source: src/Parser/MpqcParser.cpp@ a9b86d

/*
 * Project: MoleCuilder
 * Description: creates and alters molecular systems
 * Copyright (C)  2010 University of Bonn. All rights reserved.
 * Please see the LICENSE file or "Copyright notice" in builder.cpp for details.
 */

/*
 * MpqcParser.cpp
 *
 *  Created on: 12.06.2010
 *      Author: heber
 */

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

#include <iostream>
#include <boost/tokenizer.hpp>
#include <string>

#include "CodePatterns/MemDebug.hpp"

#include "MpqcParser.hpp"
#include "MpqcParser_Parameters.hpp"

#include "atom.hpp"
#include "config.hpp"
#include "Element/element.hpp"
#include "molecule.hpp"
#include "CodePatterns/Log.hpp"
#include "CodePatterns/toString.hpp"
#include "CodePatterns/Verbose.hpp"
#include "LinearAlgebra/Vector.hpp"
#include "Element/periodentafel.hpp"
#include "World.hpp"


/** Constructor of MpqcParser.
 *
 */
MpqcParser::MpqcParser()
{
  parameters = new MpqcParser_Parameters();
}

/** Destructor of MpqcParser.
 *
 */
MpqcParser::~MpqcParser()
{
  delete parameters;
}

/** Load an MPQC config file into the World.
 * \param *file input stream
 */
void MpqcParser::load(istream *file)
{
  bool MpqcSection = false;
  bool MoleculeSection = false;
  bool GeometrySection = false;
  bool BasisSection = false;
  bool AuxBasisSection = false;
  char line[MAXSTRINGSIZE];
  typedef boost::tokenizer<boost::char_separator<char> > tokenizer;
  boost::char_separator<char> sep("[]");
  boost::char_separator<char> angularsep("<>");
  boost::char_separator<char> equalitysep(" =");
  boost::char_separator<char> whitesep(" \t");
  ConvertTo<double> toDouble;

