C++ OBPairData类代码示例

  bool QEqCharges::ComputeCharges(OBMol &mol)
  {

     ///////////////////////////////////////////////////////////////////////////////
    //Some OpenBabel bookkeeping that I copied from the Gasteiger scheme
    mol.SetPartialChargesPerceived();

    OBPairData *dp = new OBPairData;
    dp->SetAttribute("PartialCharges");
    dp->SetValue("QEq");
    dp->SetOrigin(perceived);
    mol.SetData(dp);


    ///////////////////////////////////////////////////////////////////////////////
    //Read in atomic information from OpenBabel molecule and parameterize

    //Read in total number of atoms
    int i, N = mol.NumAtoms();

    Hardness = MatrixXd::Zero(N+1, N+1);
    Voltage = VectorXd::Zero(N+1);
    Electronegativity = VectorXd::Zero(N);
    VectorXd BasisSet = VectorXd::Zero(N);

    Vector3d Parameters;

    FOR_ATOMS_OF_MOL(atom, mol)
    {
       	Parameters = GetParameters(atom->GetAtomicNum(), atom->GetFormalCharge());
	i = atom->GetIdx() - 1;

	if (Parameters[0] == 0.)
        {
		stringstream msg;
		msg << "Some QEq Parameters not found!" << endl
	            << "Parameters not found for atom no. " << i+1 << endl
		    << "Atom will be ignored in the charge computation.";
		obErrorLog.ThrowError(__FUNCTION__, msg.str(), obError);
	}

	Electronegativity[i] = Parameters[0];
	Hardness(i,i) = Parameters[1];
	BasisSet[i] = Parameters[2];
    }

/**
@since version 2.3
Adds an OBPairData object to each atom and bond in a substructure.
The substructure's atoms are specified in an input parameter, a
vector of atom indx; the bonds are those in the molecule that join
these atoms. The attribute and value of the OBPairObject (the same
for all the added objects) are specified as parameters.
**/
bool AddDataToSubstruct(OBMol* pmol,
        const std::vector<int>& atomIdxs,
        const std::string& attribute,
        const std::string& value)
{
  //Add data to atoms
  for(unsigned int j=0; j<atomIdxs.size(); ++j)
  {
    OBAtom* pAtom = pmol->GetAtom(atomIdxs[j]);
    if(!pAtom)
      continue;
    OBPairData* dp = new OBPairData;
    dp->SetAttribute(attribute);
    dp->SetValue(value);
    pAtom->SetData(dp);
  }

  OBBond* pBond;
  vector<OBBond*>::iterator i;
  for(pBond = pmol->BeginBond(i); pBond; pBond = pmol->NextBond(i))
  {
    //Add data to bond if it joins two atoms in list
    if(count(atomIdxs.begin(), atomIdxs.end(), pBond->GetBeginAtomIdx())
        && count(atomIdxs.begin(), atomIdxs.end(), pBond->GetEndAtomIdx()))
    {
      OBPairData* dp = new OBPairData;
      dp->SetAttribute(attribute);
      dp->SetValue(value);
      pBond->SetData(dp);
    }
  }
  return true;
}

  unsigned int GetAtomSymClass(OBAtom *atom)
  {
    OBPairData *pd = dynamic_cast<OBPairData*>(atom->GetParent()->GetData("OpenBabel Symmetry Classes"));
    if (pd) {

      cout << "same? = " << pd->GetValue() << endl;

      istringstream iss(pd->GetValue());
      std::vector<unsigned int> symmetry_classes;
      copy(istream_iterator<unsigned int>(iss),
           istream_iterator<unsigned int>(),
           back_inserter<vector<unsigned int> >(symmetry_classes));
      // Now find the number of unique elements
      vector<unsigned int> copy_sym = symmetry_classes;
      sort(copy_sym.begin(), copy_sym.end());
      vector<unsigned int>::iterator end_pos = unique(copy_sym.begin(), copy_sym.end()); // Requires sorted elements
      int nclasses = end_pos - copy_sym.begin();

      cout << "sym_class[" << atom->GetIndex() << "] = " << symmetry_classes.at(atom->GetIndex()) << endl;
      return symmetry_classes.at(atom->GetIndex());
    }

    return 99;
  }

  //! \return whether partial charges were successfully assigned to this molecule
  bool EQEqCharges::ComputeCharges(OBMol &mol)
  {
    int i, j, a, c, N = mol.NumAtoms();
    double cellVolume;
    VectorXf chi(N), J(N), b(N), x(N);
    MatrixXf J_ij(N, N), A(N, N);
    OBUnitCell *obuc;
    matrix3x3 unitcell, fourier;
    vector3 dx;
    int numNeighbors[3];
    OBAtom *atom;

    // If parameters have not yet been loaded, do that
    if (!_paramFileLoaded)
    {
      if (ParseParamFile())
      {
        _paramFileLoaded = true;
      } else
      {
        return false;
      }
    }

    // Calculate atomic properties based around their ionic charge
    for (i = 0; i < N; i++)
    {
      atom = mol.GetAtom(i + 1);
      a = atom->GetAtomicNum();
      c = _chargeCenter[a];

      // Fail if ionization data is missing for any atom in the molecule
      if (_ionizations[a][c + 1] == -1 || _ionizations[a][c] == -1 || a > TABLE_OF_ELEMENTS_SIZE)
      {
        obErrorLog.ThrowError(__FUNCTION__, "Insufficient ionization data for atoms in the given molecule. Update `data/eqeqIonizations.txt` with missing information and re-run this function.", obError);
        return false;
      }

      J(i) = _ionizations[a][c + 1] - _ionizations[a][c];
      chi(i) = 0.5 * (_ionizations[a][c + 1] + _ionizations[a][c]) - (a == 1? 0 : c * J(i));
    }

