fileIO.py

"""For parsing input/output files"""
import os
import re
import sys
from copy import deepcopy
from io import IOBase, StringIO
from math import ceil

import numpy as np

from AaronTools import addlogger
from AaronTools.atoms import Atom
from AaronTools.const import ELEMENTS, PHYSICAL, UNIT
from AaronTools.orbitals import Orbitals
from AaronTools.spectra import Frequency, ValenceExcitations
from AaronTools.theory import *
from AaronTools.utils.utils import (
    is_alpha,
    is_int,
    is_num,
    float_num,
)

read_types = [
    "xyz",
    "log",
    "com",
    "gjf",
    "sd",
    "sdf",
    "mol",
    "mol2",
    "out",
    "dat",
    "fchk",
    "crest",
    "xtb",
    "sqmout",
    "47",
    "31",
    "qout",
]
write_types = ["xyz", "com", "inp", "inq", "in", "sqmin", "cube", "xtb", "crest"]
file_type_err = "File type not yet implemented: {}"
NORM_FINISH = "Normal termination"
ORCA_NORM_FINISH = "****ORCA TERMINATED NORMALLY****"
PSI4_NORM_FINISH = "*** Psi4 exiting successfully. Buy a developer a beer!"
ERROR = {
    "Convergence failure -- run terminated.": "SCF_CONV",
    "Inaccurate quadrature in CalDSu": "CONV_CDS",
    "Error termination request processed by link 9999": "CONV_LINK",
    "FormBX had a problem": "FBX",
    "NtrErr Called from FileIO": "CHK",
    "Wrong number of Negative eigenvalues": "EIGEN",
    "Erroneous write": "QUOTA",
    "Atoms too close": "CLASH",
    "Small interatomic distances encountered:": "CLASH",
    "The combination of multiplicity": "CHARGEMULT",
    "Bend failed for angle": "REDUND",
    "Linear angle in Bend": "REDUND",
    "Error in internal coordinate system": "COORD",
    "galloc: could not allocate memory": "GALLOC",
    "Error imposing constraints": "CONSTR",
    "End of file reading basis center.": "BASIS_READ",
    "Atomic number out of range for .* basis set.": "BASIS",
    "Unrecognized atomic symbol": "ATOM",
    "malloc failed.": "MEM",
    "A syntax error was detected in the input line": "SYNTAX",
    "Unknown message": "UNKNOWN",
}

ERROR_ORCA = {
    "SCF NOT CONVERGED AFTER": "SCF_CONV",
    # ORCA doesn't actually exit if the SCF doesn't converge...
    # "CONV_CDS": "",
    "The optimization did not converge but reached the maximum number": "OPT_CONV",
    # ORCA still prints the normal finish line if opt doesn't converge...
    # "FBX": "",
    # "CHK": "",
    # "EIGEN": "", <- ORCA doesn't seem to have this
    # "QUOTA": "",
    "Zero distance between atoms": "CLASH",  # <- only get an error if atoms are literally on top of each other
    "Error : multiplicity": "CHARGEMULT",
    # "REDUND": "",
    # "REDUND": "",
    # "GALLOC": "",
    # "CONSTR": "",
    "The basis set was either not assigned or not available for this element": "BASIS",
    "Element name/number, dummy atom or point charge expected": "ATOM",
    "Error  (ORCA_SCF): Not enough memory available!": "MEM",
    "WARNING: Analytical MP2 frequency calculations": "NUMFREQ",
    "WARNING: Analytical Hessians are not yet implemented for meta-GGA functionals": "NUMFREQ",
    "ORCA finished with error return": "UNKNOWN",
    "UNRECOGNIZED OR DUPLICATED KEYWORD(S) IN SIMPLE INPUT LINE": "TYPO",
}

# some exceptions are listed in https://psicode.org/psi4manual/master/_modules/psi4/driver/p4util/exceptions.html
ERROR_PSI4 = {
    "PsiException: Could not converge SCF iterations": "SCF_CONV",
    "psi4.driver.p4util.exceptions.SCFConvergenceError: Could not converge SCF iterations": "SCF_CONV",
    "OptimizationConvergenceError": "OPT_CONV",
    "TDSCFConvergenceError": "TDCF_CONV",
    "The INTCO_EXCEPTion handler": "INT_COORD",
    # ^ this is basically psi4's FBX
    # "CONV_CDS": "",
    # "CONV_LINK": "",
    # "FBX": "",
    # "CHK": "",
    # "EIGEN": "", <- psi4 doesn't seem to have this
    # "QUOTA": "",
    # "ValidationError:": "INPUT", <- generic input error, CHARGEMULT and CLASH would also get caught by this
    "qcelemental.exceptions.ValidationError: Following atoms are too close:": "CLASH",
    "qcelemental.exceptions.ValidationError: Inconsistent or unspecified chg/mult": "CHARGEMULT",
    "MissingMethodError": "INVALID_METHOD",
    # "REDUND": "",
    # "REDUND": "",
    # "GALLOC": "",
    # "CONSTR": "",
    "psi4.driver.qcdb.exceptions.BasisSetNotFound: BasisSet::construct: Unable to find a basis set for": "BASIS",
    "qcelemental.exceptions.NotAnElementError": "ATOM",
    "psi4.driver.p4util.exceptions.ValidationError: set_memory()": "MEM",
    # ERROR_PSI4[""] = "UNKNOWN",
    "Could not converge backtransformation.": "ICOORDS",
}


