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];
}
bool OutputFormat::ReadMolecule(OBBase* pOb, OBConversion* pConv)
{
// so we want to read through the file until we can figure out
// what program actually created it
// if we get to the end, emit a warning
istream &ifs = *pConv->GetInStream();
char buffer[BUFF_SIZE];
OBFormat *pFormat = NULL;
std::string formatName;
// the detection strings are from the Chemical MIME project
// http://chemical-mime.sourceforge.net/chemical-mime-data.html
while (ifs.getline(buffer,BUFF_SIZE)) {
if ((strstr(buffer,"GAMESS execution script") != NULL) ||
(strstr(buffer,"PC GAMESS") != NULL) ||
(strstr(buffer,"GAMESS VERSION") != NULL)) {
// GAMESS output
formatName = "gamout";
break;
} else if (strstr(buffer,"=== G A M E S S - U K === ") != NULL) {
// GAMESS-UK output
formatName = "gukout";
break;
} else if (strstr(buffer,"Gaussian, Inc") != NULL) {
// Gaussian output
formatName = "g03";
break;
} else if (strstr(buffer,"GENERAL UTILITY LATTICE PROGRAM") != NULL) {
// GULP output -- not currently supported
break;
} else if (strstr(buffer,"MOPAC") != NULL) {
// MOPAC output
formatName = "mopout";
break;
} else if (strstr(buffer,"Program PWSCF") != NULL) {
// PWSCF
formatName = "pwscf";
break;
} else if (strstr(buffer,"Welcome to Q-Chem") != NULL) {
// Q-Chem output
formatName = "qcout";
break;
} else if (strstr(buffer,"Amsterdam Density Functional") != NULL) {
// ADF output
// Determine the kind of ADF output
while (ifs.getline(buffer, BUFF_SIZE)) {
if (strstr(buffer, "| A D F |") != NULL) {
formatName = "adfout";
break;
} else if (strstr(buffer, "| B A N D |") != NULL) {
formatName = "adfband";
break;
} else if (strstr(buffer, "| D F T B |") != NULL) {
formatName = "adfdftb";
break;
} else if (strstr(buffer, "DFTB Engine") != NULL) {
// "| D F T B |" is no longer printed in ADF 2018
// Hopefully, "DFTB Engine" will work fine...
formatName = "adfdftb";
break;
}
}
break;
} else if (strstr(buffer,"Northwest Computational Chemistry") != NULL) {
// NWChem output
formatName = "nwo";
break;
} else if (strstr(buffer,"MPQC: Massively Parallel Quantum Chemistry") != NULL) {
// MPQC output
formatName = "mpqc";
break;
} else if (strstr(buffer,"PROGRAM SYSTEM MOLPRO") != NULL) {
// MOLPRO output
formatName = "mpo";
break;
} else if ((strstr(buffer,"Schrodinger, Inc.") != NULL) &&
(strstr(buffer,"Jaguar") != NULL)) {
// Jaguar
formatName = "jout";
break;
} else if (strstr(buffer, "ABINIT") != NULL) {
// Abinit
formatName = "abinit";
break;
} else if (strstr(buffer, "ACES2") != NULL) {
// ACESII
formatName = "acesout";
break;
} else if (strstr(buffer, "CRYSTAL06") != NULL ||
strstr(buffer, "CRYSTAL09") != NULL) {
// CRYSTAL09
formatName = "c09out";
break;
} else if (strstr(buffer, "* O R C A *") != NULL) {
// ORCA
formatName = "orca";
break;
} else if (strstr(buffer, "WELCOME TO SIESTA") != NULL) {
// SIESTA
formatName = "siesta";
break;
}
}
// if we assigned something above, let's try to find it
if (formatName.length())
pFormat = pConv->FindFormat(formatName);
if (pFormat) {
ifs.seekg (0, ios::beg); // reset the stream to the beginning
ifs.clear();
bool success = pFormat->ReadMolecule(pOb, pConv);
// Tag the molecule with the format (e.g., if a program wants to know the kind of "out" or "log" file)
// We have to do this *after* ReadMolecule returns, or the data might be cleared
if (pOb) {
OBPairData *dp = new OBPairData;
dp->SetAttribute("File Format");
dp->SetValue(formatName);
dp->SetOrigin(fileformatInput);
pOb->SetData(dp);
}
return success;
}
obErrorLog.ThrowError(__FUNCTION__,
"Problems reading an output file: Could not determine the format of this file. Please report it to the openbabel-discuss @ lists.sourceforge.net mailing list.", obError);
return(false); // we couldn't figure out the format
}
bool RXNFormat::ReadMolecule(OBBase* pOb, OBConversion* pConv)
{
OBMol* pmol = pOb->CastAndClear<OBMol>();
if (pmol == NULL)
return false;
OBFormat* pMolFormat = pConv->FindFormat("MOL");
if (pMolFormat==NULL)
return false;
istream &ifs = *pConv->GetInStream();
string ln;
// When MDLFormat reads the last product it may also read and discard
// the line with $RXN for the next reaction. But it then sets $RXNread option.
if(pConv->IsOption("$RXNread"))
pConv->RemoveOption("$RXNread", OBConversion::OUTOPTIONS);
else
{
if (!getline(ifs,ln))
return(false);
if(Trim(ln).find("$RXN")!=0)
return false; //Has to start with $RXN
}
if (!getline(ifs,ln))
return false; //reaction title
pmol->SetTitle(Trim(ln));
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;
}
//! \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);
mol.SetData(dp);
m_partialCharges.clear();
m_partialCharges.reserve(N);
m_formalCharges.clear();
m_formalCharges.reserve(N);
for (i = 0; i < N; i ++)
{
atom = mol.GetAtom(i + 1);
atom->SetPartialCharge(x(i));
m_partialCharges.push_back(x(i));
m_formalCharges.push_back(atom->GetFormalCharge());
}
obErrorLog.ThrowError(__FUNCTION__, "EQEq charges successfully assigned.", obInfo);
return true;
}
bool OpAlign::Do(OBBase* pOb, const char* OptionText, OpMap* pmap, OBConversion* pConv)
{
OBMol* pmol = dynamic_cast<OBMol*>(pOb);
if(!pmol)
return false;
map<string,string>::const_iterator itr;
// Is there an -s option?
if(pConv->IsFirstInput())
{
_pOpIsoM = NULL; //assume no -s option
itr = pmap->find("s");
if(itr!=pmap->end())
{
//There is an -s option; check it is ok
_pOpIsoM = static_cast<OpNewS*>(OBOp::FindType("s"));
_stext = itr->second; //get its parameter(s)
if(!_pOpIsoM || _stext.empty())
{
obErrorLog.ThrowError(__FUNCTION__,
"No parameter on -s option, or its OBOp version is not loaded", obError);
pConv->SetOneObjectOnly(); //to finish
return false;
}
}
}
// If the output format is a 2D depiction format, then we should align
// on the 2D coordinates and not the 3D coordinates (if present). This
//means we need to generate the 2D coordinates at this point.
if(pmol->GetDimension()==3 && (pConv->GetOutFormat()->Flags() & DEPICTION2D))
{
OBOp* pgen = OBOp::FindType("gen2D");
if(pgen)
pgen->Do(pmol);
}
// All molecules must have coordinates, so add them if 0D
// They may be added again later when gen2D or gen3D is called, but they will be the same.
