//
// Software License for AMDiS
//
// Copyright (c) 2010 Dresden University of Technology 
// All rights reserved.
// Authors: Simon Vey, Thomas Witkowski et al.
//
// This file is part of AMDiS
//
// See also license.opensource.txt in the distribution.


#include <algorithm>
#include <iostream>
#include <fstream>
#include <limits>
#include <stdint.h>
#include <boost/lexical_cast.hpp>
#include <boost/filesystem.hpp>

#include "parallel/MeshDistributor.h"
#include "parallel/MeshManipulation.h"
#include "parallel/ParallelDebug.h"
#include "parallel/StdMpi.h"
#include "parallel/ParMetisPartitioner.h"
#include "io/ElementFileWriter.h"
#include "io/MacroInfo.h"
#include "io/VtkWriter.h"
#include "Mesh.h"
#include "Traverse.h"
#include "ElInfo.h"
#include "Element.h"
#include "MacroElement.h"
#include "DOFMatrix.h"
#include "DOFVector.h"
#include "SystemVector.h"
#include "ElementDofIterator.h"
#include "ProblemStatBase.h"
#include "StandardProblemIteration.h"
#include "VertexVector.h"
#include "MeshStructure.h"
#include "ProblemVec.h"
#include "ProblemInstat.h"
#include "Debug.h"

namespace AMDiS {

  using boost::lexical_cast;
  using namespace boost::filesystem;
  using namespace std;

  inline bool cmpDofsByValue(const DegreeOfFreedom* dof1, const DegreeOfFreedom* dof2)
  {
    return (*dof1 < *dof2);
  }


  MeshDistributor::MeshDistributor(string str)
    : probStat(0),
      name(str),
      feSpace(NULL),
      mesh(NULL),
      refineManager(NULL),
      info(10),
      partitioner(NULL),
      nRankDofs(0),
      nOverallDofs(0),
      rstart(0),
      deserialized(false),
      writeSerializationFile(false),
      repartitioningAllowed(false),
      repartitionIthChange(20),
      nTimestepsAfterLastRepartitioning(0),
      repartCounter(0),
      debugOutputDir(""),
      lastMeshChangeIndex(0)
  {
    FUNCNAME("MeshDistributor::ParalleDomainBase()");

    mpiRank = MPI::COMM_WORLD.Get_rank();
    mpiSize = MPI::COMM_WORLD.Get_size();
    mpiComm = MPI::COMM_WORLD;

    int tmp = 0;
    GET_PARAMETER(0, name + "->repartitioning", "%d", &tmp);
    repartitioningAllowed = (tmp > 0);

    GET_PARAMETER(0, name + "->debug output dir", &debugOutputDir);
    GET_PARAMETER(0, name + "->repartition ith change", "%d", &repartitionIthChange);

    tmp = 0;
    GET_PARAMETER(0, name + "->log main rank", "%d", &tmp);
    Msg::outputMainRank = (tmp > 0);
  }


  void MeshDistributor::initParallelization()
  {
    FUNCNAME("MeshDistributor::initParallelization()");

    TEST_EXIT(mpiSize > 1)
      ("Parallelization does not work with only one process!\n");

    TEST_EXIT(feSpace)("No FE space has been defined for the mesh distributor!\n");
    TEST_EXIT(mesh)("No mesh has been defined for the mesh distributor!\n");

    elObjects.setFeSpace(feSpace);

    // If the problem has been already read from a file, we need only to set
    // isRankDofs to all matrices and rhs vector and to remove periodic 
    // boundary conditions (if there are some).
    if (deserialized) {
      updateMacroElementInfo();

      setRankDofs();

      removePeriodicBoundaryConditions();

      macroElIndexMap.clear();
      macroElIndexTypeMap.clear();

      map<int, bool>& elementInRank = partitioner->getElementInRank();
      for (vector<MacroElement*>::iterator it = allMacroElements.begin();
	   it != allMacroElements.end(); ++it) {
	elementInRank[(*it)->getIndex()] = false;
	macroElIndexMap[(*it)->getIndex()] = (*it)->getElement();
	macroElIndexTypeMap[(*it)->getIndex()] = (*it)->getElType();
      }

      for (deque<MacroElement*>::iterator it = mesh->getMacroElements().begin();
	   it != mesh->getMacroElements().end(); ++it)
	elementInRank[(*it)->getIndex()] = true;      

      return;
    }

   
    // Test, if the mesh is the macro mesh only! Paritioning of the mesh is supported
    // only for macro meshes, so it will not work yet if the mesh is already refined
    // in some way.
    testForMacroMesh();

    // For later mesh repartitioning, we need to store some information about the
    // macro mesh.
    createMacroElementInfo();

    // create an initial partitioning of the mesh
    partitioner->createPartitionData();

    // set the element weights, which are 1 at the very first begin
    setInitialElementWeights();

    // and now partition the mesh    
    partitioner->fillCoarsePartitionVec(&oldPartitionVec);

    bool partitioningSucceed = partitioner->partition(elemWeights, INITIAL);
    TEST_EXIT(partitioningSucceed)("Initial partitioning does not work!\n");

    partitioner->fillCoarsePartitionVec(&partitionVec);


#if (DEBUG != 0)
    debug::ElementIdxToDofs elMap;
    debug::createSortedDofs(mesh, elMap);
    if (mpiRank == 0) {
      int writePartMesh = 1;
      GET_PARAMETER(0, "dbg->write part mesh", "%d", &writePartMesh);

      if (writePartMesh > 0) {
	debug::writeElementIndexMesh(mesh, debugOutputDir + "elementIndex.vtu");
	writePartitioningMesh(debugOutputDir + "part.vtu");
      } else {
	MSG("Skip write part mesh!\n");
      }
    }
#endif


    // === Create interior boundary information. ===

    createInteriorBoundaryInfo();

#if (DEBUG != 0)
    ParallelDebug::printBoundaryInfo(*this);
#endif

    for (deque<MacroElement*>::iterator it = mesh->firstMacroElement();
	 it != mesh->endOfMacroElements(); ++it) {
      for (int i = 0; i < mesh->getGeo(NEIGH); i++) {
	if ((*it)->getNeighbour(i) && 
	    mesh->isPeriodicAssociation((*it)->getBoundary(i))) {
	  (*it)->getNeighbour(i)->setNeighbour((*it)->getOppVertex(i), NULL);
	  (*it)->setNeighbour(i, NULL);
	  (*it)->setBoundary(i, 0);

	  macroElementNeighbours[(*it)->getIndex()][i] = -1;
	}
      }
    }

    for (vector<MacroElement*>::iterator it = allMacroElements.begin();
	 it != allMacroElements.end(); ++it) {
      for (int i = 0; i < mesh->getGeo(NEIGH); i++) {
	if ((*it)->getNeighbour(i) && 
	    mesh->isPeriodicAssociation((*it)->getBoundary(i))) {
	  (*it)->getNeighbour(i)->setNeighbour((*it)->getOppVertex(i), NULL);
	  (*it)->setNeighbour(i, NULL);
	  (*it)->setBoundary(i, 0);

	  macroElementNeighbours[(*it)->getIndex()][i] = -1;
	}
      }
    }

    // === Remove all macro elements that are not part of the rank partition. ===

    removeMacroElements();


    // === Create new global and local DOF numbering. ===


    // We have to remove the VertexVectors, which contain periodic assoiciations, 
    // because they are not valid anymore after some macro elements have been removed
    // and the corresponding DOFs were deleted.
    for (map<BoundaryType, VertexVector*>::iterator it = mesh->getPeriodicAssociations().begin();
	 it != mesh->getPeriodicAssociations().end(); ++it)
      const_cast<DOFAdmin&>(mesh->getDofAdmin(0)).removeDOFContainer(dynamic_cast<DOFContainer*>(it->second));

    updateLocalGlobalNumbering();


    // === If in debug mode, make some tests. ===

#if (DEBUG != 0)
    MSG("AMDiS runs in debug mode, so make some test ...\n");

    ParallelDebug::testAllElements(*this);
    debug::testSortedDofs(mesh, elMap);
    ParallelDebug::testInteriorBoundary(*this);
    ParallelDebug::testCommonDofs(*this, true);
    ParallelDebug::testGlobalIndexByCoords(*this);

    debug::writeMesh(feSpace, -1, debugOutputDir + "macro_mesh");   

    MSG("Debug mode tests finished!\n");
#endif


    // === Create periodic dof mapping, if there are periodic boundaries. ===

    createPeriodicMap();


    // === Global refinements. ===
    
    int globalRefinement = 0;
    GET_PARAMETER(0, mesh->getName() + "->global refinements", "%d", &globalRefinement);
    
    if (globalRefinement > 0) {
      refineManager->globalRefine(mesh, globalRefinement);

      updateLocalGlobalNumbering();

     
      // === Update periodic mapping, if there are periodic boundaries. ===     

      createPeriodicMap();
    }


    /// === Set DOF rank information to all matrices and vectors. ===

    setRankDofs();


    // === Remove periodic boundary conditions in sequential problem definition. ===

    removePeriodicBoundaryConditions();
  }


  void MeshDistributor::addProblemStat(ProblemVec *probVec)
  {
    FUNCNAME("MeshDistributor::addProblemVec()");

    if (feSpace != NULL) {
      vector<FiniteElemSpace*> feSpaces = probVec->getFeSpaces();
      for (unsigned int i = 0; i < feSpaces.size(); i++) {
	TEST_EXIT(feSpace == feSpaces[i])
	  ("Parallelizaton is not supported for multiple FE spaces!\n");
      }
    } else {
      feSpace = probVec->getFeSpace(0);
      mesh = feSpace->getMesh();
      info = probVec->getInfo();
      
