source: src/Actions/FragmentationAction/FragmentationAction.cpp@ d635829

/*
 * Project: MoleCuilder
 * Description: creates and alters molecular systems
 * Copyright (C)  2010-2012 University of Bonn. All rights reserved.
 * Copyright (C)  2013 Frederik Heber. All rights reserved.
 * 
 *
 *   This file is part of MoleCuilder.
 *
 *    MoleCuilder 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 2 of the License, or
 *    (at your option) any later version.
 *
 *    MoleCuilder 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 MoleCuilder.  If not, see <http://www.gnu.org/licenses/>.
 */

/*
 * FragmentationAction.cpp
 *
 *  Created on: May 9, 2010
 *      Author: heber
 */

// include config.h
#ifdef HAVE_CONFIG_H
#include <config.h>
#endif

#include "CodePatterns/MemDebug.hpp"

#include "Atom/atom.hpp"
#include "CodePatterns/IteratorAdaptors.hpp"
#include "CodePatterns/Log.hpp"
#include "Descriptors/AtomSelectionDescriptor.hpp"
#include "Fragmentation/Exporters/ExportGraph_ToFiles.hpp"
#include "Fragmentation/Exporters/ExportGraph_ToJobs.hpp"
#include "Fragmentation/Exporters/SaturatedFragment.hpp"
#include "Fragmentation/Exporters/SphericalPointDistribution.hpp"
#include "Fragmentation/Fragmentation.hpp"
#include "Fragmentation/Graph.hpp"
#include "Fragmentation/HydrogenSaturation_enum.hpp"
#include "Fragmentation/Interfragmenter.hpp"
#include "Fragmentation/KeySetsContainer.hpp"
#include "Fragmentation/Summation/Containers/FragmentationResultContainer.hpp"
#include "Graph/AdjacencyList.hpp"
#include "Graph/BondGraph.hpp"
#include "Graph/CyclicStructureAnalysis.hpp"
#include "Graph/DepthFirstSearchAnalysis.hpp"
#include "Helpers/defs.hpp"
#include "molecule.hpp"
#include "World.hpp"

#include <boost/shared_ptr.hpp>
#include <boost/filesystem.hpp>
#include <algorithm>
#include <iostream>
#include <map>
#include <string>
#include <vector>

#include "Actions/FragmentationAction/FragmentationAction.hpp"

using namespace MoleCuilder;

// and construct the stuff
#include "FragmentationAction.def"
#include "Action_impl_pre.hpp"
/** =========== define the function ====================== */
ActionState::ptr FragmentationFragmentationAction::performCall() {
  clock_t start,end;
  int ExitFlag = -1;
  World &world = World::getInstance();

  // inform about used parameters
  LOG(0, "STATUS: Fragmenting molecular system with current connection matrix up to "
      << params.order.get() << " order. ");
  if (params.types.get().size() != 0)
    LOG(0, "STATUS: Fragment files begin with "
        << params.prefix.get() << " and are stored as: "
        << params.types.get() << "." << std::endl);

  // check for selected atoms
  if (world.beginAtomSelection() == world.endAtomSelection()) {
    STATUS("There are no atoms selected for fragmentation.");
    return Action::failure;
  }

  // go through all atoms, note down their molecules and group them
  typedef std::multimap<molecule *, atom *> clusters_t;
  typedef std::vector<atomId_t> atomids_t;
  atomids_t atomids;
  clusters_t clusters;
  for (World::AtomSelectionConstIterator iter = world.beginAtomSelection();
      iter != world.endAtomSelection(); ++iter) {
    clusters.insert( std::make_pair(iter->second->getMolecule(), iter->second) );
    atomids.push_back(iter->second->getId());
  }
  {
    std::vector<molecule *> molecules;
    molecules.insert( molecules.end(), MapKeyIterator<clusters_t::const_iterator>(clusters.begin()),
        MapKeyIterator<clusters_t::const_iterator>(clusters.end()) );
    molecules.erase( std::unique(molecules.begin(), molecules.end()), molecules.end() );
    LOG(1, "INFO: There are " << molecules.size()  << " molecules to consider.");
  }

  // parse in Adjacency file
  boost::shared_ptr<AdjacencyList> FileChecker;
  boost::filesystem::path filename(params.prefix.get() + std::string(ADJACENCYFILE));
  if (boost::filesystem::exists(filename) && boost::filesystem::is_regular_file(filename)) {
    std::ifstream File;
    File.open(filename.string().c_str(), ios::out);
    FileChecker.reset(new AdjacencyList(File));
    File.close();
  } else {
    LOG(1, "INFO: Could not open default adjacency file " << filename.string() << ".");
    FileChecker.reset(new AdjacencyList);
  }

