source: src/Parser/TremoloParser.cpp@ 13bfeb

/*
 * TremoloParser.cpp
 *
 *  Created on: Mar 2, 2010
 *      Author: metzler
 */

#include "Helpers/Assert.hpp"
#include "TremoloParser.hpp"
#include "World.hpp"
#include "atom.hpp"
#include "element.hpp"
#include "bond.hpp"
#include "periodentafel.hpp"
#include "Descriptors/AtomIdDescriptor.hpp"
#include <map>
#include <vector>


using namespace std;
using namespace boost;

/**
 * Constructor.
 */
TremoloParser::TremoloParser() {
  knownKeys[" "] = TremoloKey::noKey; // with this we can detect invalid keys
  knownKeys["x"] = TremoloKey::x;
  knownKeys["u"] = TremoloKey::u;
  knownKeys["F"] = TremoloKey::F;
  knownKeys["stress"] = TremoloKey::stress;
  knownKeys["Id"] = TremoloKey::Id;
  knownKeys["neighbors"] = TremoloKey::neighbors;
  knownKeys["imprData"] = TremoloKey::imprData;
  knownKeys["GroupMeasureTypeNo"] = TremoloKey::GroupMeasureTypeNo;
  knownKeys["Type"] = TremoloKey::Type;
  knownKeys["extType"] = TremoloKey::extType;
  knownKeys["name"] = TremoloKey::name;
  knownKeys["resName"] = TremoloKey::resName;
  knownKeys["chainID"] = TremoloKey::chainID;
  knownKeys["resSeq"] = TremoloKey::resSeq;
  knownKeys["occupancy"] = TremoloKey::occupancy;
  knownKeys["tempFactor"] = TremoloKey::tempFactor;
  knownKeys["segID"] = TremoloKey::segID;
  knownKeys["Charge"] = TremoloKey::Charge;
  knownKeys["charge"] = TremoloKey::charge;
  knownKeys["GrpTypeNo"] = TremoloKey::GrpTypeNo;
  knownKeys["torsion"] = TremoloKey::torsion;
}

/**
 * Destructor.
 */
TremoloParser::~TremoloParser() {
  usedFields.clear();
  additionalAtomData.clear();
  atomIdMap.clear();
  knownKeys.clear();
}

/**
 * Loads atoms from a tremolo-formatted file.
 *
 * \param tremolo file
 */
void TremoloParser::load(istream* file) {
  string line;
  string::size_type location;

  usedFields.clear();
  while (file->good()) {
    std::getline(*file, line, '\n');
    if (usedFields.empty()) {
      location = line.find("ATOMDATA", 0);
      if (location != string::npos) {
       parseAtomDataKeysLine(line, location + 8);
      }
    }
    if (line.length() > 0 && line.at(0) != '#') {
      readAtomDataLine(line);
    }
  }

  processNeighborInformation();
  adaptImprData();
  adaptTorsion();
}

/**
 * Saves the World's current state into as a tremolo file.
 *
 * \param file where to save the state
 */
void TremoloParser::save(ostream* file) {
  vector<atom*>::iterator atomIt;
  vector<string>::iterator it;

  *file << "# ATOMDATA";
  for (it=usedFields.begin(); it < usedFields.end(); it++) {
    *file << "\t" << *it;
  }
  *file << endl;
  vector<atom *> AtomList = World::getInstance().getAllAtoms();
  for (atomIt = AtomList.begin(); atomIt != AtomList.end(); atomIt++) {
    saveLine(file, *atomIt);
  }
}

/**
 * Sets the keys for which data should be written to the stream when save is
 * called.
 *
 * \param string of field names with the same syntax as for an ATOMDATA line
 *        but without the prexix "ATOMDATA"
 */
void TremoloParser::setFieldsForSave(std::string atomDataLine) {
  parseAtomDataKeysLine(atomDataLine, 0);
}


/**
 * Writes one line of tremolo-formatted data to the provided stream.
 *
 * \param stream where to write the line to
 * \param reference to the atom of which information should be written
 */
void TremoloParser::saveLine(ostream* file, atom* currentAtom) {
  vector<string>::iterator it;
  TremoloKey::atomDataKey currentField;

  for (it = usedFields.begin(); it != usedFields.end(); it++) {
    currentField = knownKeys[it->substr(0, it->find("="))];
    switch (currentField) {
      case TremoloKey::x :
        // for the moment, assume there are always three dimensions
        *file << currentAtom->x[0] << "\t";
        *file << currentAtom->x[1] << "\t";
        *file << currentAtom->x[2] << "\t";
        break;
      case TremoloKey::u :
        // for the moment, assume there are always three dimensions
        *file << currentAtom->v[0] << "\t";
        *file << currentAtom->v[1] << "\t";
        *file << currentAtom->v[2] << "\t";
        break;
      case TremoloKey::Type :
        *file << currentAtom->getType()->getSymbol() << "\t";
        break;
      case TremoloKey::Id :
        *file << currentAtom->getId() << "\t";
        break;
      case TremoloKey::neighbors :
        writeNeighbors(file, atoi(it->substr(it->find("=") + 1, 1).c_str()), currentAtom);
        break;
      default :
        *file << (additionalAtomData.find(currentAtom->getId()) != additionalAtomData.end()
          ? additionalAtomData[currentAtom->getId()].get(currentField)
          : defaultAdditionalData.get(currentField));
        *file << "\t";
        break;
    }
  }