    // If a unit cell is defined, use the periodic Ewald calculation
    if (mol.HasData(OBGenericDataType::UnitCell))
    {
      // Get unit cell and calculate its Fourier transform + volume
      obuc = (OBUnitCell *) mol.GetData(OBGenericDataType::UnitCell);
      unitcell = obuc->GetCellMatrix();
      fourier = (2 * PI * unitcell.inverse()).transpose();
      cellVolume = obuc->GetCellVolume();

      // Get the number of radial unit cells to use in x, y, and z
      numNeighbors[0] = int(ceil(minCellLength / (2.0 * (obuc->GetA())))) - 1;
      numNeighbors[1] = int(ceil(minCellLength / (2.0 * (obuc->GetB())))) - 1;
      numNeighbors[2] = int(ceil(minCellLength / (2.0 * (obuc->GetC())))) - 1;

      for (i = 0; i < N; i++)
      {
        atom = mol.GetAtom(i + 1);
        for (j = 0; j < N; j++)
        {
          dx = atom->GetVector() - (mol.GetAtom(j + 1))->GetVector();
          J_ij(i, j) = GetPeriodicEwaldJij(J(i), J(j), dx, (i == j), unitcell, fourier, cellVolume, numNeighbors);
        }
      }
    // If no unit cell, use the simplified nonperiodic calculation
    } else
    {
      for (i = 0; i < N; i++)
      {
        atom = mol.GetAtom(i + 1);
        for (j = 0; j < N; j++)
        {
          J_ij(i, j) = GetNonperiodicJij(J(i), J(j), atom->GetDistance(j + 1), (i == j));
        }
        return false;
      }
    }

    // Formulate problem as A x = b, where x is the calculated partial charges
    // First equation is a simple overall balance: sum(Q) = 0
    A.row(0) = VectorXf::Ones(N);
    b(0) = 0;

    // Remaining equations are based off of the fact that, at equilibrium, the
    // energy of the system changes equally for a change in any charge:
    //     dE/dQ_1 = dE/dQ_2 = ... = dE/dQ_N
    A.block(1, 0, N - 1, N) = J_ij.block(0, 0, N - 1, N) - J_ij.block(1, 0, N - 1, N);
    b.tail(N - 1) = chi.tail(N - 1) - chi.head(N - 1);

    // The solution is a list of charges in the system
    x = A.colPivHouseholderQr().solve(b);

    // Now we are done calculating, pass all this back to OpenBabel molecule
    mol.SetPartialChargesPerceived();
    OBPairData *dp = new OBPairData;
    dp->SetAttribute("PartialCharges");
    dp->SetValue("EQEq");
    dp->SetOrigin(perceived);
//.........这里部分代码省略.........

bool TinkerFormat::WriteMolecule(OBBase* pOb, OBConversion* pConv)
{
    OBMol* pmol = dynamic_cast<OBMol*>(pOb);
    if(pmol==NULL)
        return false;

    //Define some references so we can use the old parameter names
    ostream &ofs = *pConv->GetOutStream();
    OBMol &mol = *pmol;
    bool mmffTypes = pConv->IsOption("m",OBConversion::OUTOPTIONS) != NULL;

    unsigned int i;
    char buffer[BUFF_SIZE];
    OBBond *bond;
    vector<OBBond*>::iterator j;

    // Before we try output of MMFF94 atom types, check if it works
    OBForceField *ff = OpenBabel::OBForceField::FindForceField("MMFF94");
    if (mmffTypes && ff && ff->Setup(mol))
      mmffTypes = ff->GetAtomTypes(mol);
    else
      mmffTypes = false; // either the force field isn't available, or it doesn't work

    if (!mmffTypes)
      snprintf(buffer, BUFF_SIZE, "%6d %-20s   MM2 parameters\n",mol.NumAtoms(),mol.GetTitle());
    else
      snprintf(buffer, BUFF_SIZE, "%6d %-20s   MMFF94 parameters\n",mol.NumAtoms(),mol.GetTitle());
    ofs << buffer;

    ttab.SetFromType("INT");

    OBAtom *atom;
    string str,str1;
    for(i = 1;i <= mol.NumAtoms(); i++)
    {
        atom = mol.GetAtom(i);
        str = atom->GetType();
        ttab.SetToType("MM2");
        ttab.Translate(str1,str);

        if (mmffTypes) {
          // Override the MM2 typing
          OBPairData *type = (OpenBabel::OBPairData*)atom->GetData("FFAtomType");
          if (type)
            str1 = type->GetValue().c_str();
        }

        snprintf(buffer, BUFF_SIZE, "%6d %2s  %12.6f%12.6f%12.6f %5d",
                i,
                etab.GetSymbol(atom->GetAtomicNum()),
                atom->GetX(),
                atom->GetY(),
                atom->GetZ(),
                atoi((char*)str1.c_str()));
        ofs << buffer;

        for (bond = atom->BeginBond(j); bond; bond = atom->NextBond(j))
        {
            snprintf(buffer, BUFF_SIZE, "%6d", (bond->GetNbrAtom(atom))->GetIdx());
            ofs << buffer;
        }

        ofs << endl;
    }

    return(true);
}

//.........这里部分代码省略.........
          pmol->DeleteAtom(toDelete.at(i));


        // Discover units
        matrix3x3 conv (1);
        tokenize(vs, buffer);

        if (strstr(vs[1].c_str(), "alat")) {
          conv *= (alat * BOHR_TO_ANGSTROM);
        }
        else if (strstr(vs[1].c_str(), "crystal")) {
          // Set to the zero matrix and test below.
          conv = matrix3x3 (0.0);
        }
        // Add others if needed