      TEST_EXIT(mesh->getNumberOfDOFAdmin() == 1)
	("Only meshes with one DOFAdmin are supported!\n");
      TEST_EXIT(mesh->getDofAdmin(0).getNumberOfPreDofs(VERTEX) == 0)
	("Wrong pre dof number for DOFAdmin!\n");
      
      switch (mesh->getDim()) {
      case 2:
	refineManager = new RefinementManager2d();
	break;
      case 3:
	refineManager = new RefinementManager3d();
	break;
      default:
	ERROR_EXIT("This should not happen for dim = %d!\n", mesh->getDim());
      }

      partitioner = new ParMetisPartitioner(mesh, &mpiComm);
    }

    // Create parallel serialization file writer, if needed.
    int writeSerialization = 0;
    GET_PARAMETER(0, probVec->getName() + "->output->write serialization", "%d", 
		  &writeSerialization);
    if (writeSerialization && !writeSerializationFile) {
      string filename = "";
      GET_PARAMETER(0, name + "->output->serialization filename", &filename);
      
      TEST_EXIT(filename != "")
	("No filename defined for parallel serialization file!\n");

      int tsModulo = -1;
      GET_PARAMETER(0, probVec->getName() + "->output->write every i-th timestep", 
		    "%d", &tsModulo);
      
      probVec->getFileWriterList().push_back(new Serializer<MeshDistributor>(this, filename, tsModulo));
      writeSerializationFile = true;
    }    

    int readSerialization = 0;
    GET_PARAMETER(0, probVec->getName() + "->input->read serialization", "%d", 
		  &readSerialization);
    if (readSerialization) {
      string filename = "";
      GET_PARAMETER(0, probVec->getName() + "->input->serialization filename", &filename);
      filename += ".p" + lexical_cast<string>(mpiRank);
      MSG("Start deserialization with %s\n", filename.c_str());
      ifstream in(filename.c_str());

      TEST_EXIT(!in.fail())("Could not open deserialization file: %s\n",
			    filename.c_str());

      probVec->deserialize(in);
      in.close();
      MSG("Deserialization from file: %s\n", filename.c_str());

      filename = "";
      GET_PARAMETER(0, name + "->input->serialization filename", &filename);
      
      TEST_EXIT(filename != "")
	("No filename defined for parallel deserialization file!\n");
      
      string rankFilename = filename + ".p" + lexical_cast<string>(mpiRank);
      in.open(rankFilename.c_str());
      
      TEST_EXIT(!in.fail())("Could not open parallel deserialization file: %s\n",
			    filename.c_str());
      
      deserialize(in);
      in.close();
      MSG("Deserializtion of mesh distributor from file: %s\n", rankFilename.c_str());
      deserialized = true;
    }

    probStat.push_back(probVec);
  }


  void MeshDistributor::exitParallelization()
  {}

  
  void MeshDistributor::testForMacroMesh()
  {
    FUNCNAME("MeshDistributor::testForMacroMesh()");

    int nMacroElements = 0;

    TraverseStack stack;
    ElInfo *elInfo = stack.traverseFirst(mesh, -1, Mesh::CALL_LEAF_EL);
    while (elInfo) {
      TEST_EXIT(elInfo->getLevel() == 0)
	("Mesh is already refined! This does not work with parallelization!\n");

      TEST_EXIT(elInfo->getType() == 0)
	("Only macro elements with level 0 are supported!\n");
      
      nMacroElements++;

      elInfo = stack.traverseNext(elInfo);
    }

    TEST_EXIT(nMacroElements >= mpiSize)
      ("The mesh has less macro elements than number of mpi processes!\n");
  }


  void MeshDistributor::synchVector(DOFVector<double> &vec)
  {
    StdMpi<vector<double> > stdMpi(mpiComm);

    for (RankToDofContainer::iterator sendIt = sendDofs.begin();
	 sendIt != sendDofs.end(); ++sendIt) {
      vector<double> dofs;
      int nSendDofs = sendIt->second.size();
      dofs.reserve(nSendDofs);
      
      for (int i = 0; i < nSendDofs; i++)
	dofs.push_back(vec[*((sendIt->second)[i])]);

      stdMpi.send(sendIt->first, dofs);
    }

    for (RankToDofContainer::iterator recvIt = recvDofs.begin();
	 recvIt != recvDofs.end(); ++recvIt)
      stdMpi.recv(recvIt->first, recvIt->second.size());

    stdMpi.startCommunication<double>(MPI_DOUBLE);

    for (RankToDofContainer::iterator recvIt = recvDofs.begin();
	 recvIt != recvDofs.end(); ++recvIt)
      for (unsigned int i = 0; i < recvIt->second.size(); i++)
	vec[*(recvIt->second)[i]] = stdMpi.getRecvData(recvIt->first)[i];
  }


  void MeshDistributor::synchVector(SystemVector &vec)
  {
    int nComponents = vec.getSize();
    StdMpi<vector<double> > stdMpi(mpiComm);

    for (RankToDofContainer::iterator sendIt = sendDofs.begin();
	 sendIt != sendDofs.end(); ++sendIt) {
      vector<double> dofs;
      int nSendDofs = sendIt->second.size();
      dofs.reserve(nComponents * nSendDofs);
      
      for (int i = 0; i < nComponents; i++) {
	DOFVector<double> *dofvec = vec.getDOFVector(i);
	for (int j = 0; j < nSendDofs; j++)
	  dofs.push_back((*dofvec)[*((sendIt->second)[j])]);
      }

      stdMpi.send(sendIt->first, dofs);
    }

    for (RankToDofContainer::iterator recvIt = recvDofs.begin();
	 recvIt != recvDofs.end(); ++recvIt)
      stdMpi.recv(recvIt->first, recvIt->second.size() * nComponents);

    stdMpi.startCommunication<double>(MPI_DOUBLE);

    for (RankToDofContainer::iterator recvIt = recvDofs.begin();
	 recvIt != recvDofs.end(); ++recvIt) {
      int nRecvDofs = recvIt->second.size();

      int counter = 0;
      for (int i = 0; i < nComponents; i++) {
	DOFVector<double> *dofvec = vec.getDOFVector(i);
 	for (int j = 0; j < nRecvDofs; j++)
	  (*dofvec)[*(recvIt->second)[j]] = 
	    stdMpi.getRecvData(recvIt->first)[counter++];
      }
    }
  }


  void MeshDistributor::setRankDofs()
  {
    for (unsigned int i = 0; i < probStat.size(); i++) {
      int nComponents = probStat[i]->getNumComponents();
      for (int j = 0; j < nComponents; j++) {
	for (int k = 0; k < nComponents; k++)
	  if (probStat[i]->getSystemMatrix(j, k))
	    probStat[i]->getSystemMatrix(j, k)->setRankDofs(isRankDof);

	TEST_EXIT_DBG(probStat[i]->getRhs()->getDOFVector(j))("No RHS vector!\n");
	TEST_EXIT_DBG(probStat[i]->getSolution()->getDOFVector(j))("No solution vector!\n");
	
	probStat[i]->getRhs()->getDOFVector(j)->setRankDofs(isRankDof);
	probStat[i]->getSolution()->getDOFVector(j)->setRankDofs(isRankDof);
      }
    }
  }


  void MeshDistributor::removePeriodicBoundaryConditions()
  {
    FUNCNAME("MeshDistributor::removePeriodicBoundaryConditions()");

    // Remove periodic boundaries in boundary manager on matrices and vectors.
    for (unsigned int i = 0; i < probStat.size(); i++) {
      int nComponents = probStat[i]->getNumComponents();

      for (int j = 0; j < nComponents; j++) {
	for (int k = 0; k < nComponents; k++) {
	  DOFMatrix* mat = probStat[i]->getSystemMatrix(j, k);
	  if (mat && mat->getBoundaryManager())
	    removePeriodicBoundaryConditions(const_cast<BoundaryIndexMap&>(mat->getBoundaryManager()->getBoundaryConditionMap()));
	}
	
	if (probStat[i]->getSolution()->getDOFVector(j)->getBoundaryManager())
	  removePeriodicBoundaryConditions(const_cast<BoundaryIndexMap&>(probStat[i]->getSolution()->getDOFVector(j)->getBoundaryManager()->getBoundaryConditionMap()));
	
	if (probStat[i]->getRhs()->getDOFVector(j)->getBoundaryManager())
	  removePeriodicBoundaryConditions(const_cast<BoundaryIndexMap&>(probStat[i]->getRhs()->getDOFVector(j)->getBoundaryManager()->getBoundaryConditionMap()));
      }
    }

    // Remove periodic boundaries on elements in mesh.
    TraverseStack stack;
    ElInfo *elInfo = stack.traverseFirst(mesh,  -1, Mesh::CALL_EVERY_EL_PREORDER);
    while (elInfo) {
      elInfo->getElement()->deleteElementData(PERIODIC);
      elInfo = stack.traverseNext(elInfo);
    }    

    // Remove periodic vertex associations
    mesh->getPeriodicAssociations().clear();
  }


  void MeshDistributor::removePeriodicBoundaryConditions(BoundaryIndexMap& boundaryMap)
  {
    BoundaryIndexMap::iterator it = boundaryMap.begin();
    while (it != boundaryMap.end()) {
      if (it->second->isPeriodic())
	boundaryMap.erase(it++);
      else
	++it;      
    }    
  }


  void MeshDistributor::checkMeshChange()
  {
    FUNCNAME("MeshDistributor::checkMeshChange()");    

    double first = MPI::Wtime();

    // === If mesh has not been changed on all ranks, return. ===

    int recvAllValues = 0;
    int sendValue = static_cast<int>(mesh->getChangeIndex() != lastMeshChangeIndex);
    mpiComm.Allreduce(&sendValue, &recvAllValues, 1, MPI_INT, MPI_SUM);

    if (recvAllValues == 0)
      return;

    // === At least one rank mesh has been changed, so the boundaries must be ===
    // === adapted to the new mesh structure.                                 ===
   
    do {
      // To check the interior boundaries, the ownership of the boundaries is not 
      // important. Therefore, we add all boundaries to one boundary container.
      RankToBoundMap allBound;

      for (InteriorBoundary::iterator it(myIntBoundary); !it.end(); ++it)
	if ((mesh->getDim() == 2 && it->rankObj.subObj == EDGE) || 
	    (mesh->getDim() == 3 && it->rankObj.subObj == FACE))
 	  allBound[it.getRank()].push_back(*it);

      for (InteriorBoundary::iterator it(otherIntBoundary); !it.end(); ++it)
	if ((mesh->getDim() == 2 && it->rankObj.subObj == EDGE) || 
	    (mesh->getDim() == 3 && it->rankObj.subObj == FACE))
	  allBound[it.getRank()].push_back(*it);

      for (InteriorBoundary::iterator it(periodicBoundary); !it.end(); ++it) {
	if (it.getRank() == mpiRank) {
	  WARNING("Na, du weisst schon!\n");
	} else {
	  if ((mesh->getDim() == 2 && it->rankObj.subObj == EDGE) || 
	      (mesh->getDim() == 3 && it->rankObj.subObj == FACE))
	    allBound[it.getRank()].push_back(*it);	
	}
      }