def step2str(step):
    if int(step) == step:
        return str(int(step))
    else:
        return str(step).replace(".", "-")

def str2step(step_str):
    if "-" in step_str:
        return float(step_str.replace("-", "."))
    else:
        return float(step_str)

def expected_inp_ext(exec_type):
    """
    extension expected for an input file for exec_type
    Gaussian - .com (.gjf on windows)
    ORCA - .inp
    Psi4 - .in
    SQM - .mdin
    qchem - .inp
    """
    if exec_type.lower() == "gaussian":
        if sys.platform.startswith("win"):
            return ".gjf"
        return ".com"
    if exec_type.lower() == "orca":
        return ".inp"
    if exec_type.lower() == "psi4":
        return ".in"
    if exec_type.lower() == "sqm":
        return ".mdin"
    if exec_type.lower() == "qchem":
        return ".inp"

def expected_out_ext(exec_type):
    """
    extension expected for an input file for exec_type
    Gaussian - .log
    ORCA - .out
    Psi4 - .out
    SQM - .mdout
    qchem - .out
    """
    if exec_type.lower() == "gaussian":
        return ".log"
    if exec_type.lower() == "orca":
        return ".out"
    if exec_type.lower() == "psi4":
        return ".out"
    if exec_type.lower() == "sqm":
        return ".mdout"
    if exec_type.lower() == "qchem":
        return ".out"


class FileWriter:
    @classmethod
    def write_file(
        cls, geom, style=None, append=False, outfile=None, *args, **kwargs
    ):
        """
        Writes file from geometry in the specified style

        :geom: the Geometry to use
        :style: the file type style to generate
            Currently supported options: xyz (default), com, inp, in
            if outfile has one of these extensions, default is that style
        :append: for *.xyz, append geometry to the same file
        :outfile: output destination - default is
                  [geometry name] + [extension] or [geometry name] + [step] + [extension]
                  if outfile is False, no output file will be written, but the contents will be returned
        :theory: for com, inp, and in files, an object with a get_header and get_footer method
        """
        if isinstance(outfile, str) and style is None:
            name, ext = os.path.splitext(outfile)
            style = ext.strip(".")

        elif style is None:
            style = "xyz"

        if style.lower() not in write_types:
            if style.lower() == "gaussian":
                style = "com"
            elif style.lower() == "orca":
                style = "inp"
            elif style.lower() == "psi4":
                style = "in"
            elif style.lower() == "sqm":
                style = "sqmin"
            elif style.lower() == "qchem":
                style = "inp"
            else:
                raise NotImplementedError(file_type_err.format(style))

        if (
            outfile is None
            and os.path.dirname(geom.name)
            and not os.access(os.path.dirname(geom.name), os.W_OK)
        ):
            os.makedirs(os.path.dirname(geom.name))
        if style.lower() == "xyz":
            out = cls.write_xyz(geom, append, outfile=outfile)
        elif style.lower() == "com":
            if "theory" in kwargs:
                theory = kwargs["theory"]
                del kwargs["theory"]
                out = cls.write_com(geom, theory, outfile=outfile, **kwargs)
            else:
                raise TypeError(
                    "when writing 'com/gjf' files, **kwargs must include: theory=Aaron.Theory() (or AaronTools.Theory())"
                )
        elif style.lower() == "inp":
            if "theory" in kwargs:
                theory = kwargs["theory"]
                del kwargs["theory"]
                out = cls.write_inp(geom, theory, outfile=outfile, **kwargs)
            else:
                raise TypeError(
                    "when writing 'inp' files, **kwargs must include: theory=Aaron.Theory() (or AaronTools.Theory())"
                )
        elif style.lower() == "in":
            if "theory" in kwargs:
                theory = kwargs["theory"]
                del kwargs["theory"]
                out = cls.write_in(geom, theory, outfile=outfile, **kwargs)
            else:
                raise TypeError(
                    "when writing 'in' files, **kwargs must include: theory=Aaron.Theory() (or AaronTools.Theory())"
                )
        elif style.lower() == "sqmin":
            if "theory" in kwargs:
                theory = kwargs["theory"]
                del kwargs["theory"]
                out = cls.write_sqm(geom, theory, outfile=outfile, **kwargs)
            else:
                raise TypeError(
                    "when writing 'sqmin' files, **kwargs must include: theory=Aaron.Theory() (or AaronTools.Theory())"
                )
        elif style.lower() == "inq":
            if "theory" in kwargs:
                theory = kwargs["theory"]
                del kwargs["theory"]
                out = cls.write_inq(geom, theory, outfile=outfile, **kwargs)
            else:
                raise TypeError(
                    "when writing 'inq' files, **kwargs must include: theory=Aaron.Theory() (or AaronTools.Theory())"
                )
        elif style.lower() == "xtb":
            if "theory" in kwargs:
                theory = kwargs["theory"]
                del kwargs["theory"]
                out = cls.write_xtb(geom, theory, outfile=outfile, **kwargs)
            else:
                raise TypeError(
                    "when writing 'xtb' files, **kwargs must include: theory=Aaron.Theory() (or AaronTools.Theory())"
                )
        elif style.lower() == "crest":
            if "theory" in kwargs:
                theory = kwargs["theory"]
                del kwargs["theory"]
                out = cls.write_crest(geom, theory, outfile=outfile, **kwargs)
            else:
                raise TypeError(
                    "when writing 'crest' files, **kwargs must include: theory=Aaron.Theory() (or AaronTools.Theory())"
                )
        