// It would be better if this op was called after them, which would happen
// if its name was alphabetically after "gen" (and before "s").
if(pmol->GetDimension()==0)
{
//Will the coordinates be 2D or 3D?
itr = pmap->find("gen3D");
OBOp* pgen = (itr==pmap->end()) ? OBOp::FindType("gen2D") : OBOp::FindType("gen3D");
if(pgen)
pgen->Do(pmol);
}
//Do the alignment in 2D if the output format is svg, png etc. and there is no -xn option
if(pmol->GetDimension()==3 && pConv && !pConv->IsOption("n"))
{
OBFormat* pOutFormat = pConv->GetOutFormat();
if(pOutFormat->Flags() & DEPICTION2D)
{
OBOp* pgen = OBOp::FindType("gen2D");
if(pgen)
pgen->Do(pmol);
}
}
if(pConv->IsFirstInput() || _refMol.NumAtoms()==0)
{
_refvec.clear();
// Reference molecule is basically the first molecule
_refMol = *pmol;
if(!_pOpIsoM)
//no -s option. Use a molecule reference.
_align.SetRefMol(_refMol);
else
{
//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;
}
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);
dp->SetOrigin(userInput);
}
else {
// Pair did not exist; make new one
dp = new OBPairData;
dp->SetAttribute(attr);
dp->SetValue(val);
dp->SetOrigin(userInput);
SetData(dp);
}
}
}
itr = pOptions->find("add"); //adds new properties from descriptors in list
if(itr!=pOptions->end())
OBDescriptor::AddProperties(this, itr->second);
itr = pOptions->find("delete"); //deletes the specified properties
if(itr!=pOptions->end())
OBDescriptor::DeleteProperties(this, itr->second);
itr = pOptions->find("append"); //Appends values of descriptors or properties to title
if(itr!=pOptions->end())
{
string title(GetTitle());
title += OBDescriptor::GetValues(this, itr->second);
if(ispunct(title[0]))
title[0]=' ';//a leading punct char is used only as a separator, not at start
SetTitle(Trim(title).c_str());
}
//Filter using OBDescriptor comparison and (older) SMARTS tests
//Continue only if previous test was true.
bool fmatch = true;
itr = pOptions->find("filter");
if(itr!=pOptions->end())
{
std::istringstream optionText(itr->second);
fmatch = OBDescriptor::FilterCompare(this, optionText, false);
}
if(fmatch)
{
itr = pOptions->find("v");
if(itr!=pOptions->end() && !itr->second.empty())
{
//inverse match quoted SMARTS string which follows
OBSmartsPattern sp;
sp.Init(itr->second);
fmatch = !sp.Match(*this); //(*pmol) ;
}
}
if(fmatch)
{
itr = pOptions->find("s");
if(itr!=pOptions->end() && !itr->second.empty())
{
//SMARTS filter
//If exactmatch option set (probably in fastsearchformat) the
//number of atoms in the pattern (passed as a string in the option text)
//has to be the same as in the molecule.
itr2 = pOptions->find("exactmatch");
if(itr2!=pOptions->end() && NumHvyAtoms()!=atoi(itr2->second.c_str()))
fmatch=false;
else
{
//match quoted SMARTS string which follows
OBSmartsPattern sp;
sp.Init(itr->second.c_str());
fmatch = sp.Match(*this);
}
}
}
if(!fmatch)
{
//filter failed: delete OBMol and return NULL
delete this;
return NULL;
}
else
{
if(ret==false)
{
obErrorLog.ThrowError(__FUNCTION__, "Error executing an option", obError);
delete this; //added 9March2006
return NULL;
}
else
return this;
}
}
// Reading Gaussian output has been tested for G98 and G03 to some degree // If you have problems (or examples of older output), please contact // the [email protected] mailing list and/or post a bug bool GaussianOutputFormat::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; const char* title = pConv->GetTitle(); char buffer[BUFF_SIZE]; string str,str1,str2,thermo_method; double x,y,z; OBAtom *atom; vector<string> vs,vs2; int total_charge = 0; unsigned int spin_multiplicity = 1; bool hasPartialCharges = false; string chargeModel; // descriptor for charges (e.g. "Mulliken") // Variable for G2/G3/G4 etc. calculations double ezpe,Hcorr,Gcorr,E0,CV; bool ezpe_set=false,Hcorr_set=false,Gcorr_set=false,E0_set=false,CV_set=false; double temperature = 0; /* Kelvin */ std::vector<double> Scomponents; // Electrostatic potential OBFreeGrid *esp = NULL; // coordinates of all steps // Set conformers to all coordinates we adopted std::vector<double*> vconf; // index of all frames/conformers std::vector<double> coordinates; // coordinates in each frame int natoms = 0; // number of atoms -- ensure we don't go to a new job with a different molecule // OBConformerData stores information about multiple steps // we can change attribute later if needed (e.g., IRC) OBConformerData *confData = new OBConformerData(); confData->SetOrigin(fileformatInput); std::vector<unsigned short> confDimensions = confData->GetDimension(); // to be fair, set these all to 3D std::vector<double> confEnergies = confData->GetEnergies(); std::vector< std::vector< vector3 > > confForces = confData->GetForces(); //Vibrational data std::vector< std::vector< vector3 > > Lx; std::vector<double> Frequencies, Intensities; //Rotational data std::vector<double> RotConsts(3); int RotSymNum=1; OBRotationData::RType RotorType = OBRotationData::UNKNOWN; // Translation vectors (if present) vector3 translationVectors[3]; int numTranslationVectors = 0; //Electronic Excitation data std::vector<double> Forces, Wavelengths, EDipole, RotatoryStrengthsVelocity, RotatoryStrengthsLength; // Orbital data std::vector<double> orbitals; std::vector<std::string> symmetries; int aHOMO, bHOMO, betaStart; aHOMO = bHOMO = betaStart = -1; int i=0; bool no_symmetry=false; char coords_type[25]; //Prescan file to find second instance of "orientation:" //This will be the kind of coords used in the chk/fchk file //Unless the "nosym" keyword has been requested while (ifs.getline(buffer,BUFF_SIZE)) { if (strstr(buffer,"Symmetry turned off by external request.") != NULL) { // The "nosym" keyword has been requested no_symmetry = true; } if (strstr(buffer,"orientation:") !=NULL) { i++; tokenize (vs, buffer); // gotta check what types of orientation are present strncpy (coords_type, vs[0].c_str(), 24); strcat (coords_type, " orientation:"); } if ((no_symmetry && i==1) || i==2) break; } // Reset end-of-file pointers etc. ifs.clear(); ifs.seekg(0); //rewind mol.BeginModify(); while (ifs.getline(buffer,BUFF_SIZE)) { if(strstr(buffer, "Entering Gaussian") != NULL) { //Put some metadata into OBCommentData string comment("Gaussian "); if(NULL != strchr(buffer,'=')) { comment += strchr(buffer,'=')+2; comment += ""; for(unsigned i=0; i<115 && ifs; ++i) { ifs.getline(buffer,BUFF_SIZE); if(strstr(buffer,"Revision") != NULL) { if (buffer[strlen(buffer)-1] == ',') { buffer[strlen(buffer)-1] = '\0'; } add_unique_pairdata_to_mol(&mol,"program",buffer,0); } else if(buffer[1]=='#') { //the line describing the method comment += buffer; OBCommentData *cd = new OBCommentData; cd->SetData(comment); cd->SetOrigin(fileformatInput); mol.SetData(cd); tokenize(vs,buffer); if (vs.size() > 1) { char *str = strdup(vs[1].c_str()); char *ptr = strchr(str,'/'); if (NULL != ptr) { *ptr = ' '; add_unique_pairdata_to_mol(&mol,"basis",ptr,0); *ptr = '\0'; add_unique_pairdata_to_mol(&mol,"method",str,0); } } break; } } } } else if (strstr(buffer,"Multiplicity") != NULL) { tokenize(vs, buffer, " \t\n"); if (vs.size() == 6) { total_charge = atoi(vs[2].c_str()); spin_multiplicity = atoi(vs[5].c_str()); } ifs.getline(buffer,BUFF_SIZE); } else if (strstr(buffer, coords_type) != NULL) { numTranslationVectors = 0; // ignore old translationVectors ifs.getline(buffer,BUFF_SIZE); // --------------- ifs.getline(buffer,BUFF_SIZE); // column headings ifs.getline(buffer,BUFF_SIZE); // column headings ifs.getline(buffer,BUFF_SIZE); // --------------- ifs.getline(buffer,BUFF_SIZE); tokenize(vs,buffer); while (vs.size()>4) { int corr = vs.size()==5 ? -1 : 0; //g94; later versions have an extra column x = atof((char*)vs[3+corr].c_str()); y = atof((char*)vs[4+corr].c_str()); z = atof((char*)vs[5+corr].c_str()); int atomicNum = atoi((char*)vs[1].c_str()); if (atomicNum > 0) // translation vectors are "-2" { if (natoms == 0) { // first time reading the molecule, create each atom atom = mol.NewAtom(); atom->SetAtomicNum(atoi((char*)vs[1].c_str())); } coordinates.push_back(x); coordinates.push_back(y); coordinates.push_back(z); } else { translationVectors[numTranslationVectors++].Set(x, y, z); } if (!ifs.getline(buffer,BUFF_SIZE)) { break; } tokenize(vs,buffer); } // done with reading atoms natoms = mol.NumAtoms(); if(natoms==0) return false; // malloc / memcpy double *tmpCoords = new double [(natoms)*3]; memcpy(tmpCoords, &coordinates[0], sizeof(double)*natoms*3); vconf.push_back(tmpCoords); coordinates.clear(); confDimensions.push_back(3); // always 3D -- OBConformerData allows mixing 2D and 3D structures } else if(strstr(buffer,"Dipole moment") != NULL) { ifs.getline(buffer,BUFF_SIZE); // actual components X ### Y #### Z ### tokenize(vs,buffer); if (vs.size() >= 6) { OBVectorData *dipoleMoment = new OBVectorData; dipoleMoment->SetAttribute("Dipole Moment"); double x, y, z; x = atof(vs[1].c_str()); y = atof(vs[3].c_str()); z = atof(vs[5].c_str()); dipoleMoment->SetData(x, y, z); dipoleMoment->SetOrigin(fileformatInput); mol.SetData(dipoleMoment); } if (!ifs.getline(buffer,BUFF_SIZE)) break; } else if(strstr(buffer,"Traceless Quadrupole moment") != NULL) { ifs.getline(buffer,BUFF_SIZE); // actual components XX ### YY #### ZZ ### tokenize(vs,buffer); ifs.getline(buffer,BUFF_SIZE); // actual components XY ### XZ #### YZ ### tokenize(vs2,buffer); if ((vs.size() >= 6) && (vs2.size() >= 6)) { double Q[3][3]; OpenBabel::OBMatrixData *quadrupoleMoment = new OpenBabel::OBMatrixData; Q[0][0] = atof(vs[1].c_str()); Q[1][1] = atof(vs[3].c_str()); Q[2][2] = atof(vs[5].c_str()); Q[1][0] = Q[0][1] = atof(vs2[1].c_str()); Q[2][0] = Q[0][2] = atof(vs2[3].c_str()); Q[2][1] = Q[1][2] = atof(vs2[5].c_str()); matrix3x3 quad(Q); quadrupoleMoment->SetAttribute("Traceless Quadrupole Moment"); quadrupoleMoment->SetData(quad); quadrupoleMoment->SetOrigin(fileformatInput); mol.SetData(quadrupoleMoment); } if (!ifs.getline(buffer,BUFF_SIZE)) break; } else if(strstr(buffer,"Exact polarizability") != NULL) { // actual components XX, YX, YY, XZ, YZ, ZZ tokenize(vs,buffer); if (vs.size() >= 8) { double Q[3][3]; OpenBabel::OBMatrixData *pol_tensor = new OpenBabel::OBMatrixData; Q[0][0] = atof(vs[2].c_str()); Q[1][1] = atof(vs[4].c_str()); Q[2][2] = atof(vs[7].c_str()); Q[1][0] = Q[0][1] = atof(vs[3].c_str()); Q[2][0] = Q[0][2] = atof(vs[5].c_str()); Q[2][1] = Q[1][2] = atof(vs[6].c_str()); matrix3x3 pol(Q); pol_tensor->SetAttribute("Exact polarizability"); pol_tensor->SetData(pol); pol_tensor->SetOrigin(fileformatInput); mol.SetData(pol_tensor); } if (!ifs.getline(buffer,BUFF_SIZE)) break; } else if(strstr(buffer,"Total atomic charges") != NULL || strstr(buffer,"Mulliken atomic charges") != NULL) { hasPartialCharges = true; chargeModel = "Mulliken"; ifs.getline(buffer,BUFF_SIZE); // column headings ifs.getline(buffer,BUFF_SIZE); tokenize(vs,buffer); while (vs.size() >= 3 && strstr(buffer,"Sum of ") == NULL) { atom = mol.GetAtom(atoi(vs[0].c_str())); if (!atom) break; atom->SetPartialCharge(atof(vs[2].c_str())); if (!ifs.getline(buffer,BUFF_SIZE)) break; tokenize(vs,buffer); } } else if (strstr(buffer, "Atomic Center") != NULL) { // Data points for ESP calculation tokenize(vs,buffer); if (NULL == esp) esp = new OpenBabel::OBFreeGrid(); if (vs.size() == 8) { esp->AddPoint(atof(vs[5].c_str()),atof(vs[6].c_str()), atof(vs[7].c_str()),0); } else if (vs.size() > 5) { double x,y,z; if (3 == sscanf(buffer+32,"%10lf%10lf%10lf",&x,&y,&z)) { esp->AddPoint(x,y,z,0); } } } else if (strstr(buffer, "ESP Fit Center") != NULL) { // Data points for ESP calculation tokenize(vs,buffer); if (NULL == esp) esp = new OpenBabel::OBFreeGrid(); if (vs.size() == 9) { esp->AddPoint(atof(vs[6].c_str()),atof(vs[7].c_str()), atof(vs[8].c_str()),0); } else if (vs.size() > 6) { double x,y,z; if (3 == sscanf(buffer+32,"%10lf%10lf%10lf",&x,&y,&z)) { esp->AddPoint(x,y,z,0); } } } else if (strstr(buffer, "Electrostatic Properties (Atomic Units)") != NULL) { int i,np; OpenBabel::OBFreeGridPoint *fgp; OpenBabel::OBFreeGridPointIterator fgpi; for(i=0; (i<5); i++) { ifs.getline(buffer,BUFF_SIZE); // skip line } // Assume file is correct and that potentials are present // where they should. np = esp->NumPoints(); fgpi = esp->BeginPoints(); i = 0; for(fgp = esp->BeginPoint(fgpi); (NULL != fgp); fgp = esp->NextPoint(fgpi)) { ifs.getline(buffer,BUFF_SIZE); tokenize(vs,buffer); if (vs.size() >= 2) { fgp->SetV(atof(vs[2].c_str())); i++; } } if (i == np) { esp->SetAttribute("Electrostatic Potential"); mol.SetData(esp); } else { cout << "Read " << esp->NumPoints() << " ESP points i = " << i << "\n"; } } else if (strstr(buffer, "Charges from ESP fit") != NULL) { hasPartialCharges = true; chargeModel = "ESP"; ifs.getline(buffer,BUFF_SIZE); // Charge / dipole line ifs.getline(buffer,BUFF_SIZE); // column header ifs.getline(buffer,BUFF_SIZE); // real charges tokenize(vs,buffer); while (vs.size() >= 3 && strstr(buffer,"-----") == NULL) { atom = mol.GetAtom(atoi(vs[0].c_str())); if (!atom) break; atom->SetPartialCharge(atof(vs[2].c_str())); if (!ifs.getline(buffer,BUFF_SIZE)) break; tokenize(vs,buffer); } } else if(strstr(buffer,"Natural Population") != NULL) { hasPartialCharges = true; chargeModel = "NBO"; ifs.getline(buffer,BUFF_SIZE); // column headings ifs.getline(buffer,BUFF_SIZE); // again ifs.getline(buffer,BUFF_SIZE); // again (-----) ifs.getline(buffer,BUFF_SIZE); // real data tokenize(vs,buffer); while (vs.size() >= 3 && strstr(buffer,"=====") == NULL) { atom = mol.GetAtom(atoi(vs[1].c_str())); if (!atom) break; atom->SetPartialCharge(atof(vs[2].c_str())); if (!ifs.getline(buffer,BUFF_SIZE)) break; tokenize(vs,buffer); } } else if(strstr(buffer, " Frequencies -- ")) //vibrational frequencies { //The info should appear only once as several blocks starting with this line tokenize(vs, buffer); for(unsigned int i=2; i<vs.size(); ++i) Frequencies.push_back(atof(vs[i].c_str())); ifs.getline(buffer,BUFF_SIZE); //Red. masses ifs.getline(buffer,BUFF_SIZE); //Frc consts ifs.getline(buffer,BUFF_SIZE); //IR Inten tokenize(vs, buffer); for(unsigned int i=3; i<vs.size(); ++i) Intensities.push_back(atof(vs[i].c_str())); ifs.getline(buffer, BUFF_SIZE); // column labels or Raman intensity if(strstr(buffer, "Raman Activ")) { ifs.getline(buffer, BUFF_SIZE); // Depolar (P) ifs.getline(buffer, BUFF_SIZE); // Depolar (U) ifs.getline(buffer, BUFF_SIZE); // column labels } ifs.getline(buffer, BUFF_SIZE); // actual displacement data tokenize(vs, buffer); vector<vector3> vib1, vib2, vib3; double x, y, z; while(vs.size() >= 5) { for (unsigned int i = 2; i < vs.size()-2; i += 3) { x = atof(vs[i].c_str()); y = atof(vs[i+1].c_str()); z = atof(vs[i+2].c_str()); if (i == 2) vib1.push_back(vector3(x, y, z)); else if (i == 5) vib2.push_back(vector3(x, y, z)); else if (i == 8) vib3.push_back(vector3(x, y, z)); } if (!ifs.getline(buffer, BUFF_SIZE)) break; tokenize(vs,buffer); } Lx.push_back(vib1); if (vib2.size()) Lx.push_back(vib2); if (vib3.size()) Lx.push_back(vib3); } else if(strstr(buffer, " This molecule is "))//rotational data { if(strstr(buffer, "asymmetric")) RotorType = OBRotationData::ASYMMETRIC; else if(strstr(buffer, "symmetric")) RotorType = OBRotationData::SYMMETRIC; else if(strstr(buffer, "linear")) RotorType = OBRotationData::LINEAR; else RotorType = OBRotationData::UNKNOWN; ifs.getline(buffer,BUFF_SIZE); //symmetry number tokenize(vs, buffer); RotSymNum = atoi(vs[3].c_str()); } else if(strstr(buffer, "Rotational constant")) { tokenize(vs, buffer); RotConsts.clear(); for (unsigned int i=3; i<vs.size(); ++i) RotConsts.push_back(atof(vs[i].c_str())); } else if(strstr(buffer, "alpha electrons")) // # of electrons / orbital { tokenize(vs, buffer); if (vs.size() == 6) { // # alpha electrons # beta electrons aHOMO = atoi(vs[0].c_str()); bHOMO = atoi(vs[3].c_str()); } } else if(strstr(buffer, "rbital symmetries")) // orbital symmetries { symmetries.clear(); std::string label; // used as a temporary to remove "(" and ")" from labels int iii,offset = 0; bool bDoneSymm; // Extract both Alpha and Beta symmetries ifs.getline(buffer, BUFF_SIZE); // skip the current line for(iii=0; (iii<2); iii++) { if (strstr(buffer, "electronic state")) break; // We've gone too far! while (!ifs.eof() && ((NULL != strstr(buffer,"Alpha")) || (NULL != strstr(buffer,"Beta")))) { // skip the Alpha: and Beta: title lines ifs.getline(buffer, BUFF_SIZE); } do { bDoneSymm = (NULL == strstr(buffer, "(")); if (!bDoneSymm) { tokenize(vs, buffer); if ((NULL != strstr(buffer, "Occupied")) || (NULL != strstr(buffer, "Virtual"))) { offset = 1; // skip first token } else { offset = 0; } for (unsigned int i = offset; i < vs.size(); ++i) { label = vs[i].substr(1, vs[i].length() - 2); symmetries.push_back(label); } ifs.getline(buffer, BUFF_SIZE); // get a new line if we've been reading symmetries } // don't read a new line if we're done with symmetries } while (!ifs.eof() && !bDoneSymm); } // end alpha/beta section } else if (strstr(buffer, "Alpha") && strstr(buffer, ". eigenvalues --")) { orbitals.clear(); betaStart = 0; while (strstr(buffer, ". eigenvalues --")) { tokenize(vs, buffer); if (vs.size() < 4) break; if (vs[0].find("Beta") !=string::npos && betaStart == 0) // mark where we switch from alpha to beta betaStart = orbitals.size(); for (unsigned int i = 4; i < vs.size(); ++i) { orbitals.push_back(atof(vs[i].c_str())); } ifs.getline(buffer, BUFF_SIZE); } } else if(strstr(buffer, " Excited State")) // Force and wavelength data { // The above line appears for each state, so just append the info to the vectors tokenize(vs, buffer); if (vs.size() >= 9) { double wavelength = atof(vs[6].c_str()); double force = atof(vs[8].substr(2).c_str()); // remove the "f=" part Forces.push_back(force); Wavelengths.push_back(wavelength); } } else if(strstr(buffer, " Ground to excited state Transition electric dipole moments (Au):")) // Electronic dipole moments { ifs.getline(buffer, BUFF_SIZE); // Headings ifs.getline(buffer, BUFF_SIZE); // First entry tokenize(vs, buffer); while (vs.size() == 5) { double s = atof(vs[4].c_str()); EDipole.push_back(s); ifs.getline(buffer, BUFF_SIZE); tokenize(vs, buffer); } } else if(strstr(buffer, " state X Y Z R(velocity)")) { // Rotatory Strengths ifs.getline(buffer, BUFF_SIZE); // First entry 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) { tokenize(vs,buffer); ezpe = atof(vs[2].c_str()); ezpe_set = true; } else if(strstr(buffer,"Thermal correction to Enthalpy=") != NULL) { tokenize(vs,buffer); Hcorr = atof(vs[4].c_str()); Hcorr_set = true; } else if(strstr(buffer,"Thermal correction to Gibbs Free Energy=") != NULL) { tokenize(vs,buffer); Gcorr = atof(vs[6].c_str()); Gcorr_set = true; } else if (strstr(buffer,"CV") != NULL) { ifs.getline(buffer,BUFF_SIZE); //Headers ifs.getline(buffer,BUFF_SIZE); //Total heat capacity tokenize(vs,buffer); if (vs.size() == 4) { if (vs[0].compare("Total") == 0) { CV = atof(vs[2].c_str()); CV_set = true; } } ifs.getline(buffer,BUFF_SIZE); //Electronic ifs.getline(buffer,BUFF_SIZE); //Translational tokenize(vs,buffer); if ((vs.size() == 4) && (vs[0].compare("Translational") == 0) ) { Scomponents.push_back(atof(vs[3].c_str())); } ifs.getline(buffer,BUFF_SIZE); //Rotational tokenize(vs,buffer); if ((vs.size() == 4) && (vs[0].compare("Rotational") == 0)) { Scomponents.push_back(atof(vs[3].c_str())); } ifs.getline(buffer,BUFF_SIZE); //Vibrational tokenize(vs,buffer); if ((vs.size() == 4) && (vs[0].compare("Vibrational") == 0)) { Scomponents.push_back(atof(vs[3].c_str())); } } else if ((strstr(buffer,"Temperature=") != NULL) && (strstr(buffer,"Pressure=") != NULL)) { tokenize(vs,buffer); temperature = atof(vs[1].c_str()); } else if (strstr(buffer, "(0 K)") != NULL) { /* This must be the last else */ int i,nsearch; const char *search[] = { "CBS-QB3 (0 K)", "G2(0 K)", "G3(0 K)", "G4(0 K)", "W1BD (0 K)", "W1U (0 K)" }; const char *mymeth[] = { "CBS-QB3", "G2", "G3", "G4", "W1BD", "W1U" }; const int myindex[] = { 3, 2, 2, 2, 3, 3 }; nsearch = sizeof(search)/sizeof(search[0]); for(i=0; (i<nsearch); i++) { if(strstr(buffer,search[i]) != NULL) { tokenize(vs,buffer); E0 = atof(vs[myindex[i]].c_str()); E0_set = 1; thermo_method = mymeth[i]; break; } } } } // end while if (mol.NumAtoms() == 0) { // e.g., if we're at the end of a file PR#1737209 mol.EndModify(); return false; } mol.EndModify(); // Set conformers to all coordinates we adopted // but remove last geometry -- it's a duplicate if (vconf.size() > 1) vconf.pop_back(); mol.SetConformers(vconf); mol.SetConformer(mol.NumConformers() - 1); // Copy the conformer data too confData->SetDimension(confDimensions); confData->SetEnergies(confEnergies); confData->SetForces(confForces); mol.SetData(confData); // Check whether we have data to extract heat of formation. if (ezpe_set && Hcorr_set && Gcorr_set && E0_set && CV_set && (thermo_method.size() > 0)) { extract_thermo(&mol,thermo_method,temperature,ezpe, Hcorr,Gcorr,E0,CV,RotSymNum,Scomponents); } // Attach orbital data, if there is any if (orbitals.size() > 0) { OBOrbitalData *od = new OBOrbitalData; if (aHOMO == bHOMO) { od->LoadClosedShellOrbitals(orbitals, symmetries, aHOMO); } else { // we have to separate the alpha and beta vectors std::vector<double> betaOrbitals; std::vector<std::string> betaSymmetries; unsigned int initialSize = orbitals.size(); unsigned int symmSize = symmetries.size(); if (initialSize != symmSize || betaStart == -1) { cerr << "Inconsistency: orbitals have " << initialSize << " elements while symmetries have " << symmSize << endl; } else { for (unsigned int i = betaStart; i < initialSize; ++i) { betaOrbitals.push_back(orbitals[i]); if (symmetries.size() > 0) betaSymmetries.push_back(symmetries[i]); } // ok, now erase the end elements of orbitals and symmetries for (unsigned int i = betaStart; i < initialSize; ++i) { orbitals.pop_back(); if (symmetries.size() > 0) symmetries.pop_back(); } // and load the alphas and betas od->LoadAlphaOrbitals(orbitals, symmetries, aHOMO); od->LoadBetaOrbitals(betaOrbitals, betaSymmetries, bHOMO); } } od->SetOrigin(fileformatInput); mol.SetData(od); } //Attach vibrational data, if there is any, to molecule if(Frequencies.size()>0) { OBVibrationData* vd = new OBVibrationData; vd->SetData(Lx, Frequencies, Intensities); vd->SetOrigin(fileformatInput); mol.SetData(vd); } //Attach rotational data, if there is any, to molecule if(RotConsts[0]!=0.0) { OBRotationData* rd = new OBRotationData; rd->SetData(RotorType, RotConsts, RotSymNum); rd->SetOrigin(fileformatInput); mol.SetData(rd); } // Attach unit cell translation vectors if found if (numTranslationVectors > 0) { OBUnitCell* uc = new OBUnitCell; uc->SetData(translationVectors[0], translationVectors[1], translationVectors[2]); uc->SetOrigin(fileformatInput); mol.SetData(uc); } //Attach electronic transition data, if there is any, to molecule if(Forces.size() > 0 && Forces.size() == Wavelengths.size()) { OBElectronicTransitionData* etd = new OBElectronicTransitionData; etd->SetData(Wavelengths, Forces); if (EDipole.size() == Forces.size()) etd->SetEDipole(EDipole); if (RotatoryStrengthsLength.size() == Forces.size()) etd->SetRotatoryStrengthsLength(RotatoryStrengthsLength); if (RotatoryStrengthsVelocity.size() == Forces.size()) etd->SetRotatoryStrengthsVelocity(RotatoryStrengthsVelocity); etd->SetOrigin(fileformatInput); mol.SetData(etd); } if (!pConv->IsOption("b",OBConversion::INOPTIONS)) mol.ConnectTheDots(); if (!pConv->IsOption("s",OBConversion::INOPTIONS) && !pConv->IsOption("b",OBConversion::INOPTIONS)) mol.PerceiveBondOrders(); if (hasPartialCharges) { mol.SetPartialChargesPerceived(); // Annotate that partial charges come from Mulliken OBPairData *dp = new OBPairData; dp->SetAttribute("PartialCharges"); dp->SetValue(chargeModel); // Mulliken, ESP, etc. dp->SetOrigin(fileformatInput); mol.SetData(dp); } mol.SetTotalCharge(total_charge); mol.SetTotalSpinMultiplicity(spin_multiplicity); mol.SetTitle(title); return(true); }
// Reading Gaussian output has been tested for G98 and G03 to some degree // If you have problems (or examples of older output), please contact // the [email protected] mailing list and/or post a bug bool GaussianOutputFormat::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; const char* title = pConv->GetTitle(); char buffer[BUFF_SIZE]; string str,str1; double x,y,z; OBAtom *atom; vector<string> vs; int charge = 0; unsigned int spin = 1; bool hasPartialCharges = false; string chargeModel; // descriptor for charges (e.g. "Mulliken") // coordinates of all steps // Set conformers to all coordinates we adopted std::vector<double*> vconf; // index of all frames/conformers std::vector<double> coordinates; // coordinates in each frame int natoms = 0; // number of atoms -- ensure we don't go to a new job with a different molecule // OBConformerData stores information about multiple steps // we can change attribute later if needed (e.g., IRC) OBConformerData *confData = new OBConformerData(); confData->SetOrigin(fileformatInput); std::vector<unsigned short> confDimensions = confData->GetDimension(); // to be fair, set these all to 3D std::vector<double> confEnergies = confData->GetEnergies(); std::vector< std::vector< vector3 > > confForces = confData->GetForces(); //Vibrational data std::vector< std::vector< vector3 > > Lx; std::vector<double> Frequencies, Intensities; //Rotational data std::vector<double> RotConsts(3); int RotSymNum=1; OBRotationData::RType RotorType; // Translation vectors (if present) vector3 translationVectors[3]; int numTranslationVectors = 0; //Electronic Excitation data std::vector<double> Forces, Wavelengths, EDipole, RotatoryStrengthsVelocity, RotatoryStrengthsLength; // Orbital data std::vector<double> orbitals; std::vector<std::string> symmetries; int aHOMO, bHOMO, betaStart; //Put some metadata into OBCommentData string comment("Gaussian "); ifs.getline(buffer,BUFF_SIZE); if(*buffer) { comment += strchr(buffer,'=')+2; comment += ""; for(unsigned i=0; i<115, ifs; ++i) { ifs.getline(buffer,BUFF_SIZE); if(buffer[1]=='#') { //the line describing the method comment += buffer; OBCommentData *cd = new OBCommentData; cd->SetData(comment); cd->SetOrigin(fileformatInput); mol.SetData(cd); break; } } } int i=0; bool no_symmetry=false; char coords_type[25]; //Prescan file to find second instance of "orientation:" //This will be the kind of coords used in the chk/fchk file //Unless the "nosym" keyword has been requested while (ifs.getline(buffer,BUFF_SIZE)) { if (strstr(buffer,"Symmetry turned off by external request.") != NULL) { // The "nosym" keyword has been requested no_symmetry = true; } if (strstr(buffer,"orientation:") !