        // Load new atoms from molecule
        ifs.getline(buffer,BUFF_SIZE); // First entry
        tokenize(vs, buffer);
        int size = vs.size();
        while (size == 4) {
          atomicNum = OBElements::GetAtomicNum(vs[0].c_str());
          x = atof((char*)vs[1].c_str());
          y = atof((char*)vs[2].c_str());
          z = atof((char*)vs[3].c_str());
          // Add atom
          OBAtom *atom = pmol->NewAtom();
          atom->SetAtomicNum(atomicNum);
          vector3 coords (x,y,z);
          if (conv.determinant() == 0.0) { // Fractional coords
            atom->SetVector(cell->FractionalToCartesian(coords));
          }
          else {
            atom->SetVector(conv * coords);
          }

          // Reset vars
          ifs.getline(buffer,BUFF_SIZE); // First entry
          tokenize(vs, buffer);
          size = vs.size();
        }
      }

      // Free energy
      if (strstr(buffer, "Final energy =")) {
        tokenize(vs, buffer);
        pmol->SetEnergy(atof(vs[3].c_str()) * RYDBERG_TO_KCAL_PER_MOL);
      }

      // H - PV = U energy
      if (strstr(buffer, "!    total energy              =")) {
        tokenize(vs, buffer);
        pmol->SetEnergy(atof(vs[4].c_str()) * RYDBERG_TO_KCAL_PER_MOL);
      }

      // Enthalphy
      if (strstr(buffer, "Final enthalpy =")) {
        tokenize(vs, buffer);

        hasEnthalpy = true;
        enthalpy = atof(vs.at(3).c_str()) * RYDBERG_TO_KCAL_PER_MOL;
        pv = enthalpy - pmol->GetEnergy();
      }
    }

    // set final unit cell
    pmol->SetData(cell);

    // Set enthalpy
    if (hasEnthalpy) {
      OBPairData *enthalpyPD = new OBPairData();
      OBPairData *enthalpyPD_pv = new OBPairData();
      OBPairData *enthalpyPD_eV = new OBPairData();
      OBPairData *enthalpyPD_pv_eV = new OBPairData();
      enthalpyPD->SetAttribute("Enthalpy (kcal/mol)");
      enthalpyPD_pv->SetAttribute("Enthalpy PV term (kcal/mol)");
      enthalpyPD_eV->SetAttribute("Enthalpy (eV)");
      enthalpyPD_pv_eV->SetAttribute("Enthalpy PV term (eV)");
      double en_kcal_per_mole = enthalpy;
      double pv_kcal_per_mole = pv;
      double en_eV = enthalpy / EV_TO_KCAL_PER_MOL;
      double pv_eV = pv / EV_TO_KCAL_PER_MOL;
      snprintf(tag, BUFF_SIZE, "%f", en_kcal_per_mole);
      enthalpyPD->SetValue(tag);
      snprintf(tag, BUFF_SIZE, "%f", pv_kcal_per_mole);
      enthalpyPD_pv->SetValue(tag);
      snprintf(tag, BUFF_SIZE, "%f", en_eV);
      enthalpyPD_eV->SetValue(tag);
      snprintf(tag, BUFF_SIZE, "%f", pv_eV);
      enthalpyPD_pv_eV->SetValue(tag);
      pmol->SetData(enthalpyPD);
      pmol->SetData(enthalpyPD_pv);
      pmol->SetData(enthalpyPD_eV);
      pmol->SetData(enthalpyPD_pv_eV);
    }

    pmol->EndModify();

    return true;
  }

  bool PDBFormat::WriteMolecule(OBBase* pOb, OBConversion* pConv)
  {
    OBMol* pmol = dynamic_cast<OBMol*>(pOb);
    if(pmol==NULL)
      return false;

    //Define some references so we can use the old parameter names
    ostream &ofs = *pConv->GetOutStream();
    OBMol &mol = *pmol;

    unsigned int i;
    char buffer[BUFF_SIZE];
    char type_name[10], padded_name[10];
    char the_res[10];
    char the_chain = ' ';
    const char *element_name;
    int res_num;
    bool het=true;
    int model_num = 0;
    if (!pConv->IsLast() || pConv->GetOutputIndex() > 1)
      { // More than one molecule record
        model_num = pConv->GetOutputIndex(); // MODEL 1-based index
        snprintf(buffer, BUFF_SIZE, "MODEL %8d", model_num);
        ofs << buffer << endl;
      }

    // write back all fields (REMARKS, HELIX, SHEET, SITE, ...)
    bool compndWritten = false;
    bool authorWritten = false;
    std::vector<OBGenericData*> pairData = mol.GetAllData(OBGenericDataType::PairData);
    for (std::vector<OBGenericData*>::iterator data = pairData.begin(); data != pairData.end(); ++data) {
      OBPairData *pd = static_cast<OBPairData*>(*data);
      string attr = pd->GetAttribute();

      // filter to make sure we are writing pdb fields only
      if (attr != "HEADER" && attr != "OBSLTE" && attr != "TITLE" && attr != "SPLIT" &&
          attr != "CAVEAT" && attr != "COMPND" && attr != "SOURCE" && attr != "KEYWDS" &&
          attr != "EXPDTA" && attr != "NUMMDL" && attr != "MDLTYP" && attr != "AUTHOR" &&
          attr != "REVDAT" && attr != "SPRSDE" && attr != "JRNL" && attr != "REMARK" &&
          attr != "DBREF" && attr != "DBREF1" && attr != "DBREF2" && attr != "SEQADV" &&
          attr != "SEQRES" && attr != "MODRES" && attr != "HET" && attr != "HETNAM" &&
          attr != "HETSYN" && attr != "FORMUL" && attr != "HELIX" && attr != "SHEET" &&
          attr != "SSBOND" && attr != "LINK" && attr != "CISPEP" && attr != "SITE" &&
          attr != "ORIGX1" && attr != "ORIGX2" && attr != "ORIGX3" && attr != "SCALE1" &&
          attr != "SCALE2" && attr != "SCALE3" && attr != "MATRIX1" && attr != "MATRIX2" &&
          attr != "MATRIX3" && attr != "MODEL")
        continue;

      if (attr == "COMPND")
        compndWritten = true;
      if (attr == "AUTHOR")
        authorWritten = true;