      // === Check the boundaries and adapt mesh if necessary. ===
#if (DEBUG != 0)
      MSG("Run checkAndAdaptBoundary ...\n");
#endif

      bool meshChanged = checkAndAdaptBoundary(allBound);

      // === Check on all ranks if at least one rank's mesh has changed. ===

      int sendValue = static_cast<int>(!meshChanged);
      recvAllValues = 0;
      mpiComm.Allreduce(&sendValue, &recvAllValues, 1, MPI_INT, MPI_SUM);

#if (DEBUG != 0)
      MSG("Mesh changed on %d ranks!\n", recvAllValues);
#endif
    } while (recvAllValues != 0);

#if (DEBUG != 0)
    debug::writeMesh(feSpace, -1, debugOutputDir + "mesh");
#endif

    // === Because the mesh has been changed, update the DOF numbering and mappings. ===

    updateLocalGlobalNumbering();


    // === Update periodic mapping, if there are periodic boundaries. ===

    createPeriodicMap();

    INFO(info, 8)("Parallel mesh adaption needed %.5f seconds\n", 
		  MPI::Wtime() - first);


    // === The mesh has changed, so check if it is required to repartition the mesh. ===

    nTimestepsAfterLastRepartitioning++;

    if (repartitioningAllowed) {
      if (nTimestepsAfterLastRepartitioning >= repartitionIthChange) {
	repartitionMesh();
	nTimestepsAfterLastRepartitioning = 0;
      }
    }
  }

  
  bool MeshDistributor::checkAndAdaptBoundary(RankToBoundMap &allBound)
  {
    FUNCNAME("MeshDistributor::checkAndAdaptBoundary()");

    // === Create mesh structure codes for all ranks boundary elements. ===
       
    map<int, MeshCodeVec> sendCodes;
   
    for (RankToBoundMap::iterator it = allBound.begin(); it != allBound.end(); ++it) {

      for (vector<AtomicBoundary>::iterator boundIt = it->second.begin();
	   boundIt != it->second.end(); ++boundIt) {
	MeshStructure elCode;
	elCode.init(boundIt->rankObj);
	sendCodes[it->first].push_back(elCode);
      }
    }

    StdMpi<MeshCodeVec> stdMpi(mpiComm, true);
    stdMpi.send(sendCodes);
    stdMpi.recv(allBound);
    stdMpi.startCommunication<uint64_t>(MPI_UNSIGNED_LONG);
 
    // === Compare received mesh structure codes. ===
    
    bool meshFitTogether = true;

    for (RankToBoundMap::iterator it = allBound.begin(); it != allBound.end(); ++it) {
     
      MeshCodeVec &recvCodes = stdMpi.getRecvData()[it->first];
      int i = 0;
      
      for (vector<AtomicBoundary>::iterator boundIt = it->second.begin();
	   boundIt != it->second.end(); ++boundIt) {

	MeshStructure elCode;	
	elCode.init(boundIt->rankObj);

	if (elCode.getCode() != recvCodes[i].getCode()) {
	  TEST_EXIT_DBG(refineManager)("Refinement manager is not set correctly!\n");

	  bool b = startFitElementToMeshCode(recvCodes[i], 
					     boundIt->rankObj.el,
					     boundIt->rankObj.subObj,
					     boundIt->rankObj.ithObj, 
					     boundIt->rankObj.elType,
					     boundIt->rankObj.reverseMode);

	  if (b)
	    meshFitTogether = false;	  
 	}

	i++;
      }
    }

    return meshFitTogether;
  }


  bool MeshDistributor::startFitElementToMeshCode(MeshStructure &code, 
						  Element *el, 
						  GeoIndex subObj,
						  int ithObj, 
						  int elType,
						  bool reverseMode)
  {
    FUNCNAME("MeshDistributor::startFitElementToMeshCode()");

    TEST_EXIT_DBG(el)("No element given!\n");

    // If the code is empty, the element does not matter and the function can
    // return without chaning the mesh.
    if (code.empty())
      return false;

    // s0 and s1 are the number of the edge/face in both child of the element,
    // which contain the edge/face the function has to traverse through. If the
    // edge/face is not contained in one of the children, s0 or s1 is -1.
    int s0 = el->getSubObjOfChild(0, subObj, ithObj, elType);
    int s1 = el->getSubObjOfChild(1, subObj, ithObj, elType);

    TEST_EXIT_DBG(s0 != -1 || s1 != -1)("This should not happen!\n");

    bool meshChanged = false;
    Flag traverseFlag = 
      Mesh::CALL_EVERY_EL_PREORDER | Mesh::FILL_NEIGH | Mesh::FILL_BOUND;

    // Test for reverse mode, in which the left and right children of elements
    // are flipped.
    if (reverseMode)
      traverseFlag |= Mesh::CALL_REVERSE_MODE;    


    // === If the edge/face is contained in both children. ===

    if (s0 != -1 && s1 != -1) {
      // Create traverse stack and traverse within the mesh until the element,
      // which should be fitted to the mesh structure code, is reached.
      TraverseStack stack;
      ElInfo *elInfo = stack.traverseFirst(el->getMesh(), -1, traverseFlag);
      while (elInfo && elInfo->getElement() != el)
	elInfo = stack.traverseNext(elInfo);      

      TEST_EXIT_DBG(elInfo->getElement() == el)("This should not happen!\n");

      meshChanged = fitElementToMeshCode(code, stack, subObj, ithObj, reverseMode);
      return meshChanged;
    }


    // === The edge/face is contained in only on of the both children. ===

    if (el->isLeaf()) {

      // If element is leaf and code contains only one leaf element, we are finished.
      if (code.getNumElements() == 1 && code.isLeafElement())
	return false;     

      // Create traverse stack and traverse the mesh to the element.
      TraverseStack stack;
      ElInfo *elInfo = stack.traverseFirst(el->getMesh(), -1, traverseFlag);
      while (elInfo && elInfo->getElement() != el)
	elInfo = stack.traverseNext(elInfo);      

      TEST_EXIT_DBG(elInfo)("This should not happen!\n");

      // Code is not leaf, therefore refine the element.
      el->setMark(1);
      refineManager->setMesh(el->getMesh());
      refineManager->setStack(&stack);
      refineManager->refineFunction(elInfo);
      meshChanged = true;
    }

    Element *child0 = el->getFirstChild();
    Element *child1 = el->getSecondChild();
    if (reverseMode) {
      swap(s0, s1);
      swap(child0, child1);    
    }

    // === We know that the edge/face is contained in only one of the children. ===
    // === Therefore, traverse the mesh to this children and fit this element   ===
    // === To the mesh structure code.                                          ===

    TraverseStack stack;
    ElInfo *elInfo = stack.traverseFirst(el->getMesh(), -1, traverseFlag);

    if (s0 != -1) {
      while (elInfo && elInfo->getElement() != child0)
	elInfo = stack.traverseNext(elInfo);     

      meshChanged |= fitElementToMeshCode(code, stack, subObj, s0, reverseMode);
    } else {
      while (elInfo && elInfo->getElement() != child1) 
	elInfo = stack.traverseNext(elInfo);      

      meshChanged |= fitElementToMeshCode(code, stack, subObj, s1, reverseMode);
    }


    return meshChanged;
  }


  bool MeshDistributor::fitElementToMeshCode(MeshStructure &code, 
					     TraverseStack &stack,
					     GeoIndex subObj,
					     int ithObj, 
					     bool reverseMode)
  {
    FUNCNAME("MeshDistributor::fitElementToMeshCode()");


    // === Test if there are more elements in stack to check with the code. ===

    ElInfo *elInfo = stack.getElInfo();
    if (!elInfo)
      return false;


    // === Test if code contains a leaf element. If this is the case, the ===
    // === current element is finished. Traverse the mesh to the next     ===
    // === coarser element.                                               ===

    if (code.isLeafElement()) {
      int level = elInfo->getLevel();

      do {
	elInfo = stack.traverseNext(elInfo);
      } while (elInfo && elInfo->getLevel() > level);

      return false;
    }


    bool meshChanged = false;
    Element *el = elInfo->getElement();


    // === If element is leaf (and the code is not), refine the element. ===

    if (el->isLeaf()) {
      TEST_EXIT_DBG(elInfo->getLevel() < 255)("This should not happen!\n");

      el->setMark(1);
      refineManager->setMesh(el->getMesh());
      refineManager->setStack(&stack);
      refineManager->refineFunction(elInfo);
      meshChanged = true;
    }


    // === Continue fitting the mesh structure code to the children of the ===
    // === current element.                                                ===

    int s0 = el->getSubObjOfChild(0, subObj, ithObj, elInfo->getType());
    int s1 = el->getSubObjOfChild(1, subObj, ithObj, elInfo->getType());
    Element *child0 = el->getFirstChild();
    Element *child1 = el->getSecondChild();
    if (reverseMode) {
      swap(s0, s1);
      swap(child0, child1);
    }

    
    // === Traverse left child. ===

    if (s0 != -1) {
      // The edge/face is contained in the left child, therefore fit this
      // child to the mesh structure code.

      stack.traverseNext(elInfo);
      code.nextElement();
      meshChanged |= fitElementToMeshCode(code, stack, subObj, s0, reverseMode);
      elInfo = stack.getElInfo();
    } else {
      // The edge/face is not contained in the left child. Hence we need
      // to traverse through all subelements of the left child until we get
      // the second child of the current element.

      do {
	elInfo = stack.traverseNext(elInfo);
      } while (elInfo && elInfo->getElement() != child1); 

      TEST_EXIT_DBG(elInfo != NULL)("This should not happen!\n");
    }  

    TEST_EXIT_DBG(elInfo->getElement() == child1)
      ("Got wrong child with idx = %d! Searched for child idx = %d\n",
       elInfo->getElement()->getIndex(), child1->getIndex());