  // make sure bond degree is correct
  {
    BondGraph *BG = World::getInstance().getBondGraph();
    World::AtomComposite Set = World::getInstance().getAllAtoms(AtomsBySelection());
    BG->CorrectBondDegree(Set);
  }

  // we parse in the keysets from last time if present
  Graph StoredGraph;
  StoredGraph.ParseKeySetFile(params.prefix.get());

  start = clock();
  // go through all keys (i.e. all molecules)
  clusters_t::const_iterator advanceiter;
  Graph TotalGraph;
  int keysetcounter = 0;
  for (clusters_t::const_iterator iter = clusters.begin();
      iter != clusters.end();
      iter = advanceiter) {
    // get iterator to past last atom in this molecule
    molecule * mol = iter->first;
    advanceiter = clusters.upper_bound(mol);

    // copy molecule's atoms' ids as parameters to Fragmentation's AtomMask
    std::vector<atomId_t> mols_atomids;
    std::transform(iter, advanceiter, std::back_inserter(mols_atomids),
        boost::bind( &atom::getNr,
            boost::bind( &clusters_t::value_type::second, _1 ))
    );
    LOG(2, "INFO: Fragmenting in molecule " << mol->getName() << " in " << clusters.count(mol) 
        << " atoms, out of " << mol->getAtomCount() << ".");
    const enum HydrogenTreatment treatment =  params.HowtoTreatHydrogen.get() ? ExcludeHydrogen : IncludeHydrogen;
    Fragmentation Fragmenter(mol, *FileChecker, treatment);

    // perform fragmentation
    LOG(0, std::endl << " ========== Fragmentation of molecule " << mol->getName() << " ========================= ");
    {
      Graph StoredLocalGraph(StoredGraph.getLocalGraph(mol));
      const int tempFlag = Fragmenter.FragmentMolecule(mols_atomids, params.order.get(), params.prefix.get(), StoredLocalGraph);
      if ((ExitFlag == 2) && (tempFlag != 2))
        ExitFlag = tempFlag; // if there is one molecule that needs further fragmentation, it overrides others
      if (ExitFlag == -1)
        ExitFlag = tempFlag; // if we are the first, we set the standard
    }
    if (TotalGraph.empty()) {
      TotalGraph = Fragmenter.getGraph();
      keysetcounter = TotalGraph.size();
    } else
      TotalGraph.InsertGraph(Fragmenter.getGraph(), keysetcounter);

  }
  // add full cycles if desired
  if (params.DoCyclesFull.get()) {
    // get the BackEdgeStack from somewhere
    DepthFirstSearchAnalysis DFS;
    DFS();
    std::deque<bond::ptr> BackEdgeStack = DFS.getBackEdgeStack();
    // then we analyse the cycles and get them
    CyclicStructureAnalysis CycleAnalysis(params.HowtoTreatHydrogen.get() ? ExcludeHydrogen : IncludeHydrogen);
    CycleAnalysis(&BackEdgeStack);
    CyclicStructureAnalysis::cycles_t cycles = CycleAnalysis.getAllCycles();
    // sort them according to KeySet::operator<()
    std::sort(cycles.begin(), cycles.end());
    // store all found cycles to file
    {
      boost::filesystem::path filename(params.prefix.get() + std::string(CYCLEKEYSETFILE));
      std::ofstream File;
      LOG(1, "INFO: Storing cycle keysets to " << filename.string() << ".");
      File.open(filename.string().c_str(), ios::out);
      for (CyclicStructureAnalysis::cycles_t::const_iterator iter = cycles.begin();
          iter != cycles.end(); ++iter) {
        for (CyclicStructureAnalysis::cycle_t::const_iterator cycleiter = (*iter).begin();
            cycleiter != (*iter).end(); ++cycleiter) {
          File << *cycleiter << "\t";
        }
        File << "\n";
      }
      File.close();
    }
    // ... and to result container
    {
      KeySetsContainer cyclekeys;
      for (CyclicStructureAnalysis::cycles_t::const_iterator iter = cycles.begin();
          iter != cycles.end(); ++iter) {
        const CyclicStructureAnalysis::cycle_t &cycle = *iter;
        const size_t order = cycle.size();
        KeySetsContainer::IntVector temp_cycle(cycle.begin(), cycle.end());
        cyclekeys.insert(temp_cycle, order);
      }
      FragmentationResultContainer::getInstance().addCycles(cyclekeys);
    }
    // Create graph and insert into TotalGraph
    LOG(0, "STATUS: Adding " << cycles.size() << " cycles.");
    {
      Graph CycleGraph;
      for (CyclicStructureAnalysis::cycles_t::const_iterator iter = cycles.begin();
          iter != cycles.end(); ++iter) {
        const CyclicStructureAnalysis::cycle_t &currentcycle = *iter;
        LOG(2, "INFO: Inserting cycle " << currentcycle << ".");
#ifndef NDEBUG
        std::pair< Graph::iterator, bool > inserter =
#endif
        CycleGraph.insert( std::make_pair(currentcycle, NumberValuePair(1,1.)) );
        ASSERT( inserter.second,
            "FragmentationFragmentationAction::performCall() - keyset "
            +toString(currentcycle)+" inserted twice into CycleGraph.");
      }
      TotalGraph.InsertGraph(CycleGraph, keysetcounter);
    }
  }