      // compute spacing needed. HELIX, SITE, HET, ... are trimmed when reading
      int nSpacing = 6 - attr.size();
      for (int i = 0; i < nSpacing; ++i)
        attr += " ";


      std::string lines = pd->GetValue();
      string::size_type last = 0;
      string::size_type pos = lines.find('\n');
      while (last != string::npos) {
        string line = lines.substr(last, pos - last);
        if (pos == string::npos)
          last = string::npos;
        else
          last = pos + 1;
        pos = lines.find('\n', last);

        ofs << attr << line << endl;
      }
    }

    if (!compndWritten) {
      if (strlen(mol.GetTitle()) > 0)
        snprintf(buffer, BUFF_SIZE, "COMPND    %s ",mol.GetTitle());
      else
        snprintf(buffer, BUFF_SIZE, "COMPND    UNNAMED");
      ofs << buffer << endl;
    }

    if (!authorWritten) {
      snprintf(buffer, BUFF_SIZE, "AUTHOR    GENERATED BY OPEN BABEL %s",BABEL_VERSION);
      ofs << buffer << endl;
    }

    // Write CRYST1 record, containing unit cell parameters, space group
    // and Z value (supposed to be 1)
    if (pmol->HasData(OBGenericDataType::UnitCell))
      {
        OBUnitCell *pUC = (OBUnitCell*)pmol->GetData(OBGenericDataType::UnitCell);
        if(pUC->GetSpaceGroup()){
          string tmpHM=pUC->GetSpaceGroup()->GetHMName();
          // Do we have an extended HM symbol, with origin choice as ":1" or ":2" ? If so, remove it.
          size_t n=tmpHM.find(":");
          if(n!=string::npos) tmpHM=tmpHM.substr(0,n);
          snprintf(buffer, BUFF_SIZE,
                   "CRYST1%9.3f%9.3f%9.3f%7.2f%7.2f%7.2f %-11s 1",
                   pUC->GetA(), pUC->GetB(), pUC->GetC(),
//.........这里部分代码省略.........

//.........这里部分代码省略.........
    if (!getline(ifs,ln))
      return false; //creator
    if (!getline(ifs, ln))
      return false; //comment
    // Originally the comment was added to the reaction via:
    //     pmol->SetComment(Trim(ln));

    if (!getline(ifs, ln))
      return false; // num reactants, products, and optionally agents

    unsigned int nReactants = 0, nProducts = 0, nAgents = 0;
    bool ok = ParseComponent(ln.c_str() + 0, &nReactants);
    if (!ok)
      return false;
    ok = ParseComponent(ln.c_str() + 3, &nProducts);
    if (!ok)
      return false;
    if (ln[6] != '\0') { // optional agents
      ok = ParseComponent(ln.c_str() + 6, &nAgents);
      if (!ok)
        return false;
    }

    if(nReactants + nProducts + nAgents)
    {
      //Read the first $MOL. The others are read at the end of the previous MOL
      if(!getline(ifs, ln))
        return false;
      if(Trim(ln).find("$MOL")==string::npos)
        return false;
    }

    OBReactionFacade rxnfacade(pmol);

    // Note: If we supported it, we could read each of the rxn components directly
    // into the returned OBMol instead of having to do a copy. Unfortunately,
    // this isn't possible at the moment (MOL format will need some work first).
    // Here is some example code to do it:
    //
    //unsigned int old_numatoms = 0;
    //unsigned int compid = 1;
    //for (int i = 0; i<nReactants; i++)
    //{
    //  //Read a MOL file	using the same OBConversion object but with a different format
    //  if (!pMolFormat->ReadMolecule(pmol, pConv))
    //    obErrorLog.ThrowError(__FUNCTION__, "Failed to read a reactant", obWarning);
    //  unsigned int numatoms = pmol->NumAtoms();
    //  for (unsigned int idx = old_numatoms + 1; idx <= numatoms; ++idx) {
    //    OBAtom* atom = pmol->GetAtom(idx);
    //    rxnfacade.SetRole(atom, REACTANT);
    //    rxnfacade.SetComponentId(atom, compid);
    //  }
    //  old_numatoms = numatoms;
    //  compid++;
    //}

    const char* type[3] = {"a reactant", "a product", "an agent"};
    OBReactionRole role;
    unsigned int num_components;
    for(unsigned int N=0; N<3; N++) {
      switch(N) {
      case 0:
        role = REACTANT;
        num_components = nReactants;
        break;
      case 1:
        role = PRODUCT;
        num_components = nProducts;
        break;
      case 2:
        role = AGENT;
        num_components = nAgents;
        break;
      }
      for (int i=0; i<num_components; i++)
      {
        //Read a MOL file	using the same OBConversion object but with a different format
        OBMol mol;
        if (!pMolFormat->ReadMolecule(&mol, pConv)) {
          std::string error = "Failed to read ";
          error += type[N];
          obErrorLog.ThrowError(__FUNCTION__, error, obWarning);
          continue;
        }
        if (mol.NumAtoms() == 0) {
          OBAtom* dummy = mol.NewAtom(); // Treat the empty OBMol as having a single dummy atom
          OBPairData *pd = new OBPairData();
          pd->SetAttribute("rxndummy");
          pd->SetValue("");
          pd->SetOrigin(fileformatInput);
          dummy->SetData(pd);
        }