  molecule *newmol = World::getInstance().createMolecule();
  newmol->ActiveFlag = true;
  // TODO: Remove the insertion into molecule when saving does not depend on them anymore. Also, remove molecule.hpp include
  World::getInstance().getMolecules()->insert(newmol);
  while (file->good()) {
    file->getline(line, MAXSTRINGSIZE-1);
    std::string linestring(line);
    if ((linestring.find("atoms geometry") == string::npos) && (linestring.find("}") != string::npos)) {
      GeometrySection = false;
    }
    if ((linestring.find(")") != string::npos)) { // ends a section which do not overlap
      MpqcSection = false;
      MoleculeSection = false;
      BasisSection = false;
      AuxBasisSection = false;
    }
    if (MoleculeSection) {
      if (GeometrySection) { // we have an atom
        tokenizer tokens(linestring, sep);
  //      if (tokens.size() != 2)
  //        throw MpqcParseException;
        tokenizer::iterator tok_iter = tokens.begin();
        ASSERT(tok_iter != tokens.end(),
            "MpqcParser::load() - missing token for MoleculeSection in line "+linestring+"!");
        std::stringstream whitespacefilter(*tok_iter++);
        std::string element;
        whitespacefilter >> ws >> element;
        ASSERT(tok_iter != tokens.end(),
            "MpqcParser::load() - missing token for MoleculeSection in line "+linestring+"!");
        std::string vector = *tok_iter;
        tokenizer vectorcomponents(vector, whitesep);
        Vector X;
  //      if (vectorcomponents.size() != NDIM)
  //        throw MpqcParseException;
        tok_iter = vectorcomponents.begin();
        for (int i=0; i<NDIM; ++i) {
          X[i] = toDouble(*tok_iter++);
        }
        // create atom
        atom *newAtom = World::getInstance().createAtom();
        newAtom->setType(World::getInstance().getPeriode()->FindElement(element));
        newAtom->setPosition(X);
        newmol->AddAtom(newAtom);
        DoLog(1) && (Log() << Verbose(1) << "Adding atom " << *newAtom << std::endl);
      }
    }
    if (MpqcSection) {
      if (linestring.find("mole<") != string::npos) { // get theory
        tokenizer tokens(linestring, angularsep);
        tokenizer::iterator tok_iter = tokens.begin();
        ++tok_iter;
        ASSERT(tok_iter != tokens.end(),
            "MpqcParser::load() - missing token in brackets<> for mole< in line "+linestring+"!");
        std::string value(*tok_iter);
        std::stringstream linestream("theory = "+value);
        linestream >> getParams();
      } else if (linestring.find("integrals<") != string::npos) { // get theory
        tokenizer tokens(linestring, angularsep);
        tokenizer::iterator tok_iter = tokens.begin();
        ++tok_iter;
        ASSERT(tok_iter != tokens.end(),
            "MpqcParser::load() - missing token in brackets<> for integrals< in line "+linestring+"!");
        std::string value(*tok_iter);
        std::stringstream linestream("integration = "+value);
        linestream >> getParams();
      } else if ((linestring.find("molecule") == string::npos) && (linestring.find("basis") == string::npos)){
        // molecule and basis must not be parsed in this section
        tokenizer tokens(linestring, equalitysep);
        tokenizer::iterator tok_iter = tokens.begin();
        ASSERT(tok_iter != tokens.end(),
            "MpqcParser::load() - missing token before '=' for MpqcSection in line "+linestring+"!");
        std::stringstream whitespacefilter(*tok_iter);
        std::string key;
        whitespacefilter >> ws >> key;
        if (getParams().haveParam(key)) {
          std::stringstream linestream(linestring);
          linestream >> getParams();
        } else { // unknown keys are simply ignored as long as parser is incomplete
          DoLog(2) && (Log() << Verbose(2) << "INFO: '"+key+"' is unknown and ignored." << std::endl);
        }
      }
    }
    if (BasisSection) {
      tokenizer tokens(linestring, equalitysep);
      tokenizer::iterator tok_iter = tokens.begin();
      ASSERT(tok_iter != tokens.end(),
          "MpqcParser::load() - missing token for BasisSection in line "+linestring+"!");
      std::string key(*tok_iter++);
      ASSERT(tok_iter != tokens.end(),
          "MpqcParser::load() - missing value for BasisSection after key "+key+" in line "+linestring+"!");
      std::string value(*tok_iter);
      tok_iter++;
      // TODO: use exception instead of ASSERT
      ASSERT(tok_iter == tokens.end(),
          "MpqcParser::load() - more than (key = value) on line "+linestring+".");
      if (key == "name") {
        std::stringstream linestream("basis = "+value);
        linestream >> getParams();
      }
    }
    if (AuxBasisSection) {
      tokenizer tokens(linestring, equalitysep);
      tokenizer::iterator tok_iter = tokens.begin();
      ASSERT(tok_iter != tokens.end(),
          "MpqcParser::load() - missing token for AuxBasisSection in line "+linestring+"!");
      std::string key(*tok_iter++);
      ASSERT(tok_iter != tokens.end(),
          "MpqcParser::load() - missing value for BasisSection after key "+key+" in line "+linestring+"!");
      std::string value(*tok_iter);
      tok_iter++;
      // TODO: use exception instead of ASSERT
      ASSERT(tok_iter == tokens.end(),
          "MpqcParser::load() - more than (key = value) on line "+linestring+".");
      if (key == "name") {
        std::stringstream linestream("aux_basis = "+value);
        linestream >> getParams();
      }
    }
    // set some scan flags
    if (linestring.find("mpqc:") != string::npos) {
      MpqcSection = true;
    }
    if (linestring.find("molecule<Molecule>:") != string::npos) {
      MoleculeSection = true;
    }
    if (linestring.find("atoms geometry") != string::npos) {
      GeometrySection = true;
    }
    if ((linestring.find("basis<GaussianBasisSet>:") != string::npos) && ((linestring.find("abasis<") == string::npos))) {
      BasisSection = true;
    }
    if (linestring.find("abasis<") != string::npos) {
      AuxBasisSection = true;
    }
  }
  // refresh atom::nr and atom::name
  newmol->getAtomCount();
}

/** Saves all atoms and data into a MPQC config file.
 * \param *file output stream
 * \param atoms atoms to store
 */
void MpqcParser::save(ostream *file, const std::vector<atom *> &atoms)
{
  Vector center;
//  vector<atom *> allatoms = World::getInstance().getAllAtoms();

  // calculate center
  for (std::vector<atom *>::const_iterator runner = atoms.begin();runner != atoms.end(); ++runner)
    center += (*runner)->getPosition();
  center.Scale(1./(double)atoms.size());