        elif style.lower() == "cube":
            out = cls.write_cube(geom, outfile=outfile, **kwargs)

        return out

    @classmethod
    def write_xyz(cls, geom, append, outfile=None):
        mode = "a" if append else "w"
        fmt = "{:3s} {: 10.5f} {: 10.5f} {: 10.5f}\n"
        s = "%i\n" % len(geom.atoms)
        s += "%s\n" % geom.comment
        for atom in geom.atoms:
            s += fmt.format(atom.element, *atom.coords)

        if outfile is None:
            # if no output file is specified, use the name of the geometry
            with open(geom.name + ".xyz", mode) as f:
                f.write(s)
        elif outfile is False:
            # if no output file is desired, just return the file contents
            return s.strip()
        else:
            # write output to the requested destination
            with open(outfile, mode) as f:
                f.write(s)

        return

    @classmethod
    def write_com(
        cls, geom, theory, outfile=None, return_warnings=False, **kwargs
    ):
        """
        write Gaussian input file for given Theory() and Geometry()
        geom - Geometry()
        theory - Theory()
        outfile - None, False, or str
                  None - geom.name + ".com" is used as output destination
                  False - return contents of the input file as a str
                  str - output destination
        return_warnings - True to return a list of warnings (e.g. basis
                          set might be misspelled
        kwargs - passed to Theory methods (make_header, make_molecule, etc.)
        """
        # get file content string
        header, header_warnings = theory.make_header(
            geom, return_warnings=True, **kwargs
        )
        mol, mol_warnings = theory.make_molecule(
            geom, return_warnings=True, **kwargs
        )
        footer, footer_warnings = theory.make_footer(
            geom, return_warnings=True, **kwargs
        )

        s = header + mol + footer
        warnings = header_warnings + mol_warnings + footer_warnings

        if outfile is None:
            # if outfile is not specified, name file in Aaron format
            if "step" in kwargs:
                outfile = "{}.{}.com".format(geom.name, step2str(kwargs["step"]))
            else:
                outfile = "{}.com".format(geom.name)
        if outfile is False:
            if return_warnings:
                return s, warnings
            return s
        else:
            fname = os.path.basename(outfile)
            name, ext = os.path.splitext(fname)
            # could use jinja, but it's one thing...
            s = s.replace("{{ name }}", name)
            with open(outfile, "w") as f:
                f.write(s)

        if return_warnings:
            return warnings
        return

    @classmethod
    def write_inp(
        cls, geom, theory, outfile=None, return_warnings=False, **kwargs
    ):
        """
        write ORCA input file for the given Theory() and Geometry()
        geom - Geometry()
        theory - Theory()
        outfile - None, False, or str
                  None - geom.name + ".inp" is used as output destination
                  False - return contents of the input file as a str
                  str - output destination
        return_warnings - True to return a list of warnings (e.g. basis
                          set might be misspelled
        kwargs - passed to Theory methods (make_header, make_molecule, etc.)
        """
        fmt = "{:<3s} {: 9.5f} {: 9.5f} {: 9.5f}\n"
        header, warnings = theory.make_header(
            geom, style="orca", return_warnings=True, **kwargs
        )
        footer = theory.make_footer(
            geom, style="orca", return_warnings=False, **kwargs
        )
        s = header
        for atom in geom.atoms:
            s += fmt.format(atom.element, *atom.coords)

        s += "*\n"

        s += footer
        
        if outfile is None:
            # if outfile is not specified, name file in Aaron format
            if "step" in kwargs:
                outfile = "{}.{}.inp".format(geom.name, step2str(kwargs["step"]))
            else:
                outfile = "{}.inp".format(geom.name)
        if outfile is False:
            if return_warnings:
                return s, warnings
            return s
        else:
            fname = os.path.basename(outfile)
            name, ext = os.path.splitext(fname)
            # could use jinja, but it's one thing...
            s = s.replace("{{ name }}", name)
            with open(outfile, "w") as f:
                f.write(s)
        
        if return_warnings:
            return warnings

    @classmethod
    def write_inq(
        cls, geom, theory, outfile=None, return_warnings=False, **kwargs
    ):
        """
        write QChem input file for the given Theory() and Geometry()
        geom - Geometry()
        theory - Theory()
        outfile - None, False, or str
                  None - geom.name + ".inq" is used as output destination
                  False - return contents of the input file as a str
                  str - output destination
        return_warnings - True to return a list of warnings (e.g. basis
                          set might be misspelled
        kwargs - passed to Theory methods (make_header, make_molecule, etc.)
        """
        fmt = "{:<3s} {: 9.5f} {: 9.5f} {: 9.5f}\n"
        header, header_warnings = theory.make_header(
            geom, style="qchem", return_warnings=True, **kwargs
        )
        mol, mol_warnings = theory.make_molecule(
            geom, style="qchem", return_warnings=True, **kwargs
        )

        out = header + mol
        warnings = header_warnings + mol_warnings
        
        if outfile is None:
            # if outfile is not specified, name file in Aaron format
            if "step" in kwargs:
                outfile = "{}.{}.inq".format(geom.name, step2str(kwargs["step"]))
            else:
                outfile = "{}.inq".format(geom.name)
        if outfile is False:
            if return_warnings:
                return out, warnings
            return out
        else:
            fname = os.path.basename(outfile)
            name, ext = os.path.splitext(fname)
            # could use jinja, but it's one thing...
            out = out.replace("{{ name }}", name)
            with open(outfile, "w") as f:
                f.write(out)
        