        rxnfacade.AddComponent(&mol, role);
      }
    }

    pmol->SetIsReaction();
    return true;
}

//.........这里部分代码省略.........
        do {
          nonRingAtoms = false;
          size = mol.NumAtoms();
          for (unsigned int i = 1; i <= size; ++i)
            {
              atom = mol.GetAtom(i);
              if (!atom->IsInRing()) {
                mol.DeleteAtom(atom);
                nonRingAtoms = true;
                break; // don't know how many atoms there are
              } 
              // Previously, we changed atoms to carbon here.
              // Now we perform this alchemy in terms of string-rewriting
              // once the canonical SMILES is generated
            }
        } while (nonRingAtoms);
        
        if (mol.NumAtoms() < 3)
          continue;
        
        if (mol.NumBonds() == 0)
          continue;
        
        do {
          nonRingBonds = false;
          size = mol.NumBonds();
          for (unsigned int i = 0; i < size; ++i)
            {
              bond = mol.GetBond(i);
              if (!bond->IsInRing()) {
                mol.DeleteBond(bond);
                nonRingBonds = true;
                break; // don't know how many bonds there are
              }
            }        
        } while (nonRingBonds);

        fragments = mol.Separate();
        for (unsigned int i = 0; i < fragments.size(); ++i)
          {
            if (fragments[i].NumAtoms() < 3) // too small to care
              continue;
              
            currentCAN = conv.WriteString(&fragments[i], true);
            currentCAN = RewriteSMILES(currentCAN); // change elements to "a/A" for compression
            if (index.find(currentCAN) != index.end()) { // already got this
              index[currentCAN] += 1; // add to the count for bookkeeping
              continue;
            }

            index[currentCAN] = 1; // don't ever write this ring fragment again

            // OK, now retrieve the canonical ordering for the fragment
            vector<string> canonical_order;
            if (fragments[i].HasData("Canonical Atom Order")) {
              OBPairData *data = (OBPairData*)fragments[i].GetData("Canonical Atom Order");
              tokenize(canonical_order, data->GetValue().c_str());
            }

            // Write out an XYZ-style file with the CANSMI as the title
            cout << fragments[i].NumAtoms() << '\n';
            cout << currentCAN << '\n'; // endl causes a flush

            vector<string>::iterator can_iter;
            unsigned int order;
            OBAtom *atom;

            fragments[i].Center();
            fragments[i].ToInertialFrame();

            for (unsigned int index = 0; index < canonical_order.size(); 
                 ++index) {
              order = atoi(canonical_order[index].c_str());
              atom = fragments[i].GetAtom(order);
              
              snprintf(buffer, BUFF_SIZE, "C%8.3f%8.3f%8.3f\n",
                       atom->x(), atom->y(), atom->z());
              cout << buffer;
            }

          }
        fragments.clear();
        if (index.size() > fragmentCount) {
          fragmentCount = index.size();
          cerr << " Fragments: " << fragmentCount << endl;
        }

      } // while reading molecules (in this file)
    ifs.close();
    ifs.clear();
  } // while reading files

  // loop through the map and output frequencies
  map<string, int>::const_iterator indexItr;
  for (indexItr = index.begin(); indexItr != index.end(); ++indexItr) {
    cerr << (*indexItr).second << " INDEX " << (*indexItr).first << "\n";
  }
    
  return(0);
}

//.........这里部分代码省略.........
      //If there is a -s option, reference molecule has only those atoms that are matched
      //Call the -s option from here
      bool ret = _pOpIsoM->Do(pmol, _stext.c_str(), pmap, pConv);
      // Get the atoms that were matched
      vector<int> ats = _pOpIsoM->GetMatchAtoms();
      if(!ats.empty())
      {
        // Make a vector of the matching atom coordinates...
        for(vector<int>::iterator iter=ats.begin(); iter!=ats.end(); ++iter)
          _refvec.push_back((pmol->GetAtom(*iter))->GetVector());
        // ...and use a vector reference
        _align.SetRef(_refvec);
      }
      // Stop -s option being called normally, although it will still be called once
      //  in the DoOps loop already started for the current (first) molecule.
      pConv->RemoveOption("s",OBConversion::GENOPTIONS);
      if(!ret)
      {
        // the first molecule did not match the -s option so a reference molecule
        // could not be made. Keep trying.
        _refMol.Clear();
        //obErrorLog.ThrowError(__FUNCTION__, "The first molecule did not match the -s option\n"
        //  "so the reference structure was not derived from it", obWarning, onceOnly);
        return false; //not matched
      }
    }
  }

  //All molecules
  if(pmol->GetDimension()!= _refMol.GetDimension())
  {
    stringstream ss;
    ss << "The molecule" << pmol->GetTitle()
       << " does not have the same dimensions as the reference molecule "
       << _refMol.GetTitle() << " and is ignored.";
       obErrorLog.ThrowError(__FUNCTION__, ss.str().c_str(), obError);
    return false;
  }

  if(_pOpIsoM) //Using -s option
  {
    //Ignore mol if it does not pass -s option
    if(!_pOpIsoM->Do(pmol, "", pmap, pConv)) // "" means will use existing parameters
      return false;

    // Get the atoms equivalent to those in ref molecule
    vector<int> ats = _pOpIsoM->GetMatchAtoms();

    // Make a vector of their coordinates and get the centroid
    vector<vector3> vec;
    vector3 centroid;
    for(vector<int>::iterator iter=ats.begin(); iter!=ats.end(); ++iter) {
      vector3 v = pmol->GetAtom(*iter)->GetVector();
      centroid += v;
      vec.push_back(v);
    }
    centroid /= vec.size();

    // Do the alignment
    _align.SetTarget(vec);
    if(!_align.Align())
      return false;

    // Get the centroid of the reference atoms
    vector3 ref_centroid;
    for(vector<vector3>::iterator iter=_refvec.begin(); iter!=_refvec.end(); ++iter)
      ref_centroid += *iter;
    ref_centroid /= _refvec.size();