  LOG(0, "STATUS: There are " << TotalGraph.size() << " fragments.");

  {
    // remove OrderAtSite file
    std::string line;
    std::ofstream file;
    line = params.prefix.get() + ORDERATSITEFILE;
    file.open(line.c_str(), std::ofstream::out | std::ofstream::trunc);
    file << "";
    file.close();
  }

  // now add interfragments
  if (params.InterOrder.get() != 0) {
    LOG(0, "STATUS: Putting fragments together up to order "
        << params.InterOrder.get() << " and distance of "
        << params.distance.get() << ".");
    Interfragmenter fragmenter(TotalGraph);
    const enum HydrogenTreatment treatment =  params.HowtoTreatHydrogen.get() ? ExcludeHydrogen : IncludeHydrogen;
    fragmenter(params.InterOrder.get(), params.distance.get(), treatment);
    LOG(0, "STATUS: There are now " << TotalGraph.size() << " fragments after interfragmenting.");
  }

  // store keysets to file
  {
    TotalGraph.StoreKeySetFile(params.prefix.get());
  }

  // create global saturation positions map
  SaturatedFragment::GlobalSaturationPositions_t globalsaturationpositions;
  {
    // go through each atom
    for (World::AtomSelectionConstIterator iter = world.beginAtomSelection();
        iter != world.endAtomSelection(); ++iter) {
      const atom * const _atom = iter->second;

      // skip hydrogens if treated special
      const enum HydrogenTreatment treatment =  params.HowtoTreatHydrogen.get() ? ExcludeHydrogen : IncludeHydrogen;
      if ((treatment == ExcludeHydrogen) && (_atom->getType()->getAtomicNumber() == 1)) {
        LOG(4, "DEBUG: Skipping hydrogen atom " << *_atom);
        continue;
      }

      // gather the polygon
      SphericalPointDistribution::WeightedPolygon_t polygon;
      const BondList &bondlist = _atom->getListOfBonds();
      for (BondList::const_iterator bonditer = bondlist.begin();
          bonditer != bondlist.end(); ++bonditer) {
        const atom * const _otheratom = (*bonditer)->GetOtherAtom(_atom);
        Vector bondvector = _otheratom->getPosition() - _atom->getPosition();
        bondvector.Normalize();
        polygon.push_back(
            std::make_pair(
                bondvector,
                (*bonditer)->getDegree()
            ) );
      }
      LOG(3, "DEBUG: Polygon is " << polygon << " for this atom " << *_atom << ".");

      // get the valence
      unsigned int NumberOfPoints = _atom->getElement().getNoValenceOrbitals();
      LOG(3, "DEBUG: There are " << NumberOfPoints
          << " places to fill in in total for this atom " << *_atom << ".");

      // check whether there are any bonds with degree larger than 1
      unsigned int SumOfDegrees = 0;
      bool PresentHigherBonds = false;
      for (BondList::const_iterator bonditer = bondlist.begin();
          bonditer != bondlist.end(); ++bonditer) {
        SumOfDegrees += (*bonditer)->getDegree();
        PresentHigherBonds |= (*bonditer)->getDegree() > 1;
      }