    // === Traverse right child. ===

    if (s1 != -1) {
      // The edge/face is contained in the right child, therefore fit this
      // child to the mesh structure code.

      code.nextElement();
      meshChanged |= fitElementToMeshCode(code, stack, subObj, s1, reverseMode);
    } else {
      // The edge/face is not contained in the right child. Hence we need
      // to traverse through all subelements of the right child until we have
      // finished traversing the current element with all its subelements.

      int level = elInfo->getLevel();

      do {
	elInfo = stack.traverseNext(elInfo);
      } while (elInfo && elInfo->getLevel() > level);
    }


    return meshChanged;
  }

  
  void MeshDistributor::serialize(ostream &out, DofContainer &data)
  {    
    int vecSize = data.size();
    SerUtil::serialize(out, vecSize);
    for (int i = 0; i < vecSize; i++) {
      int dofIndex = *(data[i]);
      SerUtil::serialize(out, dofIndex);
    }
  }


  void MeshDistributor::deserialize(istream &in, DofContainer &data,
				    map<int, const DegreeOfFreedom*> &dofMap)
  {
    FUNCNAME("MeshDistributor::deserialize()");

    int vecSize = 0;
    SerUtil::deserialize(in, vecSize);
    data.clear();
    data.resize(vecSize);
    for (int i = 0; i < vecSize; i++) {
      int dofIndex = 0;
      SerUtil::deserialize(in, dofIndex);

      TEST_EXIT_DBG(dofMap.count(dofIndex) != 0)
	("Dof index could not be deserialized correctly!\n");

      data[i] = dofMap[dofIndex];
    }
  }


  void MeshDistributor::serialize(ostream &out, RankToDofContainer &data)
  {
    int mapSize = data.size();
    SerUtil::serialize(out, mapSize);
    for (RankToDofContainer::iterator it = data.begin(); it != data.end(); ++it) {
      int rank = it->first;
      SerUtil::serialize(out, rank);
      serialize(out, it->second);
    }
  }

  
  void MeshDistributor::deserialize(istream &in, RankToDofContainer &data,
				    map<int, const DegreeOfFreedom*> &dofMap)
  {
    data.clear();

    int mapSize = 0;
    SerUtil::deserialize(in, mapSize);
    for (int i = 0; i < mapSize; i++) {
      int rank = 0;
      SerUtil::deserialize(in, rank);
      deserialize(in, data[rank], dofMap);      
    }
  }


  void MeshDistributor::setInitialElementWeights() 
  {
    FUNCNAME("MeshDistributor::setInitialElementWeights()");

    elemWeights.clear();
      
    string filename = "";
    GET_PARAMETER(0, mesh->getName() + "->macro weights", &filename);
    if (filename != "") {
      MSG("Read macro weights from %s\n", filename.c_str());

      ifstream infile;
      infile.open(filename.c_str(), ifstream::in);
      while (!infile.eof()) {
	int elNum, elWeight;
	infile >> elNum;
	if (infile.eof())
	  break;
	infile >> elWeight;

	elemWeights[elNum] = elWeight;
      }

      infile.close();
    } else {           
      TraverseStack stack;
      ElInfo *elInfo = stack.traverseFirst(mesh, -1, Mesh::CALL_LEAF_EL);
      while (elInfo) {
	elemWeights[elInfo->getElement()->getIndex()] = 1.0;
	elInfo = stack.traverseNext(elInfo);
      }
    }
  }


  void MeshDistributor::repartitionMesh()
  {
    FUNCNAME("MeshDistributor::repartitionMesh()");

    TEST_EXIT(mesh->getNumberOfDOFAdmin() == 1)
      ("Only meshes with one DOFAdmin are supported!\n");

    int nDofs = mesh->getDofAdmin(0).getUsedDofs();
    int repartitioning = 0;

    vector<int> nDofsInRank(mpiSize);
    mpiComm.Gather(&nDofs, 1, MPI_INT, &(nDofsInRank[0]), 1, MPI_INT, 0);

    if (mpiRank == 0) {
      int nOverallDofs = 0;
      int minDofs = numeric_limits<int>::max();
      int maxDofs = numeric_limits<int>::min();
      for (int i = 0; i < mpiSize; i++) {
	nOverallDofs += nDofsInRank[i];
	minDofs = std::min(minDofs, nDofsInRank[i]);
	maxDofs = std::max(maxDofs, nDofsInRank[i]);
      }      
     
      MSG("Overall DOFs: %d    Min DOFs: %d    Max DOFs: %d\n", 
	  nOverallDofs, minDofs, maxDofs);

      if (static_cast<double>(maxDofs) / static_cast<double>(minDofs)) 
	repartitioning = 1;

      mpiComm.Bcast(&repartitioning, 1, MPI_INT, 0);
    } else {
      mpiComm.Bcast(&repartitioning, 1, MPI_INT, 0);
    }


    if (repartitioning == 0)
      return;

    double timePoint = MPI::Wtime();

#if (DEBUG != 0)
    ParallelDebug::testDoubleDofs(mesh);

    if (repartCounter == 0) {        
      stringstream oss;
      oss << debugOutputDir << "partitioning-" << repartCounter << ".vtu";
      MSG("Write partitioning to %s\n", oss.str().c_str());
      DOFVector<double> tmpa(feSpace, "tmp");
      tmpa.set(mpiRank);
      VtkWriter::writeFile(&tmpa, oss.str());
      
      repartCounter++;
    }
#endif


    elemWeights.clear();

    TraverseStack stack;
    ElInfo *elInfo = stack.traverseFirst(mesh, -1, Mesh::CALL_LEAF_EL);
    while (elInfo) {
      elemWeights[elInfo->getMacroElement()->getIndex()]++;      
      elInfo = stack.traverseNext(elInfo);
    }

    partitioner->useLocalGlobalDofMap(&mapLocalGlobalDofs);
    bool partitioningSucceed =
      partitioner->partition(elemWeights, ADAPTIVE_REPART, 1000.0);

    if (!partitioningSucceed) {
      MSG("ParMETIS created empty partition!\n");
      return;
    }

    oldPartitionVec = partitionVec;


    // === Create map that maps macro element indices to pointers to the ===
    // === macro elements.                                               ===

    map<int, MacroElement*> elIndexMap;
    for (vector<MacroElement*>::iterator it = allMacroElements.begin();
	 it != allMacroElements.end(); ++it)
      elIndexMap[(*it)->getIndex()] = *it;


    // === Create set of all new macro elements this rank will receive from ===
    // === other ranks.                                                     ===

    std::set<MacroElement*> newMacroEl;
    for (map<int, vector<int> >::iterator it = partitioner->getRecvElements().begin();
	 it != partitioner->getRecvElements().end(); ++it) {
      for (vector<int>::iterator elIt = it->second.begin();
	   elIt != it->second.end(); ++elIt) {
	TEST_EXIT_DBG(elIndexMap.count(*elIt) == 1)
	  ("Could not find macro element %d\n", *elIt);

	newMacroEl.insert(elIndexMap[*elIt]);
      }
    }

    std::set<MacroElement*> allMacroEl;
    for (std::set<MacroElement*>::iterator it = newMacroEl.begin(); 
	 it != newMacroEl.end(); ++it)
      allMacroEl.insert(*it);

    for (deque<MacroElement*>::iterator it = mesh->firstMacroElement();
	 it != mesh->endOfMacroElements(); ++it)
      allMacroEl.insert(*it);


    // === Add new macro elements to mesh. ===

    for (std::set<MacroElement*>::iterator it = newMacroEl.begin();
	 it != newMacroEl.end(); ++it) {
      MacroElement *mel = *it;

      // First, reset all neighbour relations. The correct neighbours will be set later.
      for (int i = 0; i < mesh->getGeo(NEIGH); i++)
	mel->setNeighbour(i, NULL);

      // Create new DOFs for the macro element.
      for (int i = 0; i < mesh->getGeo(VERTEX); i++)
	mel->getElement()->setDof(i, mesh->getDof(VERTEX));

      // The macro element now has valid DOF pointers.
      mel->getElement()->setDofValid(true);

      // Push the macro element to all macro elements in mesh.
      mesh->getMacroElements().push_back(mel);
    }


    // === Send and receive mesh structure codes. ===

    map<int, MeshCodeVec> sendCodes;
    map<int, vector<vector<double> > > sendValues;

    for (map<int, vector<int> >::iterator it = partitioner->getSendElements().begin();
	 it != partitioner->getSendElements().end(); ++it) {
      for (vector<int>::iterator elIt = it->second.begin();
	   elIt != it->second.end(); ++elIt) {
	MeshStructure elCode;
	elCode.init(mesh, *elIt);
	sendCodes[it->first].push_back(elCode);

	for (unsigned int i = 0; i < testVec.size(); i++) {
	  vector<double> valVec;
	  elCode.getMeshStructureValues(mesh, *elIt, testVec[i], valVec);
	  sendValues[it->first].push_back(valVec);
	}
      }
    }

    StdMpi<MeshCodeVec> stdMpi(mpiComm, true);
    stdMpi.send(sendCodes);
    stdMpi.recv(partitioner->getRecvElements());
    stdMpi.startCommunication<uint64_t>(MPI_UNSIGNED_LONG);

    StdMpi<vector<vector<double> > > stdMpi2(mpiComm, true);
    stdMpi2.send(sendValues);
    stdMpi2.recv(partitioner->getRecvElements());
    stdMpi2.startCommunication<double>(MPI_DOUBLE);

    // === Adapte the new macro elements due to the received mesh structure codes. ===

    for (map<int, vector<int> >::iterator it = partitioner->getRecvElements().begin();
	 it != partitioner->getRecvElements().end(); ++it) {
      MeshCodeVec &recvCodes = stdMpi.getRecvData()[it->first];
      vector<vector<double> > &recvValues = stdMpi2.getRecvData()[it->first];
    int i = 0;
    int j = 0;