    //subtract the centroid, rotate the target molecule, then add the centroid
    matrix3x3 rotmatrix = _align.GetRotMatrix();
    for (unsigned int i = 1; i <= pmol->NumAtoms(); ++i)
    {
      vector3 tmpvec = pmol->GetAtom(i)->GetVector();
      tmpvec -= centroid;
      tmpvec *= rotmatrix; //apply the rotation
      tmpvec += ref_centroid;
      pmol->GetAtom(i)->SetVector(tmpvec);
    }
  }
  else //Not using -s option)
  {
    _align.SetTargetMol(*pmol);
    if(!_align.Align())
      return false;
    _align.UpdateCoords(pmol);
  }

  //Save rmsd as a property
  OBPairData* dp = new OBPairData;
  dp->SetAttribute("rmsd");
  double val = _align.GetRMSD();
  if(val<1e-12)
    val = 0.0;
  dp->SetValue(toString(val));
  dp->SetOrigin(local);
  pmol->SetData(dp);

  return true;
}

  bool MOL2Format::WriteMolecule(OBBase* pOb, OBConversion* pConv)
  {
    OBMol* pmol = dynamic_cast<OBMol*>(pOb);
    if(pmol==NULL)
      return false;

    //Define some references so we can use the old parameter names
    ostream &ofs = *pConv->GetOutStream();
    OBMol &mol = *pmol;
    bool ligandsOnly = pConv->IsOption("l", OBConversion::OUTOPTIONS)!=NULL;
 
    //The old code follows....
    string str,str1;
    char buffer[BUFF_SIZE],label[BUFF_SIZE];
    char rnum[BUFF_SIZE],rlabel[BUFF_SIZE];

    ofs << "@<TRIPOS>MOLECULE" << endl;
    str = mol.GetTitle();
    if (str.empty())
      ofs << "*****" << endl;
    else
      ofs << str << endl;

    snprintf(buffer, BUFF_SIZE," %d %d 0 0 0", mol.NumAtoms(),mol.NumBonds());
    ofs << buffer << endl;
    ofs << "SMALL" << endl;

    OBPairData *dp = (OBPairData*)mol.GetData("PartialCharges");
    if (dp != NULL) {
        // Tripos spec says:
        // NO_CHARGES, DEL_RE, GASTEIGER, GAST_HUCK, HUCKEL, PULLMAN, 
        // GAUSS80_CHARGES, AMPAC_CHARGES, MULLIKEN_CHARGES, DICT_ CHARGES,
        // MMFF94_CHARGES, USER_CHARGES
      if (dp->GetValue() == "Mulliken")
        ofs << "MULLIKEN_CHARGES" << endl;
      else // should pick from the Tripos types
        ofs << "GASTEIGER" << endl;
    }
    else { // No idea what these charges are... all our code sets "PartialCharges"
        ofs << "GASTEIGER" << endl;
    }

    ofs << "Energy = " << mol.GetEnergy() << endl;

    if (mol.HasData(OBGenericDataType::CommentData))
      {
        OBCommentData *cd = (OBCommentData*)mol.GetData(OBGenericDataType::CommentData);
        ofs << cd->GetData();
      }

    ofs << endl;
    ofs << "@<TRIPOS>ATOM" << endl;

    OBAtom *atom;
    OBResidue *res;

    vector<OBAtom*>::iterator i;
    vector<int> labelcount;
    labelcount.resize( etab.GetNumberOfElements() );

    ttab.SetFromType("INT");
    ttab.SetToType("SYB");

    for (atom = mol.BeginAtom(i);atom;atom = mol.NextAtom(i))
      {

        //
        //  Use sequentially numbered atom names if no residues
        //

        snprintf(label,BUFF_SIZE, "%s%d",
                 etab.GetSymbol(atom->GetAtomicNum()),
                 ++labelcount[atom->GetAtomicNum()]);
        strcpy(rlabel,"<1>");
        strcpy(rnum,"1");

        str = atom->GetType();
        ttab.Translate(str1,str);

        //
        //  Use original atom names if there are residues
        //

        if (!ligandsOnly && (res = atom->GetResidue()) )
          {
            // use original atom names defined by residue
            snprintf(label,BUFF_SIZE,"%s",(char*)res->GetAtomID(atom).c_str());
            // make sure that residue name includes its number
            snprintf(rlabel,BUFF_SIZE,"%s%d",res->GetName().c_str(), res->GetNum());
            snprintf(rnum,BUFF_SIZE,"%d",res->GetNum());
          }

        snprintf(buffer,BUFF_SIZE,"%7d%1s%-6s%12.4f%10.4f%10.4f%1s%-5s%4s%1s %-8s%10.4f",
                 atom->GetIdx(),"",label,
                 atom->GetX(),atom->GetY(),atom->GetZ(),
                 "",str1.c_str(),
                 rnum,"",rlabel,
                 atom->GetPartialCharge());
        ofs << buffer << endl;
      }
//.........这里部分代码省略.........

  bool MOL2Format::ReadMolecule(OBBase* pOb, OBConversion* pConv)
  {

    OBMol* pmol = pOb->CastAndClear<OBMol>();
    if(pmol==NULL)
      return false;

    //Define some references so we can use the old parameter names
    istream &ifs = *pConv->GetInStream();
    OBMol &mol = *pmol;

    //Old code follows...
    bool foundAtomLine = false;
    char buffer[BUFF_SIZE];
    char *comment = NULL;
    string str,str1;
    vector<string> vstr;
    int len;

    mol.BeginModify();

    for (;;)
      {
        if (!ifs.getline(buffer,BUFF_SIZE))
          return(false);
        if (EQn(buffer,"@<TRIPOS>MOLECULE",17))
          break;
      }