  *file << endl;
}

/**
 * Writes the neighbor information of one atom to the provided stream.
 *
 * \param stream where to write neighbor information to
 * \param number of neighbors
 * \param reference to the atom of which to take the neighbor information
 */
void TremoloParser::writeNeighbors(ostream* file, int numberOfNeighbors, atom* currentAtom) {
  BondList::iterator currentBond = currentAtom->ListOfBonds.begin();
  for (int i = 0; i < numberOfNeighbors; i++) {
    *file << (currentBond != currentAtom->ListOfBonds.end()
        ? (*currentBond)->GetOtherAtom(currentAtom)->getId() : 0) << "\t";
  }
}

/**
 * Stores keys from the ATOMDATA line.
 *
 * \param line to parse the keys from
 * \param with which offset the keys begin within the line
 */
void TremoloParser::parseAtomDataKeysLine(string line, int offset) {
  string keyword;
  stringstream lineStream;

  lineStream << line.substr(offset);
  while (lineStream.good()) {
    lineStream >> keyword;
    if (knownKeys[keyword.substr(0, keyword.find("="))] == TremoloKey::noKey) {
      // TODO: throw exception about unknown key
      cout << "Unknown key: " << keyword << " is not part of the tremolo format specification." << endl;
      break;
    }
    usedFields.push_back(keyword);
  }
}

/**
 * Reads one data line of a tremolo file and interprets it according to the keys
 * obtained from the ATOMDATA line.
 *
 * \param line to parse as an atom
 */
void TremoloParser::readAtomDataLine(string line) {
  vector<string>::iterator it;
  stringstream lineStream;
  atom* newAtom = World::getInstance().createAtom();
  TremoloAtomInfoContainer *atomInfo = NULL;
  additionalAtomData[newAtom->getId()] = *(new TremoloAtomInfoContainer);
  atomInfo = &additionalAtomData[newAtom->getId()];
  TremoloKey::atomDataKey currentField;
  string word;
  int oldId;

  lineStream << line;
  for (it = usedFields.begin(); it < usedFields.end(); it++) {
    currentField = knownKeys[it->substr(0, it->find("="))];
    switch (currentField) {
      case TremoloKey::x :
        // for the moment, assume there are always three dimensions
        lineStream >> newAtom->x[0];
        lineStream >> newAtom->x[1];
        lineStream >> newAtom->x[2];
        break;
      case TremoloKey::u :
        // for the moment, assume there are always three dimensions
        lineStream >> newAtom->v[0];
        lineStream >> newAtom->v[1];
        lineStream >> newAtom->v[2];
        break;
      case TremoloKey::Type :
        char type[3];
        lineStream >> type;
        newAtom->setType(World::getInstance().getPeriode()->FindElement(type));
        ASSERT(newAtom->getType(), "Type was not set for this atom");
        break;
      case TremoloKey::Id :
        lineStream >> oldId;
        atomIdMap[oldId] = newAtom->getId();
        break;
      case TremoloKey::neighbors :
        readNeighbors(&lineStream,
            atoi(it->substr(it->find("=") + 1, 1).c_str()), newAtom->getId());
        break;
      default :
        lineStream >> word;
        atomInfo->set(currentField, word);
        break;
    }
  }
}

/**
 * Reads neighbor information for one atom from the input.
 *
 * \param stream where to read the information from
 * \param number of neighbors to read
 * \param world id of the atom the information belongs to
 */
void TremoloParser::readNeighbors(stringstream* line, int numberOfNeighbors, int atomId) {
  int neighborId = 0;
  for (int i = 0; i < numberOfNeighbors; i++) {
    *line >> neighborId;
    // 0 is used to fill empty neighbor positions in the tremolo file.
    if (neighborId > 0) {
      additionalAtomData[atomId].neighbors.push_back(neighborId);
    }
  }
}

/**
 * Checks whether the provided name is within the list of used fields.
 *
 * \param field name to check
 *
 * \return true if the field name is used
 */
bool TremoloParser::isUsedField(string fieldName) {
  bool fieldNameExists = false;
  for (vector<string>::iterator usedField = usedFields.begin(); usedField != usedFields.end(); usedField++) {
    if (usedField->substr(0, usedField->find("=")) == fieldName)
      fieldNameExists = true;
  }

  return fieldNameExists;
}


/**
 * Adds the collected neighbor information to the atoms in the world. The atoms
 * are found by their current ID and mapped to the corresponding atoms with the
 * Id found in the parsed file.
 */
void TremoloParser::processNeighborInformation() {
  if (!isUsedField("neighbors")) {
    return;
  }

  for(map<int, TremoloAtomInfoContainer>::iterator currentInfo = additionalAtomData.begin();
    currentInfo != additionalAtomData.end(); currentInfo++
  ) {
    for(vector<int>::iterator neighbor = currentInfo->second.neighbors.begin();
      neighbor != currentInfo->second.neighbors.end(); neighbor++
    ) {
      World::getInstance().getAtom(AtomById(currentInfo->first))
          ->addBond(World::getInstance().getAtom(AtomById(atomIdMap[*neighbor])));
    }
  }
}