      // get the association for each bond neighbor
      SphericalPointDistribution SPD;
      SphericalPointDistribution::PolygonWithIndexTuples globalinfo;
      if ((NumberOfPoints != SumOfDegrees) || (PresentHigherBonds))
        globalinfo = SPD.getAssociatedPoints(polygon, NumberOfPoints);
      else
        globalinfo = SphericalPointDistribution::getIdentityAssociation(polygon);
      LOG(3, "DEBUG: Original polygon was " << polygon << " for this atom " << *_atom << ".");
      LOG(3, "DEBUG: Got global saturation polygon " << globalinfo.polygon
          << " for atom " << *_atom);
      LOG(3, "DEBUG: Associated matching is " << globalinfo.indices
          << " for atom " << *_atom);

      // convert into the desired entry in the map
      SaturatedFragment::SaturationsPositionsPerNeighbor_t positions_per_neighbor;
      BondList::const_iterator bonditer = bondlist.begin();
      // same number of bonds and tuples
      ASSERT( bondlist.size() == globalinfo.indices.size(),
          "FragmentationAction::performCall() - number of bonds "
          +toString(bondlist.size())+" unequal to number of tuples "
          +toString(globalinfo.indices.size()));
      for (SphericalPointDistribution::IndexTupleList_t::const_iterator tupleiter =
          globalinfo.indices.begin();
          tupleiter != globalinfo.indices.end(); ++tupleiter, ++bonditer) {
        // note that we cannot check whether the designated vector matches with the
        // normalized bond vector where it should have originated because the
        // vector has been rotated.
//        Vector bondvector =
//            (*bonditer)->GetOtherAtom(_atom)->getPosition()-_atom->getPosition();
//        bondvector.Normalize();
        // no go through all indices of the tuple (and positions)
        for (SphericalPointDistribution::IndexList_t::const_iterator indexiter =
            tupleiter->begin(); indexiter != tupleiter->end(); ++indexiter) {
          // and add them, first trying with new list
          std::pair<
              SaturatedFragment::SaturationsPositionsPerNeighbor_t::iterator,
              bool
              > inserter =
                  positions_per_neighbor.insert(
                      std::make_pair(
                          (*bonditer)->GetOtherAtom(_atom)->getId(),
                          SaturatedFragment::SaturationsPositions_t(
                              1,
                              globalinfo.polygon[*indexiter])
                      )
                  );
          // if already pressent, add to this present list
          if (!inserter.second) {
            inserter.first->second.push_back(globalinfo.polygon[*indexiter]);
          }
        }
      }
      // bonditer follows nicely
      ASSERT( bonditer == bondlist.end(),
          "FragmentationAction::performCall() - bonditer is out of step, it still points at bond "
          +toString(*bonditer)+".");

      // and insert
      globalsaturationpositions.insert(
          std::make_pair( _atom->getId(),
              positions_per_neighbor
          ));
    }
  }

  {
    const enum HydrogenSaturation saturation =  params.DoSaturation.get() ? DoSaturate : DontSaturate;
    const enum HydrogenTreatment treatment =  params.HowtoTreatHydrogen.get() ? ExcludeHydrogen : IncludeHydrogen;
    if (params.types.get().size() != 0) {
      // store molecule's fragment to file
      ExportGraph_ToFiles exporter(TotalGraph, treatment, saturation, globalsaturationpositions);
      exporter.setPrefix(params.prefix.get());
      exporter.setOutputTypes(params.types.get());
      exporter();
    } else {
      // store molecule's fragment in FragmentJobQueue
      ExportGraph_ToJobs exporter(TotalGraph, treatment, saturation, globalsaturationpositions);
      exporter.setLevel(params.level.get());
      exporter();
    }
  }

  // store Adjacency to file
  {
    std::string filename = params.prefix.get() + ADJACENCYFILE;
    std::ofstream AdjacencyFile;
    AdjacencyFile.open(filename.c_str(), ios::out);
    AdjacencyList adjacency(atomids);
    adjacency.StoreToFile(AdjacencyFile);
    AdjacencyFile.close();
  }

  World::getInstance().setExitFlag(ExitFlag);
  end = clock();
  LOG(0, "STATUS: Clocks for this operation: " << (end-start) << ", time: " << ((double)(end-start)/CLOCKS_PER_SEC) << "s.");

  return Action::success;
}

ActionState::ptr FragmentationFragmentationAction::performUndo(ActionState::ptr _state) {
  return Action::success;
}

ActionState::ptr FragmentationFragmentationAction::performRedo(ActionState::ptr _state){
  return Action::success;
}

bool FragmentationFragmentationAction::canUndo() {
  return true;
}

bool FragmentationFragmentationAction::shouldUndo() {
  return true;
}
/** =========== end of function ====================== */