        if return_warnings:
            return warnings

    @classmethod
    def write_in(
        cls, geom, theory, outfile=None, return_warnings=False, **kwargs
    ):
        """
        write Psi4 input file for the given Theory() and Geometry()
        geom - Geometry()
        theory - Theory()
        outfile - None, False, or str
                  None - geom.name + ".com" is used as output destination
                  False - return contents of the input file as a str
                  str - output destination
        return_warnings - True to return a list of warnings (e.g. basis
                          set might be misspelled
        kwargs - passed to Theory methods (make_header, make_molecule, etc.)
        """
        header, header_warnings = theory.make_header(
            geom, style="psi4", return_warnings=True, **kwargs
        )
        mol, mol_warnings = theory.make_molecule(
            geom, style="psi4", return_warnings=True, **kwargs
        )
        footer, footer_warnings = theory.make_footer(
            geom, style="psi4", return_warnings=True, **kwargs
        )

        s = header + mol + footer
        warnings = header_warnings + mol_warnings + footer_warnings

        if outfile is None:
            # if outfile is not specified, name file in Aaron format
            if "step" in kwargs:
                outfile = "{}.{}.in".format(geom.name, step2str(kwargs["step"]))
            else:
                outfile = "{}.in".format(geom.name)
        if outfile is False:
            if return_warnings:
                return s, warnings
            return s
        else:
            fname = os.path.basename(outfile)
            name, ext = os.path.splitext(fname)
            # could use jinja, but it's one thing...
            s = s.replace("{{ name }}", name)
            with open(outfile, "w") as f:
                f.write(s)
        
        if return_warnings:
            return warnings

    @classmethod
    def write_sqm(
        cls, geom, theory, outfile=None, return_warnings=False, **kwargs
    ):
        """
        write SQM input file for the given Theory() and Geometry()
        geom - Geometry()
        theory - Theory()
        outfile - None, False, or str
                  None - geom.name + ".com" is used as output destination
                  False - return contents of the input file as a str
                  str - output destination
        return_warnings - True to return a list of warnings (e.g. basis
                          set might be misspelled
        kwargs - passed to Theory methods (make_header, make_molecule, etc.)
        """
        header, header_warnings = theory.make_header(
            geom, style="sqm", return_warnings=True, **kwargs
        )
        mol, mol_warnings = theory.make_molecule(
            geom, style="sqm", return_warnings=True, **kwargs
        )

        s = header + mol
        warnings = header_warnings + mol_warnings

        if outfile is None:
            # if outfile is not specified, name file in Aaron format
            if "step" in kwargs:
                outfile = "{}.{}.com".format(
                    geom.name, step2str(kwargs["step"])
                )
            else:
                outfile = "{}.com".format(geom.name)
        if outfile is False:
            if return_warnings:
                return s, warnings
            return s
        else:
            fname = os.path.basename(outfile)
            name, ext = os.path.splitext(fname)
            # could use jinja, but it's one thing...
            s = s.replace("{{ name }}", name)
            with open(outfile, "w") as f:
                f.write(s)

        if return_warnings:
            return warnings

    @classmethod
    def write_cube(
        cls,
        geom,
        orbitals=None,
        outfile=None,
        kind="homo",
        padding=4.0,
        spacing=0.2,
        alpha=True,
        xyz=False,
        n_jobs=1,
        delta=0.1,
        **kwargs,
    ):
        """
        write a cube file for a molecular orbital
        geom - geometry
        orbitals - Orbitals()
        outfile - output destination
        mo - index of molecular orbital or "homo" for ground state
             highest occupied molecular orbital or "lumo" for first
             ground state unoccupied MO
             can also be an array of MO coefficients
        ao - index of atomic orbital to print
        padding - padding around geom's coordinates
        spacing - targeted spacing between points
        n_jobs - number of parallel threads to use
                 this is on top of NumPy's multithreading, so
                 if NumPy uses 8 threads and n_jobs=2, you can
                 expect to see 16 threads in use
        delta - see Orbitals.fukui_donor_value or fukui_acceptor_value
        """
        if orbitals is None:
            raise RuntimeError(
                "no Orbitals() instance given to FileWriter.write_cube"
            )

        n_pts1, n_pts2, n_pts3, v1, v2, v3, com, u = orbitals.get_cube_array(
            geom,
            standard_axes=xyz,
            spacing=spacing,
            padding=padding,
        )

        mo = None
        if kind.lower() == "homo":
            mo = max(orbitals.n_alpha, orbitals.n_beta) - 1
        elif kind.lower() == "lumo":
            mo = max(orbitals.n_alpha, orbitals.n_beta)
        elif kind.lower().startswith("mo"):
            mo = int(kind.split()[-1])
        elif kind.lower().startswith("ao"):
            mo = np.zeros(orbitals.n_mos)
            mo[int(kind.split()[-1])] = 1
        
        s = ""
        s += " %s\n" % geom.comment
        s += " %s\n" % kind

        # the '-' in front of the number of atoms indicates that this is
        # MO info so there's an extra data entry between the molecule
        # and the function values
        bohr_com = com / UNIT.A0_TO_BOHR
        if isinstance(mo, int):
            s += " -"
        else:
            s += "  "
        s += "%i %13.5f %13.5f %13.5f 1\n" % (
            len(geom.atoms), *bohr_com,
        )