/**
 * Replaces atom IDs read from the file by the corresponding world IDs. All IDs
 * IDs of the input string will be replaced; expected separating characters are
 * "-" and ",".
 *
 * \param string in which atom IDs should be adapted
 *
 * \return input string with modified atom IDs
 */
string TremoloParser::adaptIdDependentDataString(string data) {
  // there might be no IDs
  if (data == "-") {
    return "-";
  }

  char separator;
  int id;
  stringstream line, result;
  line << data;

  line >> id;
  result << atomIdMap[id];
  while (line.good()) {
    line >> separator >> id;
    result << separator << atomIdMap[id];
  }

  return result.str();
}

/**
 * Corrects the atom IDs in each imprData entry to the corresponding world IDs
 * as they might differ from the originally read IDs.
 */
void TremoloParser::adaptImprData() {
  if (!isUsedField("imprData")) {
    return;
  }

  for(map<int, TremoloAtomInfoContainer>::iterator currentInfo = additionalAtomData.begin();
    currentInfo != additionalAtomData.end(); currentInfo++
  ) {
    currentInfo->second.imprData = adaptIdDependentDataString(currentInfo->second.imprData);
  }
}

/**
 * Corrects the atom IDs in each torsion entry to the corresponding world IDs
 * as they might differ from the originally read IDs.
 */
void TremoloParser::adaptTorsion() {
  if (!isUsedField("torsion")) {
    return;
  }

  for(map<int, TremoloAtomInfoContainer>::iterator currentInfo = additionalAtomData.begin();
    currentInfo != additionalAtomData.end(); currentInfo++
  ) {
    currentInfo->second.torsion = adaptIdDependentDataString(currentInfo->second.torsion);
  }
}


TremoloAtomInfoContainer::TremoloAtomInfoContainer() {
  F = "0";
  stress = "0";
  imprData = "-";
  GroupMeasureTypeNo = "0";
  extType = "-";
  name = "-";
  resName = "-";
  chainID = "0";
  resSeq = "0";
  occupancy = "0";
  tempFactor = "0";
  segID = "0";
  Charge = "0";
  charge = "0";
  GrpTypeNo = "0";
  torsion = "-";
  neighbors = vector<int>(0, 5);
}

void TremoloAtomInfoContainer::set(TremoloKey::atomDataKey key, string value) {
  switch (key) {
    case TremoloKey::F :
      F = value;
      break;
    case TremoloKey::stress :
      stress = value;
     break;
    case TremoloKey::imprData :
      imprData = value;
      break;
    case TremoloKey::GroupMeasureTypeNo :
      GroupMeasureTypeNo = value;
      break;
    case TremoloKey::extType :
      extType = value;
      break;
    case TremoloKey::name :
      name = value;
      break;
    case TremoloKey::resName :
      resName = value;
      break;
    case TremoloKey::chainID :
      chainID = value;
      break;
    case TremoloKey::resSeq :
      resSeq = value;
      break;
    case TremoloKey::occupancy :
      occupancy = value;
      break;
    case TremoloKey::tempFactor :
      tempFactor = value;
      break;
    case TremoloKey::segID :
      segID = value;
      break;
    case TremoloKey::Charge :
      Charge = value;
      break;
    case TremoloKey::charge :
      charge = value;
      break;
    case TremoloKey::GrpTypeNo :
      GrpTypeNo = value;
      break;
    case TremoloKey::torsion :
      torsion = value;
      break;
    default :
      cout << "Unknown key: " << key << ", value: " << value << endl;
      break;
  }
}

string TremoloAtomInfoContainer::get(TremoloKey::atomDataKey key) {
  switch (key) {
    case TremoloKey::F :
      return F;
    case TremoloKey::stress :
      return stress;
    case TremoloKey::imprData :
      return imprData;
    case TremoloKey::GroupMeasureTypeNo :
      return GroupMeasureTypeNo;
    case TremoloKey::extType :
      return extType;
    case TremoloKey::name :
      return name;
    case TremoloKey::resName :
      return resName;
    case TremoloKey::chainID :
      return chainID;
    case TremoloKey::resSeq :
      return resSeq;
    case TremoloKey::occupancy :
      return occupancy;
    case TremoloKey::tempFactor :
      return tempFactor;
    case TremoloKey::segID :
      return segID;
    case TremoloKey::Charge :
      return Charge;
    case TremoloKey::charge :
      return charge;
    case TremoloKey::GrpTypeNo :
      return GrpTypeNo;
    case TremoloKey::torsion :
      return torsion;
    default :
      cout << "Unknown key: " << key << endl;
      return "";
  }
}