      TEST_EXIT_DBG(recvCodes.size() == it->second.size())
	("Should not happen!\n");

      for (vector<int>::iterator elIt = it->second.begin();
	   elIt != it->second.end(); ++elIt) {
	recvCodes[i].fitMeshToStructure(mesh, refineManager, false, *elIt);

	for (unsigned int k = 0; k < testVec.size(); k++)
	  recvCodes[i].setMeshStructureValues(mesh, *elIt, testVec[k], recvValues[j++]);    
	
	i++;
      }
    }


    // === Set correct neighbour information on macro elements. ===

    for (std::set<MacroElement* , bool(*)(MacroElement*, MacroElement*)>::iterator it = allMacroEl.begin();
	 it != allMacroEl.end(); ++it) {
      int elIndex = (*it)->getIndex();
      for (int i = 0; i < mesh->getGeo(NEIGH); i++) {

	TEST_EXIT_DBG(macroElementNeighbours.count(elIndex) == 1)
	  ("Should not happen!\n");
	TEST_EXIT_DBG(static_cast<int>(macroElementNeighbours[elIndex].size()) == 
		      mesh->getGeo(NEIGH))
	  ("Should not happen!\n");

	int neighIndex = macroElementNeighbours[elIndex][i];

	if (neighIndex == -1 || 
	    partitioner->getElementInRank()[neighIndex] == false) {
	  (*it)->setNeighbour(i, NULL);
	} else {
	  TEST_EXIT_DBG(elIndexMap.count(neighIndex) == 1)("Should not happen!\n");

	  (*it)->setNeighbour(i, elIndexMap[neighIndex]);
	}
      }
    }


    // === Delete macros from mesh. ===

    std::set<MacroElement*> deleteMacroElements;

    for (map<int, vector<int> >::iterator it = partitioner->getSendElements().begin();
	 it != partitioner->getSendElements().end(); ++it) {
      for (vector<int>::iterator elIt = it->second.begin();
	   elIt != it->second.end(); ++elIt) {
	TEST_EXIT_DBG(elIndexMap.count(*elIt) == 1)
	  ("Could not find macro element %d\n", *elIt);

	deleteMacroElements.insert(elIndexMap[*elIt]);
      }
    }


    // Note that also if there are no macros to be deleted, this function will
    // update the number of elements, vertices, etc. of the mesh.
    mesh->removeMacroElements(deleteMacroElements, feSpace);


    // === Remove double DOFs. ===

    MeshManipulation meshManipulation(feSpace);
    meshManipulation.deleteDoubleDofs(newMacroEl, elObjects);

    mesh->dofCompress();

#if (DEBUG != 0)
    stringstream oss;
    oss << debugOutputDir << "partitioning-" << repartCounter << ".vtu";    
    DOFVector<double> tmpa(feSpace, "tmp");
    tmpa.set(mpiRank);
    VtkWriter::writeFile(&tmpa, oss.str());
    repartCounter++;

    ParallelDebug::testAllElements(*this);
    ParallelDebug::testDoubleDofs(mesh);
#endif

    partitioner->fillCoarsePartitionVec(&partitionVec);

    updateInteriorBoundaryInfo();

#if (DEBUG != 0)
    ParallelDebug::printBoundaryInfo(*this);
#endif

    updateLocalGlobalNumbering();

#if (DEBUG != 0)
    MSG("AMDiS runs in debug mode, so make some test ...\n");

    ParallelDebug::testAllElements(*this);
    ParallelDebug::testInteriorBoundary(*this);
    ParallelDebug::printBoundaryInfo(*this);

    MSG("Debug mode tests finished!\n");
#endif

    MSG("Mesh repartitioning needed %.5f seconds\n", MPI::Wtime() - timePoint);
  }


  void MeshDistributor::createInteriorBoundaryInfo()
  {
    FUNCNAME("MeshDistributor::createInteriorBoundaryInfo()");

    createMeshElementData();
    createBoundaryData();
  }


  void MeshDistributor::updateInteriorBoundaryInfo()
  {
    FUNCNAME("MeshDistributor::updateInteriorBoundaryInfo()");

    elObjects.createRankData(partitionVec);
    createBoundaryData();
  }


  void MeshDistributor::createMeshElementData()
  {
    FUNCNAME("MeshDistributor::createMeshElementData()");


   // === Fills macro element data structures. ===

    TraverseStack stack;
    ElInfo *elInfo = 
      stack.traverseFirst(mesh, -1, 
			  Mesh::CALL_LEAF_EL | Mesh::FILL_NEIGH | Mesh::FILL_BOUND);
    while (elInfo) {
      TEST_EXIT_DBG(elInfo->getLevel() == 0)("Should not happen!\n");

      Element *el = elInfo->getElement();
      macroElIndexMap[el->getIndex()] = el;
      macroElIndexTypeMap[el->getIndex()] = elInfo->getType();
     
      // === Add all sub object of the element to the variable elObjects. ===
      elObjects.addElement(elInfo);

      elInfo = stack.traverseNext(elInfo);
    }

    // === Create periodic data, if there are periodic boundary conditions. ===
    elObjects.createPeriodicData();


    // === Create mesh element data for this rank. ===
    elObjects.createRankData(partitionVec);
  }


  void MeshDistributor::createBoundaryData()
  {
    FUNCNAME("MeshDistributor::createBoundaryData()");


    // === Clear all relevant data structures. ===

    myIntBoundary.clear();
    otherIntBoundary.clear();
    periodicBoundary.clear();


    // === Create interior boundary data structure. ===

    for (int geoPos = 0; geoPos < mesh->getDim(); geoPos++) {
      GeoIndex geoIndex = INDEX_OF_DIM(geoPos, mesh->getDim());

      while (elObjects.iterate(geoIndex)) {
	map<int, ElementObjectData>& objData = elObjects.getIterateData();
	if (objData.count(mpiRank) && objData.size() > 1) {
	  int owner = elObjects.getIterateOwner();
	  ElementObjectData& rankBoundEl = objData[mpiRank];

	  TEST_EXIT_DBG(macroElIndexMap[rankBoundEl.elIndex])("Should not happen!\n");

	  AtomicBoundary bound;	    	    
	  bound.rankObj.el = macroElIndexMap[rankBoundEl.elIndex];
	  bound.rankObj.elIndex = rankBoundEl.elIndex;
	  bound.rankObj.elType = macroElIndexTypeMap[rankBoundEl.elIndex];
	  bound.rankObj.subObj = geoIndex;
	  bound.rankObj.ithObj = rankBoundEl.ithObject;

 	  if (geoIndex == FACE) {
 	    for (int edgeNo = 0; edgeNo < 3; edgeNo++) {
	      int edgeOfFace = 
		bound.rankObj.el->getEdgeOfFace(bound.rankObj.ithObj, edgeNo);

	      bound.rankObj.excludedSubstructures.push_back(make_pair(EDGE, edgeOfFace));
	    }
	  }
	  
	  
	  if (owner == mpiRank) {
	    for (map<int, ElementObjectData>::iterator it2 = objData.begin();
		 it2 != objData.end(); ++it2) {
	      if (it2->first == mpiRank)
		continue;
	      
	      bound.neighObj.el = macroElIndexMap[it2->second.elIndex];
	      bound.neighObj.elIndex = it2->second.elIndex;
	      bound.neighObj.elType = macroElIndexTypeMap[it2->second.elIndex];
	      bound.neighObj.subObj = geoIndex;
	      bound.neighObj.ithObj = it2->second.ithObject;

	      bound.type = INTERIOR;
	      
	      AtomicBoundary& b = myIntBoundary.getNewAtomic(it2->first);
	      b = bound;
	      b.rankObj.setReverseMode(b.neighObj, feSpace, bound.type);
	    }
	    
	  } else {
	    TEST_EXIT_DBG(objData.count(owner) == 1)
	      ("Should not happen!\n");
	    
	    ElementObjectData& ownerBoundEl = objData[owner];
	    
	    bound.neighObj.el = macroElIndexMap[ownerBoundEl.elIndex];
	    bound.neighObj.elIndex = ownerBoundEl.elIndex;
	    bound.neighObj.elType = -1;
	    bound.neighObj.subObj = geoIndex;
	    bound.neighObj.ithObj = ownerBoundEl.ithObject;

	    bound.type = INTERIOR;
	    
	    AtomicBoundary& b = otherIntBoundary.getNewAtomic(owner);
	    b = bound;	 
	    b.neighObj.setReverseMode(b.rankObj, feSpace, bound.type);
	  }
	}
      }
    }
    
    
    // === Create periodic boundary data structure. ===

    for (PerBoundMap<DegreeOfFreedom>::iterator it = elObjects.getPeriodicVertices().begin();
	 it != elObjects.getPeriodicVertices().end(); ++it) {
      if (elObjects.isInRank(it->first.first, mpiRank) == false)
	continue;

      WorldVector<double> c0, c1;
      mesh->getDofIndexCoords(it->first.first, feSpace, c0);
      mesh->getDofIndexCoords(it->first.second, feSpace, c1);

      //      MSG("CREATE BOUNDARY FOR DOF MAP: %d (%.3f %.3f %.3f)<-> %d (%.3f %.3f %.3f)\n", it->first.first, 
      // 	  c0[0], c0[1], c0[2], it->first.second, c1[0], c1[1], c1[2]);
      ElementObjectData& perDofEl0 = elObjects.getElementsInRank(it->first.first)[mpiRank];
      //      MSG("DATA: %d %d %d\n", perDofEl0.elIndex, VERTEX, perDofEl0.ithObject);

      for (map<int, ElementObjectData>::iterator elIt = elObjects.getElementsInRank(it->first.second).begin();
	   elIt != elObjects.getElementsInRank(it->first.second).end(); ++elIt) {

	int otherElementRank = elIt->first;
	ElementObjectData& perDofEl1 = elIt->second;

	AtomicBoundary bound;
	bound.rankObj.el = macroElIndexMap[perDofEl0.elIndex];
	bound.rankObj.elIndex = perDofEl0.elIndex;
	bound.rankObj.elType = macroElIndexTypeMap[perDofEl0.elIndex];
	bound.rankObj.subObj = VERTEX;
	bound.rankObj.ithObj = perDofEl0.ithObject;

	bound.neighObj.el = macroElIndexMap[perDofEl1.elIndex];
	bound.neighObj.elIndex = perDofEl1.elIndex;
	bound.neighObj.elType = macroElIndexTypeMap[perDofEl1.elIndex];
	bound.neighObj.subObj = VERTEX;
	bound.neighObj.ithObj = perDofEl1.ithObject;

	bound.type = it->second;