        # the basis vectors of cube files are ordered based on the
        # spacing between points along that axis
        # or maybe it's the number of points?
        # we use the first one
        for n, v in sorted(
            zip([n_pts1, n_pts2, n_pts3], [v1, v2, v3]),
            key=lambda p: np.linalg.norm(p[1]),
        ):
            bohr_v = v / UNIT.A0_TO_BOHR
            s += " %5i %13.5f %13.5f %13.5f\n" % (
                n, *bohr_v
            )
        # contruct an array of points for the grid
        coords, n_list = orbitals.get_cube_points(
            n_pts1, n_pts2, n_pts3, v1, v2, v3, com
        )

        # write the structure in bohr
        for atom in geom.atoms:
            s += " %5i %13.5f %13.5f %13.5f %13.5f\n" % (
                ELEMENTS.index(atom.element),
                ELEMENTS.index(atom.element),
                atom.coords[0] / UNIT.A0_TO_BOHR,
                atom.coords[1] / UNIT.A0_TO_BOHR,
                atom.coords[2] / UNIT.A0_TO_BOHR,
            )

        # extra section - only for MO data
        if isinstance(mo, int):
            s += " %5i %5i\n" % (1, mo + 1)

        # get values for this MO
        if kind.lower() == "density":
            val = orbitals.density_value(coords, n_jobs=n_jobs)
        elif kind.lower() == "fukui donor":
            val = orbitals.fukui_donor_value(
                coords, n_jobs=n_jobs, delta=delta
            )
        elif kind.lower() == "fukui acceptor":
            val = orbitals.fukui_acceptor_value(
                coords, n_jobs=n_jobs, delta=delta
            )
        elif kind.lower() == "fukui dual":
            val = orbitals.fukui_dual_value(
                coords, n_jobs=n_jobs, delta=delta
            )
        else:
            val = orbitals.mo_value(mo, coords, n_jobs=n_jobs)

        # write to a file
        for n1 in range(0, n_list[0]):
            for n2 in range(0, n_list[1]):
                val_ndx = n1 * n_list[2] * n_list[1] + n2 * n_list[2]
                val_subset = val[val_ndx : val_ndx + n_list[2]]
                for i, v in enumerate(val_subset):
                    if abs(v) < 1e-30:
                        v = 0
                    s += "%13.5e" % v
                    if (i + 1) % 6 == 0:
                        s += "\n"
                if (i + 1) % 6 != 0:
                    s += "\n"

        if outfile is None:
            # if no output file is specified, use the name of the geometry
            with open(geom.name + ".cube", "w") as f:
                f.write(s)
        elif outfile is False:
            # if no output file is desired, just return the file contents
            return s
        else:
            # write output to the requested destination
            with open(outfile, "w") as f:
                f.write(s)
        return

    @classmethod
    def write_xtb(
        cls,
        geom,
        theory,
        outfile=None,
        return_warnings=False,
        **kwargs,
    ):
        if theory.job_type:
            for job in theory.job_type:
                if hasattr(job, "geometry"):
                    job.geometry = geom
                    
        contents = dict()
        cli, cli_warnings = theory.get_xtb_cmd(
            return_warnings=True, **kwargs
        )
        contents["cmd"] = cli
        xcontrol, xc_warnings, write_ref = theory.get_xtb_control(
            return_warnings=True, **kwargs
        )
        contents["xc"] = xcontrol

        contents["xyz"] = cls.write_xyz(geom, append=False, outfile=False)

        warnings = cli_warnings + xc_warnings
        
        if write_ref:
            contents[write_ref] = contents["xyz"]

        if outfile is False:
            if return_warnings:
                return contents, warnings
            return contents
        
        if outfile is None:
            if "step" in kwargs:
                outfile = "{}.{}".format(geom.name, step2str(kwargs["step"]))
            else:
                outfile = geom.name
        
        dirname, basename = os.path.split(outfile)
        name, ext = os.path.splitext(basename)

        cls.write_dict_files(contents, dirname, name)
        
        if return_warnings:
            return warnings

    @classmethod
    def write_crest(
        cls,
        geom,
        theory,
        outfile=None,
        return_warnings=False,
        **kwargs,
    ):
        if theory.job_type:
            for job in theory.job_type:
                if hasattr(job, "geometry"):
                    job.geometry = geom
                 
        contents = dict()
        cli, cli_warnings = theory.get_crest_cmd(
            return_warnings=True, **kwargs
        )
        contents["cmd"] = cli
        xcontrol, xc_warnings, write_ref = theory.get_xtb_control(
            return_warnings=True, crest=True, **kwargs
        )
        contents["xc"] = xcontrol
        contents["xyz"] = cls.write_xyz(geom, append=False, outfile=False)
        if write_ref:
            contents[write_ref] = contents["xyz"]
 
        warnings = cli_warnings + xc_warnings
 
        if outfile is False:
            if return_warnings:
                return contents, warnings
            return contents
        
        if outfile is None:
            if "step" in kwargs:
                outfile = "{}.{}".format(geom.name, step2str(kwargs["step"]))
            else:
                outfile = geom.name

        dirname, basename = os.path.split(outfile)
        name, ext = os.path.splitext(basename)

        cls.write_dict_files(contents, dirname, name)
        
        if return_warnings:
            return warnings

    @staticmethod
    def write_dict_files(contents, dirname, name):
        """
        write data to different files
        contents - dict(), keys are either a file name (includes a ".") or
            a file extension (no ".")
        dirname - where to write files
        e.g. calling with contents as
        {"run.sh": "cat {{ name }}.txt", "txt": "hi"}
        and name as "test"
        will write run.sh and test.txt to dirname
        """
        for key, data in contents.items():
            if "." in key:
                output_path = os.path.join(dirname, key)
            else:
                output_path = os.path.join(dirname, "%s.%s" % (name, key))
                
            with open(output_path, "w") as f:
                f.write(data.replace("{{ name }}", name))