    // OK, just read MOLECULE line
    int lcount;
    int natoms,nbonds;
    bool hasPartialCharges = true;
    for (lcount=0;;lcount++)
      {
        if (!ifs.getline(buffer,BUFF_SIZE))
          return(false);
        if (EQn(buffer,"@<TRIPOS>ATOM",13))
          {
            foundAtomLine = true;
            break;
          }

        if (lcount == 0)
          {
            tokenize(vstr,buffer);
            if (!vstr.empty())
              mol.SetTitle(buffer);
          }
        else if (lcount == 1)
          sscanf(buffer,"%d%d",&natoms,&nbonds);
        else if (lcount == 3) // charge descriptions
          {
            // Annotate origin of partial charges
            OBPairData *dp = new OBPairData;
            dp->SetAttribute("PartialCharges");
            dp->SetValue(buffer);
            dp->SetOrigin(fileformatInput);
            mol.SetData(dp);

            if (strncasecmp(buffer, "NO_CHARGES", 10) == 0)
              hasPartialCharges = false;
          }
        else if (lcount == 4) //energy (?)
          {
            tokenize(vstr,buffer);
            if (!vstr.empty() && vstr.size() == 3)
              if (vstr[0] == "Energy")
                mol.SetEnergy(atof(vstr[2].c_str()));
          }
        else if (lcount == 5) //comment
          {
            if ( buffer[0] )
              {
                len = (int) strlen(buffer)+1;
                //! @todo allow better multi-line comments
                // which don't allow ill-formed data to consume memory
                // Thanks to Andrew Dalke for the pointer
                if (comment != NULL)
                  delete [] comment;
                comment = new char [len];
                memcpy(comment,buffer,len);
              }
          }
      }

    if (!foundAtomLine)
      {
        mol.EndModify();
        mol.Clear();
        obErrorLog.ThrowError(__FUNCTION__, "Unable to read Mol2 format file. No atoms found.", obWarning);
        return(false);
      }

    mol.ReserveAtoms(natoms);

    int i;
    vector3 v;
    OBAtom atom;
    double x,y,z,pcharge;
//.........这里部分代码省略.........

  bool TinkerFormat::WriteMolecule(OBBase* pOb, OBConversion* pConv)
  {
    OBMol* pmol = dynamic_cast<OBMol*>(pOb);
    if(pmol==NULL)
      return false;

    //Define some references so we can use the old parameter names
    ostream &ofs = *pConv->GetOutStream();
    OBMol &mol = *pmol;
    bool mm2Types = false;
    bool mmffTypes = pConv->IsOption("m",OBConversion::OUTOPTIONS) != NULL;
    bool mm3Types = pConv->IsOption("3",OBConversion::OUTOPTIONS) != NULL;
    bool classTypes = pConv->IsOption("c", OBConversion::OUTOPTIONS) != NULL;

    unsigned int i;
    char buffer[BUFF_SIZE];
    OBBond *bond;
    vector<OBBond*>::iterator j;

    // Before we try output of MMFF94 atom types, check if it works
    OBForceField *ff = OpenBabel::OBForceField::FindForceField("MMFF94");
    if (mmffTypes && ff && ff->Setup(mol))
      mmffTypes = ff->GetAtomTypes(mol);
    else
      mmffTypes = false; // either the force field isn't available, or it doesn't work

    if (!mmffTypes && !mm3Types && !classTypes) {
      snprintf(buffer, BUFF_SIZE, "%6d %-20s   MM2 parameters\n",mol.NumAtoms(),mol.GetTitle());
      mm2Types = true;
    }
    else if (mm3Types)
      snprintf(buffer, BUFF_SIZE, "%6d %-20s   MM3 parameters\n",mol.NumAtoms(),mol.GetTitle());
    else if (classTypes)
      snprintf(buffer, BUFF_SIZE, "%6d %-20s   Custom parameters\n",mol.NumAtoms(),mol.GetTitle());
    else
      snprintf(buffer, BUFF_SIZE, "%6d %-20s   MMFF94 parameters\n",mol.NumAtoms(),mol.GetTitle());
    ofs << buffer;

    ttab.SetFromType("INT");

    OBAtom *atom;
    string str,str1;
    int atomType;
    for(i = 1;i <= mol.NumAtoms(); i++)
      {
        atom = mol.GetAtom(i);
        str = atom->GetType();
        atomType = 0; // Something is very wrong if this doesn't get set below

        if (mm2Types) {
          ttab.SetToType("MM2");
          ttab.Translate(str1,str);
          atomType = atoi((char*)str1.c_str());
        }
        if (mmffTypes) {
          // Override the MM2 typing
          OBPairData *type = (OpenBabel::OBPairData*)atom->GetData("FFAtomType");
          if (type) {
            str1 = type->GetValue().c_str();
            atomType = atoi((char*)str1.c_str());
          }
        }
        if (mm3Types) {
          // convert to integer for MM3 typing
          atomType = SetMM3Type(atom);
        }
        if (classTypes) {
          // Atom classes are set by the user, so use those
          OBGenericData *data = atom->GetData("Atom Class");
          if (data) {
            OBPairInteger* acdata = dynamic_cast<OBPairInteger*>(data); // Could replace with C-style cast if willing to live dangerously
            if (acdata) {
              int ac = acdata->GetGenericValue();
              if (ac >= 0)
                atomType = ac;
            }
          }
        }

        snprintf(buffer, BUFF_SIZE, "%6d %2s  %12.6f%12.6f%12.6f %5d",
                 i,
                 OBElements::GetSymbol(atom->GetAtomicNum()),
                 atom->GetX(),
                 atom->GetY(),
                 atom->GetZ(),
                 atomType);
        ofs << buffer;

        for (bond = atom->BeginBond(j); bond; bond = atom->NextBond(j))
          {
            snprintf(buffer, BUFF_SIZE, "%6d", (bond->GetNbrAtom(atom))->GetIdx());
            ofs << buffer;
          }

        ofs << endl;
      }

    return(true);
  }

  OBBase* OBMol::DoTransformations(const std::map<std::string, std::string>* pOptions, OBConversion* pConv)
  {
    // Perform any requested transformations
    // on a OBMol
    //The input map has option letters or name as the key and
    //any associated text as the value.
    //For normal(non-filter) transforms:
    // returns a pointer to the OBMol (this) if ok or NULL if not.
    //For filters returns a pointer to the OBMol (this) if there is a  match,
    //and NULL when not and in addition the OBMol object is deleted NULL.