	AtomicBoundary& b = periodicBoundary.getNewAtomic(otherElementRank);
	b = bound;
      }
    }


    for (PerBoundMap<DofEdge>::iterator it = elObjects.getPeriodicEdges().begin();
	 it != elObjects.getPeriodicEdges().end(); ++it) {
      if (elObjects.isInRank(it->first.first, mpiRank) == false)
	continue;

      ElementObjectData& perEdgeEl0 = elObjects.getElementsInRank(it->first.first)[mpiRank];

      for (map<int, ElementObjectData>::iterator elIt = elObjects.getElementsInRank(it->first.second).begin();
 	   elIt != elObjects.getElementsInRank(it->first.second).end(); ++elIt) {
      
	int otherElementRank = elIt->first;
	ElementObjectData& perEdgeEl1 = elIt->second;

	AtomicBoundary bound;	    	    
	bound.rankObj.el = macroElIndexMap[perEdgeEl0.elIndex];
	bound.rankObj.elIndex = perEdgeEl0.elIndex;
	bound.rankObj.elType = macroElIndexTypeMap[perEdgeEl0.elIndex];
	bound.rankObj.subObj = EDGE;
	bound.rankObj.ithObj = perEdgeEl0.ithObject;
	
	bound.neighObj.el = macroElIndexMap[perEdgeEl1.elIndex];
	bound.neighObj.elIndex = perEdgeEl1.elIndex;
	bound.neighObj.elType = macroElIndexTypeMap[perEdgeEl1.elIndex];
	bound.neighObj.subObj = EDGE;
	bound.neighObj.ithObj = perEdgeEl1.ithObject;
	
	bound.type = it->second;
	
	AtomicBoundary& b = periodicBoundary.getNewAtomic(otherElementRank);
	b = bound;
     
	if (mpiRank > otherElementRank)
	  b.rankObj.setReverseMode(b.neighObj, feSpace, bound.type);
	else
	  b.neighObj.setReverseMode(b.rankObj, feSpace, bound.type);
      }
    }


    for (PerBoundMap<DofFace>::iterator it = elObjects.getPeriodicFaces().begin();
	 it != elObjects.getPeriodicFaces().end(); ++it) {
      if (elObjects.isInRank(it->first.first, mpiRank) == false)
	continue;

      TEST_EXIT_DBG(elObjects.getElements(it->first.first).size() == 1)
 	("Should not happen!\n");
      TEST_EXIT_DBG(elObjects.getElements(it->first.second).size() == 1)
 	("Should not happen!\n");

      ElementObjectData& perEdgeEl0 = elObjects.getElementsInRank(it->first.first)[mpiRank];

      for (map<int, ElementObjectData>::iterator elIt = elObjects.getElementsInRank(it->first.second).begin();
 	   elIt != elObjects.getElementsInRank(it->first.second).end(); ++elIt) {
      
	int otherElementRank = elIt->first;
	ElementObjectData& perEdgeEl1 = elIt->second;

	AtomicBoundary bound;	    	    
	bound.rankObj.el = macroElIndexMap[perEdgeEl0.elIndex];
	bound.rankObj.elIndex = perEdgeEl0.elIndex;
	bound.rankObj.elType = macroElIndexTypeMap[perEdgeEl0.elIndex];
	bound.rankObj.subObj = FACE;
	bound.rankObj.ithObj = perEdgeEl0.ithObject;
	
	bound.neighObj.el = macroElIndexMap[perEdgeEl1.elIndex];
	bound.neighObj.elIndex = perEdgeEl1.elIndex;
	bound.neighObj.elType = macroElIndexTypeMap[perEdgeEl1.elIndex];
	bound.neighObj.subObj = FACE;
	bound.neighObj.ithObj = perEdgeEl1.ithObject;
	
	bound.type = it->second;
	
	AtomicBoundary& b = periodicBoundary.getNewAtomic(otherElementRank);
	b = bound;
     
	if (mpiRank > otherElementRank)
	  b.rankObj.setReverseMode(b.neighObj, feSpace, bound.type);
	else
	  b.neighObj.setReverseMode(b.rankObj, feSpace, bound.type);
      }
    }
    

    // === Once we have this information, we must care about the order of the atomic ===
    // === bounds in the three boundary handling object. Eventually all the bound-   ===
    // === aries have to be in the same order on both ranks that share the bounday.  ===

    StdMpi<vector<AtomicBoundary> > stdMpi(mpiComm);
    stdMpi.send(myIntBoundary.boundary);
    stdMpi.recv(otherIntBoundary.boundary);
    stdMpi.startCommunication<int>(MPI_INT);


    // === The information about all neighbouring boundaries has been received. So ===
    // === the rank tests if its own atomic boundaries are in the same order. If   ===
    // === not, the atomic boundaries are swaped to the correct order.             ===

    for (RankToBoundMap::iterator rankIt = otherIntBoundary.boundary.begin();
	 rankIt != otherIntBoundary.boundary.end(); ++rankIt) {

      // === We have received from rank "rankIt->first" the ordered list of element ===
      // === indices. Now, we have to sort the corresponding list in this rank to   ===
      // === get the same order.                                                    ===
     
      for (unsigned int j = 0; j < rankIt->second.size(); j++) {

	// If the expected object is not at place, search for it.

	BoundaryObject &recvedBound = stdMpi.getRecvData()[rankIt->first][j].rankObj;

	if ((rankIt->second)[j].neighObj != recvedBound) {
	  unsigned int k = j + 1;

	  for (; k < rankIt->second.size(); k++)
 	    if ((rankIt->second)[k].neighObj == recvedBound)
	      break;

	  // The element must always be found, because the list is just in another order.
	  TEST_EXIT_DBG(k < rankIt->second.size())("Should never happen!\n");

	  // Swap the current with the found element.
	  AtomicBoundary tmpBound = (rankIt->second)[k];
	  (rankIt->second)[k] = (rankIt->second)[j];
	  (rankIt->second)[j] = tmpBound;	
	}
      }
    }


    // === Do the same for the periodic boundaries. ===

    if (periodicBoundary.boundary.size() > 0) {
      stdMpi.clear();

      InteriorBoundary sendBounds, recvBounds;     
      for (RankToBoundMap::iterator rankIt = periodicBoundary.boundary.begin();
	   rankIt != periodicBoundary.boundary.end(); ++rankIt) {

	if (rankIt->first == mpiRank)
	  continue;

	if (rankIt->first < mpiRank)
	  sendBounds.boundary[rankIt->first] = rankIt->second;
	else
	  recvBounds.boundary[rankIt->first] = rankIt->second;	
      }

      stdMpi.send(sendBounds.boundary);
      stdMpi.recv(recvBounds.boundary);
      stdMpi.startCommunication<int>(MPI_INT);

      /*
      for (RankToBoundMap::iterator rankIt = periodicBoundary.boundary.begin();
	   rankIt != periodicBoundary.boundary.end(); ++rankIt) {

	for (unsigned int j = 0; j < rankIt->second.size(); j++) {

 	  MSG("%-ith boundary with rank %d\n", j, rankIt->first);
 	  MSG("  bound: el = %d   subobj = %d   ithobj = %d\n",
 	      periodicBoundary.boundary[rankIt->first][j].rankObj.elIndex,
 	      periodicBoundary.boundary[rankIt->first][j].rankObj.subObj,
 	      periodicBoundary.boundary[rankIt->first][j].rankObj.ithObj);
 	  MSG("  neigh: el = %d   subobj = %d   ithobj = %d\n",
	      periodicBoundary.boundary[rankIt->first][j].neighObj.elIndex,
 	      periodicBoundary.boundary[rankIt->first][j].neighObj.subObj,
 	      periodicBoundary.boundary[rankIt->first][j].neighObj.ithObj);
	}
      }
      
      */

      for (RankToBoundMap::iterator rankIt = periodicBoundary.boundary.begin();
	   rankIt != periodicBoundary.boundary.end(); ++rankIt) {

 	if (rankIt->first <= mpiRank)
 	  continue;
  
	for (unsigned int j = 0; j < rankIt->second.size(); j++) {
	  BoundaryObject &recvRankObj = stdMpi.getRecvData()[rankIt->first][j].rankObj;
	  BoundaryObject &recvNeighObj = stdMpi.getRecvData()[rankIt->first][j].neighObj;

	  if (periodicBoundary.boundary[rankIt->first][j].neighObj != recvRankObj ||
	      periodicBoundary.boundary[rankIt->first][j].rankObj != recvNeighObj) {
	    unsigned int k = j + 1;	    
	    for (; k < rankIt->second.size(); k++)
	      if (periodicBoundary.boundary[rankIt->first][k].neighObj == recvRankObj &&
		  periodicBoundary.boundary[rankIt->first][k].rankObj == recvNeighObj)
		break;
	    
	    // The element must always be found, because the list is just in 
	    // another order.
	    TEST_EXIT_DBG(k < rankIt->second.size())("Should never happen!\n");

	    // Swap the current with the found element.
	    AtomicBoundary tmpBound = (rankIt->second)[k];
	    (rankIt->second)[k] = (rankIt->second)[j];
	    (rankIt->second)[j] = tmpBound;	
	  } 
	}
      }     
    } // periodicBoundary.boundary.size() > 0
  }


  void MeshDistributor::removeMacroElements()
  {
    FUNCNAME("MeshDistributor::removeMacroElements()");

    std::set<MacroElement*> macrosToRemove;
    for (deque<MacroElement*>::iterator it = mesh->firstMacroElement();
	 it != mesh->endOfMacroElements(); ++it)
      if (partitioner->getElementInRank()[(*it)->getIndex()] == false)
	macrosToRemove.insert(*it);    

    mesh->removeMacroElements(macrosToRemove, feSpace);
  }


  void MeshDistributor::updateLocalGlobalNumbering()
  {
    FUNCNAME("MeshDistributor::updateLocalGlobalNumbering()");

    mesh->dofCompress();