=NULL) { i++; tokenize (vs, buffer); strcpy (coords_type, vs[0].c_str()); strcat (coords_type, " orientation:"); } if ((no_symmetry && i==1) || i==2) break; // Check for the last line of normal output and exit loop, otherwise, // the rewind below will no longer work. if (strstr(buffer,"Normal termination of Gaussian") != NULL) break; } ifs.seekg(0); //rewind mol.BeginModify(); while (ifs.getline(buffer,BUFF_SIZE)) { if (strstr(buffer,"Multiplicity") != NULL) { tokenize(vs, buffer, " \t\n"); if (vs.size() == 6) { charge = atoi(vs[2].c_str()); spin = atoi(vs[5].c_str()); } ifs.getline(buffer,BUFF_SIZE); } else if (strstr(buffer, coords_type) != NULL) { numTranslationVectors = 0; // ignore old translationVectors ifs.getline(buffer,BUFF_SIZE); // --------------- ifs.getline(buffer,BUFF_SIZE); // column headings ifs.getline(buffer,BUFF_SIZE); // column headings ifs.getline(buffer,BUFF_SIZE); // --------------- ifs.getline(buffer,BUFF_SIZE); tokenize(vs,buffer); while (vs.size() == 6) { x = atof((char*)vs[3].c_str()); y = atof((char*)vs[4].c_str()); z = atof((char*)vs[5].c_str()); int atomicNum = atoi((char*)vs[1].c_str()); if (atomicNum > 0) // translation vectors are "-2" { if (natoms == 0) { // first time reading the molecule, create each atom atom = mol.NewAtom(); atom->SetAtomicNum(atoi((char*)vs[1].c_str())); } coordinates.push_back(x); coordinates.push_back(y); coordinates.push_back(z); } else { translationVectors[numTranslationVectors++].Set(x, y, z); } if (!ifs.getline(buffer,BUFF_SIZE)) { break; } tokenize(vs,buffer); } // done with reading atoms natoms = mol.NumAtoms(); // malloc / memcpy double *tmpCoords = new double [(natoms)*3]; memcpy(tmpCoords, &coordinates[0], sizeof(double)*natoms*3); vconf.push_back(tmpCoords); coordinates.clear(); confDimensions.push_back(3); // always 3D -- OBConformerData allows mixing 2D and 3D structures } else if(strstr(buffer,"Dipole moment") != NULL) { ifs.getline(buffer,BUFF_SIZE); // actual components X ### Y #### Z ### tokenize(vs,buffer); if (vs.size() >= 6) { OBVectorData *dipoleMoment = new OBVectorData; dipoleMoment->SetAttribute("Dipole Moment"); double x, y, z; x = atof(vs[1].c_str()); y = atof(vs[3].c_str()); z = atof(vs[5].c_str()); dipoleMoment->SetData(x, y, z); dipoleMoment->SetOrigin(fileformatInput); mol.SetData(dipoleMoment); } if (!ifs.getline(buffer,BUFF_SIZE)) break; } else if(strstr(buffer,"Total atomic charges") != NULL || strstr(buffer,"Mulliken atomic charges") != NULL) { hasPartialCharges = true; chargeModel = "Mulliken"; ifs.getline(buffer,BUFF_SIZE); // column headings ifs.getline(buffer,BUFF_SIZE); tokenize(vs,buffer); while (vs.size() >= 3 && strstr(buffer,"Sum of ") == NULL) { atom = mol.GetAtom(atoi(vs[0].c_str())); if (!atom) break; atom->SetPartialCharge(atof(vs[2].c_str())); if (!ifs.getline(buffer,BUFF_SIZE)) break; tokenize(vs,buffer); } } else if (strstr(buffer, "Charges from ESP fit") != NULL) { hasPartialCharges = true; chargeModel = "ESP"; ifs.getline(buffer,BUFF_SIZE); // Charge / dipole line ifs.getline(buffer,BUFF_SIZE); // column header ifs.getline(buffer,BUFF_SIZE); // real charges tokenize(vs,buffer); while (vs.size() >= 3 && strstr(buffer,"-----") == NULL) { atom = mol.GetAtom(atoi(vs[0].c_str())); if (!atom) break; atom->SetPartialCharge(atof(vs[2].c_str())); if (!ifs.getline(buffer,BUFF_SIZE)) break; tokenize(vs,buffer); } } else if(strstr(buffer,"Natural Population") != NULL) { hasPartialCharges = true; chargeModel = "NBO"; ifs.getline(buffer,BUFF_SIZE); // column headings ifs.getline(buffer,BUFF_SIZE); // again ifs.getline(buffer,BUFF_SIZE); // again (-----) ifs.getline(buffer,BUFF_SIZE); // real data tokenize(vs,buffer); while (vs.size() >= 3 && strstr(buffer,"=====") == NULL) { atom = mol.GetAtom(atoi(vs[1].c_str())); if (!atom) break; atom->SetPartialCharge(atof(vs[2].c_str())); if (!ifs.getline(buffer,BUFF_SIZE)) break; tokenize(vs,buffer); } } else if(strstr(buffer, " Frequencies -- ")) //vibrational frequencies { //The info should appear only once as several blocks starting with this line tokenize(vs, buffer); for(unsigned int i=2; i<vs.size(); ++i) Frequencies.push_back(atof(vs[i].c_str())); ifs.getline(buffer,BUFF_SIZE); //Red. masses ifs.getline(buffer,BUFF_SIZE); //Frc consts ifs.getline(buffer,BUFF_SIZE); //IR Inten tokenize(vs, buffer); for(unsigned int i=3; i<vs.size(); ++i) Intensities.push_back(atof(vs[i].c_str())); ifs.getline(buffer, BUFF_SIZE); // column labels or Raman intensity if(strstr(buffer, "Raman Activ")) { ifs.getline(buffer, BUFF_SIZE); // Depolar (P) ifs.getline(buffer, BUFF_SIZE); // Depolar (U) ifs.getline(buffer, BUFF_SIZE); // column labels } ifs.getline(buffer, BUFF_SIZE); // actual displacement data tokenize(vs, buffer); vector<vector3> vib1, vib2, vib3; double x, y, z; while(vs.size() > 5) { for (unsigned int i = 2; i < vs.size()-2; i += 3) { x = atof(vs[i].c_str()); y = atof(vs[i+1].c_str()); z = atof(vs[i+2].c_str()); if (i == 2) vib1.push_back(vector3(x, y, z)); else if (i == 5) vib2.push_back(vector3(x, y, z)); else if (i == 8) vib3.push_back(vector3(x, y, z)); } if (!ifs.getline(buffer, BUFF_SIZE)) break; tokenize(vs,buffer); } Lx.push_back(vib1); if (vib2.size()) Lx.push_back(vib2); if (vib3.size()) Lx.push_back(vib3); } else if(strstr(buffer, " This molecule is "))//rotational data { if(strstr(buffer, "asymmetric")) RotorType = OBRotationData::ASYMMETRIC; else if(strstr(buffer, "symmetric")) RotorType = OBRotationData::SYMMETRIC; else if(strstr(buffer, "linear")) RotorType = OBRotationData::LINEAR; else RotorType = OBRotationData::UNKNOWN; ifs.getline(buffer,BUFF_SIZE); //symmetry number tokenize(vs, buffer); RotSymNum = atoi(vs[3].c_str()); } else if(strstr(buffer, "Rotational constant")) { tokenize(vs, buffer); RotConsts.clear(); for (unsigned int i=3; i<vs.size(); ++i) RotConsts.push_back(atof(vs[i].c_str())); } else if(strstr(buffer, "alpha electrons")) // # of electrons / orbital { tokenize(vs, buffer); if (vs.size() == 6) { // # alpha electrons # beta electrons aHOMO = atoi(vs[0].c_str()); bHOMO = atoi(vs[3].c_str()); } } else if(strstr(buffer, "rbital symmetries")) // orbital symmetries { symmetries.clear(); std::string label; // used as a temporary to remove "(" and ")" from labels int offset = 0; while(true) { ifs.getline(buffer, BUFF_SIZE); tokenize(vs, buffer); // parse first line "Occupied" ... for (unsigned int i = 1; i < vs.size(); ++i) { label = vs[i].substr(1, vs[i].length() - 2); symmetries.push_back(label); } ifs.getline(buffer, BUFF_SIZE); // Parse remaining lines while (strstr(buffer, "(")) { tokenize(vs, buffer); if (strstr(buffer, "Virtual")) { offset = 1; // skip first token } else { offset = 0; } for (unsigned int i = offset; i < vs.size(); ++i) { label = vs[i].substr(1, vs[i].length() - 2); symmetries.push_back(label); } ifs.getline(buffer, BUFF_SIZE); // get next line } // end parsing symmetry labels if (!strstr(buffer, "Beta")) // no beta orbitals break; } // end alpha/beta section } else if (strstr(buffer, "Alpha") && strstr(buffer, ". eigenvalues --")) { orbitals.clear(); betaStart = 0; while (strstr(buffer, ". eigenvalues --")) { tokenize(vs, buffer); if (vs.size() < 4) break; if (vs[0].find("Beta") !