k if there are at least \c k arc-disjoint paths from

    /// \c s to \c t in the digraph. Otherwise it returns the number of

    /// arc-disjoint paths found.

    ///

    /// \note Apart from the return value, <tt>s.start(t, k)</tt> is

    /// just a shortcut of the following code.

    /// \code

    ///   s.findFlow(t, k);

    ///   s.findPaths();

    /// \endcode

    int start(const Node& t, int k = 2) {

      findFlow(t, k);

      findPaths();

      return _path_num;

    }

    /// \brief Execute the algorithm to find an optimal flow.

    ///

    /// This function executes the successive shortest path algorithm to

    /// find a minimum cost flow, which is the union of \c k (or less)

    /// arc-disjoint paths.

    ///

    /// \param t The target node.

    /// \param k The number of paths to be found.

    ///

    /// \return \c k if there are at least \c k arc-disjoint paths from

    /// the source node to the given node \c t in the digraph.

    /// Otherwise it returns the number of arc-disjoint paths found.

    ///

    /// \pre \ref init() must be called before using this function.

    int findFlow(const Node& t, int k = 2) {

      _t = t;

      ResidualDijkstra dijkstra(*this);

      // Initialization

      for (ArcIt e(_graph); e != INVALID; ++e) {

        (*_flow)[e] = 0;

      }

      if (_full_init) {

        for (NodeIt n(_graph); n != INVALID; ++n) {

          (*_potential)[n] = (*_init_dist)[n];

        }

        Node u = _t;

        Arc e;

        while ((e = (*_init_pred)[u]) != INVALID) {

          (*_flow)[e] = 1;

          u = _graph.source(e);

        }        

        _path_num = 1;

      } else {

        for (NodeIt n(_graph); n != INVALID; ++n) {

          (*_potential)[n] = 0;

        }

        _path_num = 0;

      }

      // Find shortest paths

      while (_path_num < k) {

        // Run Dijkstra

        if (!dijkstra.run(_path_num)) break;

        ++_path_num;

        // Set the flow along the found shortest path

        Node u = _t;

        Arc e;

        while ((e = _pred[u]) != INVALID) {

          if (u == _graph.target(e)) {

            (*_flow)[e] = 1;

            u = _graph.source(e);

          } else {

            (*_flow)[e] = 0;

            u = _graph.target(e);

          }

        }

      }

      return _path_num;

    }

    /// \brief Compute the paths from the flow.

    ///

    /// This function computes arc-disjoint paths from the found minimum

    /// cost flow, which is the union of them.

    ///

    /// \pre \ref init() and \ref findFlow() must be called before using

    /// this function.

    void findPaths() {

      FlowMap res_flow(_graph);

      for(ArcIt a(_graph); a != INVALID; ++a) res_flow[a] = (*_flow)[a];

      _paths.clear();

      _paths.resize(_path_num);

      for (int i = 0; i < _path_num; ++i) {

        Node n = _s;

        while (n != _t) {

          OutArcIt e(_graph, n);

          for ( ; res_flow[e] == 0; ++e) ;

          n = _graph.target(e);

          _paths[i].addBack(e);

          res_flow[e] = 0;

        }

      }

    }

    /// @}

    /// \name Query Functions

    /// The results of the algorithm can be obtained using these

    /// functions.

    /// \n The algorithm should be executed before using them.

    /// @{

    /// \brief Return the total length of the found paths.

    ///

    /// This function returns the total length of the found paths, i.e.

    /// the total cost of the found flow.

    /// The complexity of the function is O(e).

    ///

    /// \pre \ref run() or \ref findFlow() must be called before using

    /// this function.

    Length totalLength() const {

      Length c = 0;

      for (ArcIt e(_graph); e != INVALID; ++e)

        c += (*_flow)[e] * _length[e];

      return c;

    }

    /// \brief Return the flow value on the given arc.

    ///

    /// This function returns the flow value on the given arc.

    /// It is \c 1 if the arc is involved in one of the found arc-disjoint

    /// paths, otherwise it is \c 0.

    ///

    /// \pre \ref run() or \ref findFlow() must be called before using

    /// this function.

    int flow(const Arc& arc) const {

      return (*_flow)[arc];

    }

    /// \brief Return a const reference to an arc map storing the

    /// found flow.

    ///

    /// This function returns a const reference to an arc map storing

    /// the flow that is the union of the found arc-disjoint paths.

    ///

    /// \pre \ref run() or \ref findFlow() must be called before using

    /// this function.

    const FlowMap& flowMap() const {

      return *_flow;

    }

    /// \brief Return the potential of the given node.

    ///

    /// This function returns the potential of the given node.

    /// The node potentials provide the dual solution of the

    /// underlying \ref min_cost_flow "minimum cost flow problem".

    ///

    /// \pre \ref run() or \ref findFlow() must be called before using

    /// this function.

    Length potential(const Node& node) const {

      return (*_potential)[node];

    }

    /// \brief Return a const reference to a node map storing the

    /// found potentials (the dual solution).

    ///

    /// This function returns a const reference to a node map storing

    /// the found potentials that provide the dual solution of the

    /// underlying \ref min_cost_flow "minimum cost flow problem".

    ///

    /// \pre \ref run() or \ref findFlow() must be called before using

    /// this function.

    const PotentialMap& potentialMap() const {

      return *_potential;

    }

    /// \brief Return the number of the found paths.

    ///

    /// This function returns the number of the found paths.

    ///

    /// \pre \ref run() or \ref findFlow() must be called before using

    /// this function.

    int pathNum() const {

      return _path_num;

    }

    /// \brief Return a const reference to the specified path.

    ///

    /// This function returns a const reference to the specified path.

    ///

    /// \param i The function returns the <tt>i</tt>-th path.

    /// \c i must be between \c 0 and <tt>%pathNum()-1</tt>.

    ///

    /// \pre \ref run() or \ref findPaths() must be called before using

    /// this function.

    const Path& path(int i) const {

      return _paths[i];

    }

    /// @}

  }; //class Suurballe

  ///@}

} //namespace lemon

#endif //LEMON_SUURBALLE_H