#if (DEBUG != 0)
    debug::ElementIdxToDofs elMap;
    debug::createSortedDofs(mesh, elMap);   
#endif

    typedef std::set<const DegreeOfFreedom*> DofSet;

    // === Get all DOFs in ranks partition. ===

    ElementDofIterator elDofIt(feSpace);
    DofSet rankDofSet;
    
    TraverseStack stack;
    ElInfo *elInfo = stack.traverseFirst(mesh, -1, Mesh::CALL_LEAF_EL);
    while (elInfo) {
      elDofIt.reset(elInfo->getElement());
      do {
	rankDofSet.insert(elDofIt.getDofPtr());
      } while(elDofIt.next());

      elInfo = stack.traverseNext(elInfo);
    }


    DofContainer rankDofs;
    for (DofSet::iterator it = rankDofSet.begin(); it != rankDofSet.end(); ++it)
      rankDofs.push_back(*it);    
    sort(rankDofs.begin(), rankDofs.end(), cmpDofsByValue);
    int nRankAllDofs = rankDofs.size();


    // === Traverse on interior boundaries and move all not ranked owned DOFs from ===
    // === rankDofs to boundaryDofs.                                               ===

    sendDofs.clear();
    recvDofs.clear();


    for (InteriorBoundary::iterator it(myIntBoundary); !it.end(); ++it) {
      DofContainer dofs;	
      it->rankObj.el->getVertexDofs(feSpace, it->rankObj, dofs);
      it->rankObj.el->getNonVertexDofs(feSpace, it->rankObj, dofs);

      for (unsigned int i = 0; i < dofs.size(); i++)
	sendDofs[it.getRank()].push_back(dofs[i]);
    }


    for (InteriorBoundary::iterator it(otherIntBoundary); !it.end(); ++it) {
      DofContainer dofs;
      it->rankObj.el->getVertexDofs(feSpace, it->rankObj, dofs);
      it->rankObj.el->getNonVertexDofs(feSpace, it->rankObj, dofs);

      for (unsigned int i = 0; i < dofs.size(); i++) {
	DofContainer::iterator eraseIt = 
	  find(rankDofs.begin(), rankDofs.end(), dofs[i]);
	if (eraseIt != rankDofs.end()) 
	  rankDofs.erase(eraseIt);

	recvDofs[it.getRank()].push_back(dofs[i]);
      }
    }


    nRankDofs = rankDofs.size();

    // === Get starting position for global rank dof ordering. ====

    mpiComm.Scan(&nRankDofs, &rstart, 1, MPI_INT, MPI_SUM);
    rstart -= nRankDofs;


    // === Calculate number of overall DOFs of all partitions. ===

    nOverallDofs = 0;
    mpiComm.Allreduce(&nRankDofs, &nOverallDofs, 1, MPI_INT, MPI_SUM);

    // First, we set all dofs in ranks partition to be owend by the rank. Later,
    // the dofs in ranks partition that are owned by other rank are set to false.
    isRankDof.clear();
    for (int i = 0; i < nRankAllDofs; i++)
      isRankDof[i] = true;    

    // Stores for all rank owned dofs a new global index.
    DofIndexMap rankDofsNewGlobalIndex;
    for (int i = 0; i < nRankDofs; i++)
      rankDofsNewGlobalIndex[rankDofs[i]] = i + rstart;


    // === Send new DOF indices. ===

#if (DEBUG != 0)
    ParallelDebug::testDofContainerCommunication(*this, sendDofs, recvDofs);
#endif
    
    int i = 0;
    StdMpi<vector<DegreeOfFreedom> > stdMpi(mpiComm, false);
    for (RankToDofContainer::iterator sendIt = sendDofs.begin();
	 sendIt != sendDofs.end(); ++sendIt, i++) {
      stdMpi.getSendData(sendIt->first).resize(0);
      stdMpi.getSendData(sendIt->first).reserve(sendIt->second.size());
      for (DofContainer::iterator dofIt = sendIt->second.begin();
	   dofIt != sendIt->second.end(); ++dofIt)
	stdMpi.getSendData(sendIt->first).push_back(rankDofsNewGlobalIndex[*dofIt]);
    }
    stdMpi.updateSendDataSize();
    stdMpi.recv(recvDofs);
    stdMpi.startCommunication<int>(MPI_INT);


    for (RankToDofContainer::iterator recvIt = recvDofs.begin();
	 recvIt != recvDofs.end(); ++recvIt) {      
      int j = 0;
      for (DofContainer::iterator dofIt = recvIt->second.begin();
	   dofIt != recvIt->second.end(); ++dofIt) {
	rankDofsNewGlobalIndex[*dofIt] = stdMpi.getRecvData(recvIt->first)[j++];
	isRankDof[**dofIt] = false;
      }
    }


    // === Create now the local to global index and local to dof index mappings.  ===

    mapLocalGlobalDofs.clear();
    mapLocalDofIndex.clear();

    for (DofIndexMap::iterator dofIt = rankDofsNewGlobalIndex.begin();
	 dofIt != rankDofsNewGlobalIndex.end(); ++dofIt)
      mapLocalGlobalDofs[*(dofIt->first)] = dofIt->second;   

    for (int i = 0; i < nRankDofs; i++)
      mapLocalDofIndex[i] = *(rankDofs[i]);    


    // === Update dof admins due to new number of dofs. ===
  
    lastMeshChangeIndex = mesh->getChangeIndex();


#if (DEBUG != 0)
    stringstream oss;
    oss << debugOutputDir << "elementIndex-" << mpiRank << ".vtu";

    debug::writeElementIndexMesh(mesh, oss.str());
    debug::testSortedDofs(mesh, elMap);
    ParallelDebug::testCommonDofs(*this, true);
    ParallelDebug::testGlobalIndexByCoords(*this);
    ParallelDebug::writeDebugFile(*this, debugOutputDir + "mpi-dbg", "dat");
    
    MSG("------------- Debug information -------------\n");
    MSG("nRankDofs = %d\n", nRankDofs);
    MSG("nOverallDofs = %d\n", nOverallDofs);
    MSG("rstart %d\n", rstart);
#endif
  }


  void MeshDistributor::createPeriodicMap()
  {
    FUNCNAME("MeshDistributor::createPeriodicMap()");

    if (periodicBoundary.boundary.size() == 0)
      return;

    // Clear all periodic dof mappings calculated before. We do it from scratch.
    periodicDof.clear();
    periodicDofAssociations.clear();

    StdMpi<vector<int> > stdMpi(mpiComm, false);

    // === Each rank traverse its periodic boundaries and sends the DOF indices ===
    // === to the rank "on the other side" of the periodic boundary.            ===

    RankToDofContainer rankPeriodicDofs;
    map<int, vector<int> > rankToDofType;

    for (RankToBoundMap::iterator it = periodicBoundary.boundary.begin();
	 it != periodicBoundary.boundary.end(); ++it) {

      if (it->first == mpiRank) {
	TEST_EXIT_DBG(it->second.size() % 2 == 0)("Should not happen!\n");

	for (unsigned int i = 0; i < it->second.size(); i++) {
	  AtomicBoundary &bound = it->second[i];
	  DofContainer dofs0, dofs1;
	  bound.rankObj.el->getVertexDofs(feSpace, bound.rankObj, dofs0);
	  bound.rankObj.el->getNonVertexDofs(feSpace, bound.rankObj, dofs0); 
	  bound.neighObj.el->getVertexDofs(feSpace, bound.neighObj, dofs1);
	  bound.neighObj.el->getNonVertexDofs(feSpace, bound.neighObj, dofs1);

	  TEST_EXIT_DBG(dofs0.size() == dofs1.size())("Should not happen!\n");

	  BoundaryType type = bound.type;

	  for (unsigned int j = 0; j < dofs0.size(); j++) {
	    DegreeOfFreedom globalDof0 = mapLocalGlobalDofs[*(dofs0[j])];
	    DegreeOfFreedom globalDof1 = mapLocalGlobalDofs[*(dofs1[j])];

	    if (periodicDofAssociations[globalDof0].count(type) == 0) {
	      periodicDof[type][globalDof0] = globalDof1;
	      periodicDofAssociations[globalDof0].insert(type);
	    }
	  }
	}

      } else {
	// Create DOF indices on the boundary. 
	DofContainer& dofs = rankPeriodicDofs[it->first];
	for (vector<AtomicBoundary>::iterator boundIt = it->second.begin();
	     boundIt != it->second.end(); ++boundIt) {

	  int nDofs = dofs.size();
	  boundIt->rankObj.el->getVertexDofs(feSpace, boundIt->rankObj, dofs);
	  boundIt->rankObj.el->getNonVertexDofs(feSpace, boundIt->rankObj, dofs);

	  for (unsigned int i = 0; i < (dofs.size() - nDofs); i++)
	    rankToDofType[it->first].push_back(boundIt->type);
	}
	
	// Send the global indices to the rank on the other side.
	stdMpi.getSendData(it->first).reserve(dofs.size());
	for (unsigned int i = 0; i < dofs.size(); i++)
	  stdMpi.getSendData(it->first).push_back(mapLocalGlobalDofs[*(dofs[i])]);	
	
	// Receive from this rank the same number of dofs.
	stdMpi.recv(it->first, dofs.size());
      }
    }

    stdMpi.updateSendDataSize();
    stdMpi.startCommunication<int>(MPI_INT);


    // === The rank has received the dofs from the rank on the other side of ===
    // === the boundary. Now it can use them to create the mapping between   ===
    // === the periodic dofs in this rank and the corresponding periodic     ===
    // === dofs from the other ranks.                                        ===


    for (RankToBoundMap::iterator it = periodicBoundary.boundary.begin();
	 it != periodicBoundary.boundary.end(); ++it) {
      DofContainer& dofs = rankPeriodicDofs[it->first];
      vector<int>& types = rankToDofType[it->first];

      TEST_EXIT_DBG(dofs.size() == types.size())("Should not happen!\n");

      // Added the received dofs to the mapping.
      for (unsigned int i = 0; i < dofs.size(); i++) {
	int globalDofIndex = mapLocalGlobalDofs[*(dofs[i])];
	int mapGlobalDofIndex = stdMpi.getRecvData(it->first)[i];
	BoundaryType type = types[i];