=string::npos && betaStart == 0) // mark where we switch from alpha to beta betaStart = orbitals.size(); for (unsigned int i = 4; i < vs.size(); ++i) { orbitals.push_back(atof(vs[i].c_str())); } ifs.getline(buffer, BUFF_SIZE); } } else if(strstr(buffer, " Excited State")) // Force and wavelength data { // The above line appears for each state, so just append the info to the vectors tokenize(vs, buffer); if (vs.size() == 9) { double wavelength = atof(vs[6].c_str()); double force = atof(vs[8].substr(2).c_str()); Forces.push_back(force); Wavelengths.push_back(wavelength); } } else if(strstr(buffer, " Ground to excited state Transition electric dipole moments (Au):")) // Electronic dipole moments { ifs.getline(buffer, BUFF_SIZE); // Headings ifs.getline(buffer, BUFF_SIZE); // First entry tokenize(vs, buffer); while (vs.size() == 5) { double s = atof(vs[4].c_str()); EDipole.push_back(s); ifs.getline(buffer, BUFF_SIZE); tokenize(vs, buffer); } } else if(strstr(buffer, " state X Y Z R(velocity)")) { // Rotatory Strengths ifs.getline(buffer, BUFF_SIZE); // First entry 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) { #define HARTREE_TO_KCAL 627.509469 tokenize(vs,buffer); mol.SetEnergy(atof(vs[4].c_str()) * HARTREE_TO_KCAL); 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()); } */ } // end while if (mol.NumAtoms() == 0) { // e.g., if we're at the end of a file PR#1737209 mol.EndModify(); return false; } mol.EndModify(); // Set conformers to all coordinates we adopted // but remove last geometry -- it's a duplicate if (vconf.size() > 1) vconf.pop_back(); mol.SetConformers(vconf); mol.SetConformer(mol.NumConformers() - 1); // Copy the conformer data too confData->SetDimension(confDimensions); confData->SetEnergies(confEnergies); confData->SetForces(confForces); mol.SetData(confData); // Attach orbital data, if there is any if (orbitals.size() > 0) { OBOrbitalData *od = new OBOrbitalData; if (aHOMO == bHOMO) { od->LoadClosedShellOrbitals(orbitals, symmetries, aHOMO); } else { // we have to separate the alpha and beta vectors std::vector<double> betaOrbitals; std::vector<std::string> betaSymmetries; unsigned int initialSize = orbitals.size(); for (unsigned int i = betaStart; i < initialSize; ++i) { betaOrbitals.push_back(orbitals[i]); if (symmetries.size() > 0) betaSymmetries.push_back(symmetries[i]); } // ok, now erase the end elements of orbitals and symmetries for (unsigned int i = betaStart; i < initialSize; ++i) { orbitals.pop_back(); if (symmetries.size() > 0) symmetries.pop_back(); } // and load the alphas and betas od->LoadAlphaOrbitals(orbitals, symmetries, aHOMO); od->LoadBetaOrbitals(betaOrbitals, betaSymmetries, bHOMO); } od->SetOrigin(fileformatInput); mol.SetData(od); } //Attach vibrational data, if there is any, to molecule if(Frequencies.size()>0) { OBVibrationData* vd = new OBVibrationData; vd->SetData(Lx, Frequencies, Intensities); vd->SetOrigin(fileformatInput); mol.SetData(vd); } //Attach rotational data, if there is any, to molecule if(RotConsts[0]!=0.0) { OBRotationData* rd = new OBRotationData; rd->SetData(RotorType, RotConsts, RotSymNum); rd->SetOrigin(fileformatInput); mol.SetData(rd); } // Attach unit cell translation vectors if found if (numTranslationVectors > 0) { OBUnitCell* uc = new OBUnitCell; uc->SetData(translationVectors[0], translationVectors[1], translationVectors[2]); uc->SetOrigin(fileformatInput); mol.SetData(uc); } //Attach electronic transition data, if there is any, to molecule if(Forces.size() > 0 && Forces.size() == Wavelengths.size()) { OBElectronicTransitionData* etd = new OBElectronicTransitionData; etd->SetData(Wavelengths, Forces); if (EDipole.size() == Forces.size()) etd->SetEDipole(EDipole); if (RotatoryStrengthsLength.size() == Forces.size()) etd->SetRotatoryStrengthsLength(RotatoryStrengthsLength); if (RotatoryStrengthsVelocity.size() == Forces.size()) etd->SetRotatoryStrengthsVelocity(RotatoryStrengthsVelocity); etd->SetOrigin(fileformatInput); mol.SetData(etd); } if (!pConv->IsOption("b",OBConversion::INOPTIONS)) mol.ConnectTheDots(); if (!pConv->IsOption("s",OBConversion::INOPTIONS) && !pConv->IsOption("b",OBConversion::INOPTIONS)) mol.PerceiveBondOrders(); if (hasPartialCharges) { mol.SetPartialChargesPerceived(); // Annotate that partial charges come from Mulliken OBPairData *dp = new OBPairData; dp->SetAttribute("PartialCharges"); dp->SetValue(chargeModel); // Mulliken, ESP, etc. dp->SetOrigin(fileformatInput); mol.SetData(dp); } mol.SetTotalCharge(charge); mol.SetTotalSpinMultiplicity(spin); mol.SetTitle(title); return(true); }
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;
char temp_type[BUFF_SIZE], resname[BUFF_SIZE], atmid[BUFF_SIZE];
int elemno, resnum = -1;
ttab.SetFromType("SYB");
for (i = 0;i < natoms;i++)
{
if (!ifs.getline(buffer,BUFF_SIZE))
return(false);
sscanf(buffer," %*s %1024s %lf %lf %lf %1024s %d %1024s %lf",
atmid, &x,&y,&z, temp_type, &resnum, resname, &pcharge);
atom.SetVector(x, y, z);
// Handle "CL" and "BR" and other mis-typed atoms
str = temp_type;
if (strncmp(temp_type, "CL", 2) == 0) {
str = "Cl";
} else if (strncmp(temp_type,"BR",2) == 0) {
str = "Br";
} else if (strncmp(temp_type,"S.o2", 4) == 02) {
str = "S.O2";
} else if (strncmp(temp_type,"S.o", 3) == 0) {
str = "S.O";
} else if (strncmp(temp_type,"SI", 2) == 0) {
str = "Si";
// The following cases are entries which are not in openbabel/data/types.txt
// and should probably be added there
} else if (strncmp(temp_type,"S.1", 3) == 0) {
str = "S.2"; // no idea what the best type might be here
} else if (strncmp(temp_type,"P.", 2) == 0) {
str = "P.3";
} else if (strncasecmp(temp_type,"Ti.", 3) == 0) { // e.g. Ti.th
str = "Ti";
} else if (strncasecmp(temp_type,"Ru.", 3) == 0) { // e.g. Ru.oh
str = "Ru";
}
ttab.SetToType("ATN");
ttab.Translate(str1,str);
elemno = atoi(str1.c_str());
ttab.SetToType("IDX");
// We might have missed some SI or FE type things above, so here's
// another check
if( !elemno && isupper(temp_type[1]) )
{
temp_type[1] = (char)tolower(temp_type[1]);
str = temp_type;
ttab.Translate(str1,str);
elemno = atoi(str1.c_str());
}
// One last check if there isn't a period in the type,
// it's a malformed atom type, but it may be the element symbol
// GaussView does this (PR#1739905)
if ( !elemno ) {
obErrorLog.ThrowError(__FUNCTION__, "This Mol2 file is non-standard. Cannot interpret atom types correctly, instead attempting to interpret as elements instead.", obWarning);
string::size_type dotPos = str.find('.');
if (dotPos == string::npos) {
elemno = etab.GetAtomicNum(str.c_str());
}
}
atom.SetAtomicNum(elemno);
ttab.SetToType("INT");
ttab.Translate(str1,str);
atom.SetType(str1);
atom.SetPartialCharge(pcharge);
if (!mol.AddAtom(atom))
return(false);
if (!IsNearZero(pcharge))
hasPartialCharges = true;
// Add residue information if it exists
if (resnum != -1 && resnum != 0 &&
strlen(resname) != 0 && strncmp(resname,"<1>", 3) != 0)
{
OBResidue *res = (mol.NumResidues() > 0) ?