@addlogger
class FileReader:
    """
    Attributes:
        name ''
        file_type ''
        comment ''
        atoms [Atom]
        other {}
    """

    LOG = None
    LOGLEVEL = "DEBUG"

    def __init__(
        self,
        fname,
        get_all=False,
        just_geom=True,
        freq_name=None,
        conf_name=None,
        nbo_name=None,
        max_length=10000000,
    ):
        """
        :fname: either a string specifying the file name of the file to read
            or a tuple of (str(name), str(file_type), str(content))
        :get_all: if true, optimization steps are  also saved in
            self.all_geom; otherwise only saves last geometry
        :just_geom: if true, does not store other information, such as
            frequencies, only what is needed to construct a Geometry() obj
        :freq_name: Name of the file containing the frequency output. Only use
            if this information is in a different file than `fname` (eg: xtb runs
            using the --hess runtype option)
        :nbo_name: Name of the file containing the NBO orbital coefficients
            in the AO basis. Only used when reading *.47 files.
        :max_length: maximum array size to store from FCHK files
            any array that would be larger than this will be the
            size the array would be
        """
        # Initialization
        self.name = ""
        self.file_type = ""
        self.comment = ""
        self.atoms = []
        self.other = {}
        self.content = None
        self.all_geom = None

        # get file name and extention
        if isinstance(fname, str):
            self.name, self.file_type = os.path.splitext(fname)
            self.file_type = self.file_type.lower()[1:]
        elif isinstance(fname, (tuple, list)):
            self.name = fname[0]
            self.file_type = fname[1]
            self.content = fname[2]
        if self.file_type not in read_types:
            raise NotImplementedError(file_type_err.format(self.file_type))

        # Fill in attributes with geometry information
        if self.content is None:
            self.read_file(
                get_all, just_geom,
                freq_name=freq_name,
                conf_name=conf_name,
                nbo_name=nbo_name,
                max_length=max_length,
            )
        elif isinstance(self.content, str):
            f = StringIO(self.content)
        elif isinstance(self.content, IOBase):
            f = self.content

        if self.content is not None:
            if self.file_type == "log":
                self.read_log(f, get_all, just_geom)
            elif any(self.file_type == ext for ext in ["sd", "sdf", "mol"]):
                self.read_sd(f)
            elif self.file_type == "xyz":
                self.read_xyz(f, get_all)
            elif self.file_type == "mol2":
                self.read_mol2(f, get_all)
            elif any(self.file_type == ext for ext in ["com", "gjf"]):
                self.read_com(f)
            elif self.file_type == "out":
                self.read_orca_out(f, get_all, just_geom)
            elif self.file_type == "dat":
                self.read_psi4_out(f, get_all, just_geom)
            elif self.file_type == "fchk":
                self.read_fchk(f, just_geom, max_length=max_length)
            elif self.file_type == "crest":
                self.read_crest(f, conf_name=conf_name)
            elif self.file_type == "xtb":
                self.read_xtb(f, freq_name=freq_name)
            elif self.file_type == "sqmout":
                self.read_sqm(f)
            elif self.file_type == "47":
                self.read_nbo_47(f, nbo_name=nbo_name)
            elif self.file_type == "31":
                self.read_nbo_31(f, nbo_name=nbo_name)
            elif self.file_type == "qout":
                self.read_qchem_out(f, get_all, just_geom)

    def __getitem__(self, key):
        if hasattr(self, key):
            return getattr(self, key)
        return self.other[key]
    
    def __setitem__(self, key, value):
        if hasattr(self, key):
            setattr(self, key, value)
        self.other[key] = value

    def __contains__(self, key):
        if hasattr(self, key):
            return True
        return key in self.other

    def __delitem__(self, key):
        if hasattr(self, key):
            del self.key
        if key in self.other:
            del self.other["key"]
    
    def keys(self):
        attr_keys = set(self.__dict__.keys())
        other_keys = set(self.other.keys())
        return tuple(attr_keys.union(other_keys))
    
    def values(self):
        keys = self.keys()
        return tuple(self[key] for key in keys)
    
    def items(self):
        keys = self.keys()
        return tuple((key, self[key]) for key in keys)
    
    def read_file(
        self, get_all=False, just_geom=True,
        freq_name=None, conf_name=None, nbo_name=None,
        max_length=10000000,
    ):
        """
        Reads geometry information from fname.
        Parameters:
            get_all     If false (default), only keep the last geom
                        If true, self is last geom, but return list
                            of all others encountered
            nbo_name    nbo output file containing coefficients to
                        map AO's to orbitals
            max_length  max. array size for arrays to store in FCHK
                        files - anything larger will be the size
                        the array would be
        """
        if os.path.isfile(self.name):
            f = open(self.name, "r", encoding="utf8")
        else:
            fname = ".".join([self.name, self.file_type])
            fname = os.path.expanduser(fname)
            if os.path.isfile(fname):
                f = open(fname, "r", encoding="utf8")
            else:
                raise FileNotFoundError(
                    "Error while looking for %s: could not find %s or %s in %s"
                    % (self.name, fname, self.name, os.getcwd())
                )