	// Check if this global dof with the corresponding boundary type was
	// not added before by another periodic boundary from other rank.
	if (periodicDofAssociations[globalDofIndex].count(type) == 0) {
	  periodicDof[type][globalDofIndex] = mapGlobalDofIndex;
	  periodicDofAssociations[globalDofIndex].insert(type);
	}
      }
    }


    StdMpi<PeriodicDofMap> stdMpi2(mpiComm);



    for (RankToBoundMap::iterator it = periodicBoundary.boundary.begin();
	 it != periodicBoundary.boundary.end(); ++it) {
      if (it->first == mpiRank)
	continue;

      PeriodicDofMap perObjMap;

      for (vector<AtomicBoundary>::iterator boundIt = it->second.begin();
	   boundIt != it->second.end(); ++boundIt) {
	if (boundIt->type >= -3)
	  continue;
	
	DofContainer dofs;
	boundIt->rankObj.el->getVertexDofs(feSpace, boundIt->rankObj, dofs);
	boundIt->rankObj.el->getNonVertexDofs(feSpace, boundIt->rankObj, dofs);

	for (unsigned int i = 0; i < dofs.size(); i++) {
	  DegreeOfFreedom globalDof = mapLocalGlobalDofs[*dofs[i]];

	  for (std::set<BoundaryType>::iterator perAscIt = periodicDofAssociations[globalDof].begin();
	       perAscIt != periodicDofAssociations[globalDof].end(); ++perAscIt)
	    if (*perAscIt >= -3) {
	      TEST_EXIT_DBG(periodicDof[*perAscIt].count(globalDof) == 1)
		("Should not happen!\n");
	      perObjMap[*perAscIt][globalDof] = periodicDof[*perAscIt][globalDof];
	    }
	}
      }

      stdMpi2.send(it->first, perObjMap);
      stdMpi2.recv(it->first);
    }

    stdMpi2.startCommunication<int>(MPI_INT);

    for (std::map<int, PeriodicDofMap>::iterator it = stdMpi2.getRecvData().begin();
	 it != stdMpi2.getRecvData().end(); ++it)
      for (PeriodicDofMap::iterator perIt = it->second.begin();
	   perIt != it->second.end(); ++perIt)
	for (DofMapping::iterator dofIt = perIt->second.begin();
	     dofIt != perIt->second.end(); ++dofIt) {
	  TEST_EXIT_DBG(periodicDof[perIt->first].count(dofIt->second) == 0 ||
			periodicDof[perIt->first][dofIt->second] == dofIt->first)
	    ("Should not happen!\n");

	  periodicDof[perIt->first][dofIt->second] = dofIt->first;
	}

#if (DEBUG != 0)
    ParallelDebug::testPeriodicBoundary(*this);
#endif
  }


  void MeshDistributor::createMacroElementInfo()
  {
    FUNCNAME("MeshDistributor::createMacroElementInfo()");

    for (deque<MacroElement*>::iterator it = mesh->firstMacroElement();
	 it != mesh->endOfMacroElements(); ++it) {

      allMacroElements.push_back(*it);

      macroElementNeighbours[(*it)->getIndex()].resize(mesh->getGeo(NEIGH));
      for (int i = 0; i < mesh->getGeo(NEIGH); i++)
	macroElementNeighbours[(*it)->getIndex()][i] = 
	  (*it)->getNeighbour(i) ? (*it)->getNeighbour(i)->getIndex() : -1;
    }     
  }


  void MeshDistributor::updateMacroElementInfo()
  {
    deque<MacroElement*>& meshMacros = mesh->getMacroElements();

    for (unsigned int i = 0; i < allMacroElements.size(); i++) {
      for (deque<MacroElement*>::iterator it = meshMacros.begin();
	   it != meshMacros.end(); ++it) {
	if ((*it)->getIndex() == allMacroElements[i]->getIndex() &&
	    *it != allMacroElements[i]) {
	  delete allMacroElements[i];
	  allMacroElements[i] = *it;
	}
      }
    }
  }


  void MeshDistributor::serialize(ostream &out)
  {
    FUNCNAME("MeshDistributor::serialize()");

    checkMeshChange();

    SerUtil::serialize(out, elemWeights);
    SerUtil::serialize(out, partitionVec);
    SerUtil::serialize(out, oldPartitionVec);
    SerUtil::serialize(out, nRankDofs);
    SerUtil::serialize(out, nOverallDofs);

    elObjects.serialize(out);

    myIntBoundary.serialize(out);
    otherIntBoundary.serialize(out);
    periodicBoundary.serialize(out);

    serialize(out, sendDofs);
    serialize(out, recvDofs);

    SerUtil::serialize(out, mapLocalGlobalDofs);
    SerUtil::serialize(out, mapLocalDofIndex);
    SerUtil::serialize(out, isRankDof);

    serialize(out, periodicDof);
    serialize(out, periodicDofAssociations);

    SerUtil::serialize(out, rstart);
    SerUtil::serialize(out, macroElementNeighbours);

    int nSize = allMacroElements.size();
    SerUtil::serialize(out, nSize);
    for (int i = 0; i < nSize; i++)
      allMacroElements[i]->serialize(out);

    SerUtil::serialize(out, nTimestepsAfterLastRepartitioning);
    SerUtil::serialize(out, repartCounter);
  }


  void MeshDistributor::deserialize(istream &in)
  {
    FUNCNAME("MeshDistributor::deserialize()");

    SerUtil::deserialize(in, elemWeights);
    SerUtil::deserialize(in, partitionVec);
    SerUtil::deserialize(in, oldPartitionVec);
    SerUtil::deserialize(in, nRankDofs);    
    SerUtil::deserialize(in, nOverallDofs);

    // Create two maps: one from from element indices to the corresponding element 
    // pointers, and one map from Dof indices to the corresponding dof pointers.
    map<int, Element*> elIndexMap;
    map<int, const DegreeOfFreedom*> dofMap;
    ElementDofIterator elDofIter(feSpace);
    TraverseStack stack;
    ElInfo *elInfo = stack.traverseFirst(mesh, -1, Mesh::CALL_EVERY_EL_PREORDER);
    while (elInfo) {
      Element *el = elInfo->getElement();
      elIndexMap[el->getIndex()] = el;

      if (el->isLeaf()) {
	elDofIter.reset(el);
	do {
	  dofMap[elDofIter.getDof()] = elDofIter.getDofPtr();
	} while (elDofIter.next());      
      }

      elInfo = stack.traverseNext(elInfo);
    }

    elObjects.deserialize(in);
   
    myIntBoundary.deserialize(in, elIndexMap);
    otherIntBoundary.deserialize(in, elIndexMap);
    periodicBoundary.deserialize(in, elIndexMap);

    deserialize(in, sendDofs, dofMap);
    deserialize(in, recvDofs, dofMap);

    SerUtil::deserialize(in, mapLocalGlobalDofs);
    SerUtil::deserialize(in, mapLocalDofIndex);
    SerUtil::deserialize(in, isRankDof);

    deserialize(in, periodicDof);
    deserialize(in, periodicDofAssociations);

    SerUtil::deserialize(in, rstart);
    SerUtil::deserialize(in, macroElementNeighbours);

    int nSize = 0;
    SerUtil::deserialize(in, nSize);
    allMacroElements.resize(nSize);
    for (int i = 0; i < nSize; i++) {
      // We do not need the neighbour indices, but must still read them.
      vector<int> indices;
      allMacroElements[i] = new MacroElement(mesh->getDim());
      allMacroElements[i]->writeNeighboursTo(&indices);
      allMacroElements[i]->deserialize(in);
      allMacroElements[i]->getElement()->setMesh(mesh);
    }

    SerUtil::deserialize(in, nTimestepsAfterLastRepartitioning);
    SerUtil::deserialize(in, repartCounter);

    deserialized = true;
  }


  void MeshDistributor::serialize(ostream &out, PeriodicDofMap &data)
  {
    int mapSize = data.size();
    SerUtil::serialize(out, mapSize);

    for (PeriodicDofMap::iterator it = data.begin(); it != data.end(); ++it) {
      int type = it->first;
      DofMapping dofMap = it->second;

      SerUtil::serialize(out, type);
      SerUtil::serialize(out, dofMap);
    }
  }


  void MeshDistributor::serialize(ostream &out, map<int, std::set<int> >& data)
  {
    int mapSize = data.size();
    SerUtil::serialize(out, mapSize);

    for (map<int, std::set<int> >::iterator it = data.begin(); it != data.end(); ++it) {
      int dof = it->first;
      std::set<int> typeSet = it->second;

      SerUtil::serialize(out, dof);
      SerUtil::serialize(out, typeSet);
    }
  }


  void MeshDistributor::deserialize(istream &in, PeriodicDofMap &data)
  {
    data.clear();

    int mapSize = 0;
    SerUtil::deserialize(in, mapSize);

    for (int i = 0; i < mapSize; i++) {
      int type;
      DofMapping dofMap;

      SerUtil::deserialize(in, type);
      SerUtil::deserialize(in, dofMap);

      data[type] = dofMap;
    }
  }


  void MeshDistributor::deserialize(istream &in, 
				    map<int, std::set<int> >& data)
  {
    data.clear();

    int mapSize = 0;
    SerUtil::deserialize(in, mapSize);

    for (int i = 0; i < mapSize; i++) {
      int dof;
      std::set<int> typeSet;

      SerUtil::deserialize(in, dof);
      SerUtil::deserialize(in, typeSet);

      data[dof] = typeSet;
    }
  }

 
  void MeshDistributor::writePartitioningMesh(string filename)
  {
    FUNCNAME("MeshDistributor::writePartitioningMesh()");

    map<int, double> vec;    
    TraverseStack stack;
    ElInfo *elInfo = stack.traverseFirst(mesh, -1, 
					 Mesh::CALL_LEAF_EL | Mesh::FILL_COORDS);
    
    while (elInfo) {		  
      int index = elInfo->getElement()->getIndex();
      vec[index] = partitionVec[index];
      elInfo = stack.traverseNext(elInfo);
    }

    ElementFileWriter::writeFile(vec, mesh, filename);
  }

}