mol.GetResidue(mol.NumResidues()-1) : NULL;
if (res == NULL || res->GetName() != resname ||
static_cast<int>(res->GetNum()) != resnum)
{
vector<OBResidue*>::iterator ri;
for (res = mol.BeginResidue(ri) ; res ; res = mol.NextResidue(ri))
if (res->GetName() == resname &&
static_cast<int>(res->GetNum()) == resnum)
break;
if (res == NULL)
{
res = mol.NewResidue();
res->SetName(resname);
res->SetNum(resnum);
}
}
OBAtom *atomPtr = mol.GetAtom(mol.NumAtoms());
res->AddAtom(atomPtr);
res->SetAtomID(atomPtr, atmid);
} // end adding residue info
}
for (;;)
{
if (!ifs.getline(buffer,BUFF_SIZE))
return(false);
str = buffer;
if (!strncmp(buffer,"@<TRIPOS>BOND",13))
break;
}
int start,end,order;
for (i = 0; i < nbonds; i++)
{
if (!ifs.getline(buffer,BUFF_SIZE))
return(false);
sscanf(buffer,"%*d %d %d %1024s",&start,&end,temp_type);
str = temp_type;
order = 1;
if (str == "ar" || str == "AR" || str == "Ar")
order = 5;
else if (str == "AM" || str == "am" || str == "Am")
order = 1;
else
order = atoi(str.c_str());
mol.AddBond(start,end,order);
}
// Suggestion by Liu Zhiguo 2008-01-26
// Mol2 files define atom types -- there is no need to re-perceive
mol.SetAtomTypesPerceived();
mol.EndModify();
//must add generic data after end modify - otherwise it will be blown away
if (comment)
{
OBCommentData *cd = new OBCommentData;
cd->SetData(comment);
cd->SetOrigin(fileformatInput);
mol.SetData(cd);
delete [] comment;
comment = NULL;
}
if (hasPartialCharges)
mol.SetPartialChargesPerceived();
/* Disabled due to PR#3048758 -- seekg is very slow with gzipped mol2
// continue untill EOF or untill next molecule record
streampos pos;
for(;;)
{
pos = ifs.tellg();
if (!ifs.getline(buffer,BUFF_SIZE))
break;
if (EQn(buffer,"@<TRIPOS>MOLECULE",17))
break;
}
ifs.seekg(pos); // go back to the end of the molecule
*/
return(true);
}
bool MacroModFormat::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;
const char* defaultTitle = pConv->GetTitle();
// Get Title
char buffer[BUFF_SIZE];
int natoms;
vector<vector<pair<int,int> > > connections;
if (ifs.getline(buffer,BUFF_SIZE))
{
vector<string> vs;
tokenize(vs,buffer," \n");
if ( !vs.empty() && vs.size() > 0)
sscanf(buffer,"%i%*s",&natoms);
if (natoms == 0)
return false;
if ( !vs.empty() && vs.size() > 1)
mol.SetTitle(vs[1]);
else
{
string s = defaultTitle;
mol.SetTitle(defaultTitle);
}
}
else
return(false);
mol.BeginModify();
mol.ReserveAtoms(natoms);
connections.resize(natoms+1);
/***********************************************************************/
// Get Type Bonds, BondOrder, X, Y, Z
double x,y,z;
vector3 v;
char temp_type[10];
int i,j;
double charge;
OBAtom atom;
ttab.SetFromType("MMD");
for (i = 1; i <= natoms; i++)
{
if (!ifs.getline(buffer,BUFF_SIZE))
break;
int end[6], order[6];
sscanf(buffer,"%9s%d%d%d%d%d%d%d%d%d%d%d%d%lf%lf%lf",
temp_type,&end[0],&order[0],&end[1],&order[1],&end[2],&order[2],
&end[3], &order[3], &end[4], &order[4], &end[5], &order[5],
&x, &y, &z);
pair<int,int> tmp;
for ( j = 0 ; j <=5 ; j++ )
{
if ( end[j] > 0 && end[j] > i)
{
tmp.first = end[j];
tmp.second = order[j];
connections[i].push_back(tmp);
}
}
v.SetX(x);
v.SetY(y);
v.SetZ(z);
atom.SetVector(v);
string str = temp_type,str1;
ttab.SetToType("ATN");
ttab.Translate(str1,str);
atom.SetAtomicNum(atoi(str1.c_str()));
ttab.SetToType("INT");
ttab.Translate(str1,str);
atom.SetType(str1);
// stuff for optional fields
buffer[109]='\0';
sscanf(&buffer[101],"%lf", &charge);
atom.SetPartialCharge(charge);
mol.AddAtom(atom);
}
for (i = 1; i <= natoms; i++)
for (j = 0; j < (signed)connections[i].size(); j++)
mol.AddBond(i, connections[i][j].first, connections[i][j].second);
mol.EndModify();
mol.SetPartialChargesPerceived();
// Annotate origin of partial charges
OBPairData *dp = new OBPairData;
dp->SetAttribute("PartialCharges");
dp->SetValue("MACROMODEL");
dp->SetOrigin(fileformatInput);
mol.SetData(dp);
OBBond *bond;
vector<OBBond*>::iterator bi;
for (bond = mol.BeginBond(bi);bond;bond = mol.NextBond(bi))
if (bond->GetBondOrder() == 5 && !bond->IsInRing())
bond->SetBondOrder(1);
if ( natoms != (signed)mol.NumAtoms() )
return(false);
// clean out remaining blank lines
std::streampos ipos;
do
{
ipos = ifs.tellg();
ifs.getline(buffer,BUFF_SIZE);
}
while(strlen(buffer) == 0 && !ifs.eof() );
ifs.seekg(ipos);
return(true);
}