    //This is now a virtual function. The OBBase version just returns the OBMol pointer.
    //This is declared in mol.h

    //The filter options, s and v allow a obgrep facility.
    //Used together they must both be true to allow a molecule through.

    //Parse GeneralOptions
    if(pOptions->empty())
      return this;

    // DoOps calls Do() for each of the plugin options in the map
    // It normally returns true, even if there are no options but
    // can return false if one of the options decides that the
    // molecule should not be output. If it is a filtering op, it
    // should delete the molecule itself (unlike the -s, --filter options,
    // which delete it in this function).
    if(!OBOp::DoOps(this, pOptions, pConv))
      return (OBBase *)NULL;

    bool ret=true;

    map<string,string>::const_iterator itr, itr2;

    if(pOptions->find("b")!=pOptions->end())
      if(!ConvertDativeBonds())
        ret=false;

    if(pOptions->find("d")!=pOptions->end())
      if(!DeleteHydrogens())
        ret=false;

    if(pOptions->find("h")!=pOptions->end())
      if(!AddHydrogens(false, false))
        ret=false;

    if(pOptions->find("r")!=pOptions->end()) {
      StripSalts();
      ret = true;
    }

    itr = pOptions->find("p");
    if(itr!=pOptions->end()) {
      double pH = strtod(itr->second.c_str(), 0);
      if(!AddHydrogens(false, true, pH))
        ret=false;
    }

    if(pOptions->find("c")!=pOptions->end())
      Center();

    itr = pOptions->find("title"); //Replaces title
    if(itr!=pOptions->end())
      SetTitle(itr->second.c_str());

    itr = pOptions->find("addtotitle"); //Appends text to title
    if(itr!=pOptions->end())
      {
        string title(GetTitle());
        title += itr->second;
        SetTitle(title.c_str());
      }

/*    itr = pOptions->find("addformula"); //Appends tab + formula to title
    if(itr!=pOptions->end())
      {
        string title(GetTitle());
        title += '\t' + GetSpacedFormula(1,"");//actually unspaced
        SetTitle(title.c_str());
      }
*/
    //Add an extra property to the molecule.
    //Parameter has atrribute and value separated by a space
    itr = pOptions->find("property");
    if(itr!=pOptions->end())
      {
        string txt(itr->second);
        string::size_type pos = txt.find(' ');
        if(pos==string::npos)
          {
            obErrorLog.ThrowError(__FUNCTION__, "Missing property value", obError);
            ret=false;
          }
        else
          {
            string attr(txt.substr(0,pos)), val(txt.substr(pos+1));
            //Update value if it already exists
            OBPairData* dp = dynamic_cast<OBPairData*>(GetData(attr));
            if(dp) {
              dp->SetValue(val);
//.........这里部分代码省略.........

//.........这里部分代码省略.........
            tokenize(vs, buffer);
            while (vs.size() == 5) {
                double s = atof(vs[4].c_str());
                RotatoryStrengthsVelocity.push_back(s);
                ifs.getline(buffer, BUFF_SIZE);
                tokenize(vs, buffer);
            }
        }
        else if(strstr(buffer, "       state          X           Y           Z     R(length)")) {
            // Rotatory Strengths
            ifs.getline(buffer, BUFF_SIZE); // First entry
            tokenize(vs, buffer);
            while (vs.size() == 5) {
                double s = atof(vs[4].c_str());
                RotatoryStrengthsLength.push_back(s);
                ifs.getline(buffer, BUFF_SIZE);
                tokenize(vs, buffer);
            }
        }

        else if (strstr(buffer, "Forces (Hartrees/Bohr)"))
        {
            ifs.getline(buffer, BUFF_SIZE); // column headers
            ifs.getline(buffer, BUFF_SIZE); // ------
            ifs.getline(buffer, BUFF_SIZE); // real data
        }

        else if (strstr(buffer, "Isotropic = ")) // NMR shifts
        {
            tokenize(vs, buffer);
            if (vs.size() >= 4)
            {
                atom = mol.GetAtom(atoi(vs[0].c_str()));
                OBPairData *nmrShift = new OBPairData();
                nmrShift->SetAttribute("NMR Isotropic Shift");

                string shift = vs[4].c_str();
                nmrShift->SetValue(shift);

                atom->SetData(nmrShift);
            }
        }
        else if(strstr(buffer,"SCF Done:") != NULL)
        {
            tokenize(vs,buffer);
            mol.SetEnergy(atof(vs[4].c_str()) * HARTEE_TO_KCALPERMOL);
            confEnergies.push_back(mol.GetEnergy());
        }
        /* Temporarily commented out until the handling of energy in OBMol is sorted out
                // MP2 energies also use a different syntax

                // PM3 energies use a different syntax
                else if(strstr(buffer,"E (Thermal)") != NULL)
                  {
                    ifs.getline(buffer,BUFF_SIZE); //Headers
                    ifs.getline(buffer,BUFF_SIZE); //Total energy; what we want
                    tokenize(vs,buffer);
                    mol.SetEnergy(atof(vs[1].c_str()));
                    confEnergies.push_back(mol.GetEnergy());
                    }
        */
        else if(strstr(buffer,"Standard basis:") != NULL)
        {
            add_unique_pairdata_to_mol(&mol,"basis",buffer,2);
        }
        else if(strstr(buffer,"Zero-point correction=") != NULL)