        if self.file_type == "xyz":
            self.read_xyz(f, get_all)
        elif self.file_type == "log":
            self.read_log(f, get_all, just_geom)
        elif any(self.file_type == ext for ext in ["com", "gjf"]):
            self.read_com(f)
        elif any(self.file_type == ext for ext in ["sd", "sdf", "mol"]):
            self.read_sd(f)
        elif self.file_type == "mol2":
            self.read_mol2(f)
        elif self.file_type == "out":
            self.read_orca_out(f, get_all, just_geom)
        elif self.file_type == "dat":
            self.read_psi4_out(f, get_all, just_geom)
        elif self.file_type == "fchk":
            self.read_fchk(f, just_geom, max_length=max_length)
        elif self.file_type == "crest":
            self.read_crest(f, conf_name=conf_name)
        elif self.file_type == "xtb":
            self.read_xtb(f, freq_name=freq_name)
        elif self.file_type == "sqmout":
            self.read_sqm(f)
        elif self.file_type == "47":
            self.read_nbo_47(f, nbo_name=nbo_name)
        elif self.file_type == "31":
            self.read_nbo_31(f, nbo_name=nbo_name)
        elif self.file_type == "qout":
            self.read_qchem_out(f, get_all, just_geom)

        f.close()
        return

    def skip_lines(self, f, n):
        for i in range(n):
            f.readline()
        return

    def read_xyz(self, f, get_all=False):
        self.all_geom = []
        # number of atoms
        f.readline()
        # comment
        self.comment = f.readline().strip()
        # atom info
        atom_count = 0
        for line in f:
            line = line.strip()
            if line == "":
                continue
            try:
                int(line)
                if get_all:
                    self.all_geom += [
                        (deepcopy(self.comment), deepcopy(self.atoms))
                    ]
                self.comment = f.readline().strip()
                self.atoms = []
                atom_count = 0
            except ValueError:
                line = line.split()
                atom_count += 1
                self.atoms += [Atom(element=line[0], coords=line[1:4], name=str(atom_count))]

        # if get_all:
        #     self.all_geom += [(deepcopy(self.comment), deepcopy(self.atoms))]

    def read_sd(self, f, get_all=False):
        self.all_geom = []
        lines = f.readlines()
        progress = 0
        for i, line in enumerate(lines):
            progress += 1
            if "$$$$" in line:
                progress = 0
                if get_all:
                    self.all_geom.append(
                        [deepcopy(self.comment), deepcopy(self.atoms)]
                    )

                continue

            if progress == 3:
                self.comment = line.strip()

            if progress == 4:
                counts = line.split()
                natoms = int(counts[0])
                nbonds = int(counts[1])

            if progress == 5:
                self.atoms = []
                for line in lines[i : i + natoms]:
                    atom_info = line.split()
                    self.atoms += [
                        Atom(element=atom_info[3], coords=atom_info[0:3])
                    ]

                for line in lines[i + natoms : i + natoms + nbonds]:
                    a1, a2 = [int(x) - 1 for x in line.split()[0:2]]
                    self.atoms[a1].connected.add(self.atoms[a2])
                    self.atoms[a2].connected.add(self.atoms[a1])

                for j, a in enumerate(self.atoms):
                    a.name = str(j + 1)
                
                self.other["charge"] = 0
                for line in lines[i + natoms + nbonds:]:
                    if "CHG" in line:
                        self.other["charge"] += int(line.split()[-1])
                    if "$$$$" in line:
                        break

    def read_mol2(self, f, get_all=False):
        """
        read TRIPOS mol2
        """
        atoms = []

        lines = f.readlines()
        i = 0
        while i < len(lines):
            if lines[i].startswith("@<TRIPOS>MOLECULE"):
                self.comment = lines[i + 1]
                info = lines[i + 2].split()
                n_atoms = int(info[0])
                n_bonds = int(info[1])
                i += 3

            elif lines[i].startswith("@<TRIPOS>ATOM"):
                for j in range(0, n_atoms):
                    i += 1
                    info = lines[i].split()
                    # name = info[1]
                    coords = np.array([float(x) for x in info[2:5]])
                    element = re.match("([A-Za-z]+)", info[5]).group(1)
                    atoms.append(
                        Atom(element=element, coords=coords, name=str(j + 1))
                    )

                self.atoms = atoms

            elif lines[i].startswith("@<TRIPOS>BOND"):
                for j in range(0, n_bonds):
                    i += 1
                    info = lines[i].split()
                    a1, a2 = [int(ndx) - 1 for ndx in info[1:3]]
                    self.atoms[a1].connected.add(self.atoms[a2])
                    self.atoms[a2].connected.add(self.atoms[a1])