  // first without hessian
  if (file->fail()) {
    DoeLog(1) && (eLog()<< Verbose(1) << "Cannot open mpqc output file." << endl);
  } else {
    *file << "% Created by MoleCuilder" << endl;
    *file << "mpqc: (" << endl;
    *file << "\tsavestate = " << getParams().getString(MpqcParser_Parameters::savestateParam) << endl;
    *file << "\tdo_gradient = " << getParams().getString(MpqcParser_Parameters::do_gradientParam) << endl;
    if (getParams().getBool(MpqcParser_Parameters::hessianParam)) {
      *file << "\tfreq<MolecularFrequencies>: (" << endl;
      *file << "\t\tmolecule=$:molecule" << endl;
      *file << "\t)" << endl;
    }
    switch (getParams().getTheory()) {
      case MpqcParser_Parameters::CLHF:
        *file << "\tmole<" << getParams().getString(MpqcParser_Parameters::theoryParam) << ">: (" << endl;
        *file << "\t\tmolecule = $:molecule" << endl;
        *file << "\t\tbasis = $:basis" << endl;
        *file << "\t\tmaxiter = " << toString(getParams().getInt(MpqcParser_Parameters::maxiterParam))<< endl;
        *file << "\t\tmemory = " << toString(getParams().getInt(MpqcParser_Parameters::memoryParam)) << endl;
        *file << "\t)" << endl;
        break;
      case MpqcParser_Parameters::CLKS:
        *file << "\tmole<" << getParams().getString(MpqcParser_Parameters::theoryParam) << ">: (" << endl;
        *file << "\t\tfunctional<StdDenFunctional>:(name=B3LYP)" << endl;
        *file << "\t\tmolecule = $:molecule" << endl;
        *file << "\t\tbasis = $:basis" << endl;
        *file << "\t\tmaxiter = " << toString(getParams().getInt(MpqcParser_Parameters::maxiterParam))<< endl;
        *file << "\t\tmemory = " << toString(getParams().getInt(MpqcParser_Parameters::memoryParam)) << endl;
        *file << "\t)" << endl;
        break;
      case MpqcParser_Parameters::MBPT2:
        *file << "\tmole<" << getParams().getString(MpqcParser_Parameters::theoryParam) << ">: (" << endl;
        *file << "\t\tbasis = $:basis" << endl;
        *file << "\t\tmolecule = $:molecule" << endl;
        *file << "\t\tmemory = " << toString(getParams().getInt(MpqcParser_Parameters::memoryParam)) << endl;
        *file << "\t\treference<CLHF>: (" << endl;
        *file << "\t\t\tmaxiter = " << toString(getParams().getInt(MpqcParser_Parameters::maxiterParam))<< endl;
        *file << "\t\t\tbasis = $:basis" << endl;
        *file << "\t\t\tmolecule = $:molecule" << endl;
        *file << "\t\t\tmemory = " << toString(getParams().getInt(MpqcParser_Parameters::memoryParam)) << endl;
        *file << "\t\t)" << endl;
        *file << "\t)" << endl;
        break;
      case MpqcParser_Parameters::MBPT2_R12:
        *file << "\tmole<" << getParams().getString(MpqcParser_Parameters::theoryParam) << ">: (" << endl;
        *file << "\t\tmolecule = $:molecule" << endl;
        *file << "\t\tbasis = $:basis" << endl;
        *file << "\t\taux_basis = $:abasis" << endl;
        *file << "\t\tstdapprox = \"" << getParams().getString(MpqcParser_Parameters::stdapproxParam) << "\"" << endl;
        *file << "\t\tnfzc = " << toString(getParams().getInt(MpqcParser_Parameters::nfzcParam)) << endl;
        *file << "\t\tmemory = " << toString(getParams().getInt(MpqcParser_Parameters::memoryParam)) << endl;
        *file << "\t\tintegrals<IntegralCints>:()" << endl;
        *file << "\t\treference<CLHF>: (" << endl;
        *file << "\t\t\tmolecule = $:molecule" << endl;
        *file << "\t\t\tbasis = $:basis" << endl;
        *file << "\t\t\tmaxiter = " << toString(getParams().getInt(MpqcParser_Parameters::maxiterParam)) << endl;
        *file << "\t\t\tmemory = " << toString(getParams().getInt(MpqcParser_Parameters::memoryParam)) << endl;
        *file << "\t\t\tintegrals<" << getParams().getString(MpqcParser_Parameters::integrationParam) << ">:()" << endl;
        *file << "\t\t)" << endl;
        *file << "\t)" << endl;
        break;
      default:
        DoeLog(0) && (eLog() << Verbose(0)
            << "Unknown level of theory requested for MPQC output file." << std::endl);
        break;
    }
    *file << ")" << endl;
    *file << "molecule<Molecule>: (" << endl;
    *file << "\tunit = " << (World::getInstance().getConfig()->GetIsAngstroem() ? "angstrom" : "bohr" ) << endl;
    *file << "\t{ atoms geometry } = {" << endl;
    // output of atoms
    for (std::vector<atom *>::const_iterator AtomRunner = atoms.begin(); AtomRunner != atoms.end(); ++AtomRunner) {
      (*AtomRunner)->OutputMPQCLine(file, &center);
    }
    *file << "\t}" << endl;
    *file << ")" << endl;
    *file << "basis<GaussianBasisSet>: (" << endl;
    *file << "\tname = \"" << getParams().getString(MpqcParser_Parameters::basisParam) << "\"" << endl;
    *file << "\tmolecule = $:molecule" << endl;
    *file << ")" << endl;
    if (getParams().getTheory() == MpqcParser_Parameters::MBPT2_R12) {
      *file << "% auxiliary basis set specification" << endl;
      *file << "\tabasis<GaussianBasisSet>: (" << endl;
      *file << "\tname = \"" << getParams().getString(MpqcParser_Parameters::aux_basisParam) << "\"" << endl;
      *file << "\tmolecule = $:molecule" << endl;
      *file << ")" << endl;
    }
  }
}