            i += 1

    def read_psi4_out(self, f, get_all=False, just_geom=True):
        uv_vis = ""
        coord_unit_bohr = False
        def get_atoms(f, n, bohr):
            rv = []
            self.skip_lines(f, 1)
            n += 2
            line = f.readline()
            i = 0
            mass = 0
            while line.strip():
                i += 1
                line = line.strip()
                atom_info = line.split()
                element = atom_info[0]
                # might be a ghost atom - like for sapt
                if "Gh" in element:
                    element = element.strip("Gh(").strip(")")
                coords = np.array([float(x) for x in atom_info[1:-1]])
                if bohr:
                    coords *= UNIT.A0_TO_BOHR
                rv += [Atom(element=element, coords=coords, name=str(i))]
                mass += float(atom_info[-1])

                line = f.readline()
                n += 1

            return rv, mass, n

        line = f.readline()
        n = 1
        read_geom = False
        while line != "":
            if "* O   R   C   A *" in line:
                self.file_type = "out"
                return self.read_orca_out(
                    f, get_all=get_all, just_geom=just_geom
                )

            if "A Quantum Leap Into The Future Of Chemistry" in line:
                self.file_type = "qout"
                return self.read_qchem_out(
                    f, get_all=get_all, just_geom=just_geom
                )

            if line.startswith("    Geometry (in Angstrom), charge"):
                coord_unit_bohr = False
                if not just_geom:
                    self.other["charge"] = int(line.split()[5].strip(","))
                    self.other["multiplicity"] = int(
                        line.split()[8].strip(":")
                    )

            elif line.strip() == "SCF":
                read_geom = True

            elif line.startswith("    Geometry (in Bohr), charge"):
                coord_unit_bohr = True

            elif line.strip().startswith("Center") and read_geom:
                read_geom = False
                if get_all and len(self.atoms) > 0:
                    if self.all_geom is None:
                        self.all_geom = []

                    self.all_geom += [
                        (deepcopy(self.atoms), deepcopy(self.other))
                    ]

                self.atoms, mass, n = get_atoms(f, n, coord_unit_bohr)
                if not just_geom:
                    self.other["mass"] = mass
                    self.other["mass"] *= UNIT.AMU_TO_KG

            if just_geom:
                line = f.readline()
                n += 1
                continue
            else:
                if line.strip().startswith("Total Energy ="):
                    self.other["energy"] = float(line.split()[-1])

                elif line.strip().startswith("Total E0"):
                    self.other["energy"] = float(line.split()[-2])

                elif line.strip().startswith("Correction ZPE"):
                    self.other["ZPVE"] = float(line.split()[-4])

                elif line.strip().startswith("Total ZPE"):
                    self.other["E_ZPVE"] = float(line.split()[-2])

                elif line.strip().startswith("Total H, Enthalpy"):
                    self.other["enthalpy"] = float(line.split()[-2])

                elif line.strip().startswith("Total G, Free"):
                    self.other["free_energy"] = float(line.split()[-2])
                    self.other["temperature"] = float(line.split()[-4])

                elif "symmetry no. =" in line:
                    self.other["rotational_symmetry_number"] = int(
                        line.split()[-1].strip(")")
                    )

                elif (
                    line.strip().startswith("Rotational constants:")
                    and line.strip().endswith("[cm^-1]")
                    and "rotational_temperature" not in self.other
                ):
                    self.other["rotational_temperature"] = [
                        float(x) if is_num(x) else 0
                        for x in line.split()[-8:-1:3]
                    ]
                    self.other["rotational_temperature"] = [
                        x
                        * PHYSICAL.SPEED_OF_LIGHT
                        * PHYSICAL.PLANCK
                        / PHYSICAL.KB
                        for x in self.other["rotational_temperature"]
                    ]

                elif line.startswith("  Vibration "):
                    freq_str = ""
                    while not line.strip().startswith("=="):
                        freq_str += line
                        line = f.readline()
                        n += 1

                    self.other["frequency"] = Frequency(
                        freq_str, hpmodes=False, style="psi4"
                    )

                elif PSI4_NORM_FINISH in line:
                    self.other["finished"] = True

                elif line.startswith("    Convergence Criteria"):
                    # for tolerances:
                    # psi4 puts '*' next to converged values and 'o' in place of things that aren't monitored
                    grad = {}

                    dE_tol = line[24:38]
                    if "o" in dE_tol:
                        dE_tol = None
                    else:
                        dE_tol = dE_tol.split()[0]

                    max_f_tol = line[38:52]
                    if "o" in max_f_tol:
                        max_f_tol = None
                    else:
                        max_f_tol = max_f_tol.split()[0]

                    rms_f_tol = line[52:66]
                    if "o" in rms_f_tol:
                        rms_f_tol = None
                    else:
                        rms_f_tol = rms_f_tol.split()[0]

                    max_d_tol = line[66:80]
                    if "o" in max_d_tol:
                        max_d_tol = None
                    else:
                        max_d_tol = max_d_tol.split()[0]

                    rms_d_tol = line[80:94]
                    if "o" in rms_d_tol:
                        rms_d_tol = None
                    else:
                        rms_d_tol = rms_d_tol.split()[0]

                    line = f.readline()
                    line = f.readline()
                    n += 2

                    # for convergence:
                    # psi4 puts '*' next to converged values and 'o' next to things that aren't monitored
                    if dE_tol is not None:
                        dE_conv = line[24:38]
                        dE = float(dE_conv.split()[0])
                        grad["Delta E"] = {}
                        grad["Delta E"]["value"] = dE
                        grad["Delta E"]["converged"] = "*" in dE_conv

                    if max_f_tol is not None:
                        max_f_conv = line[38:52]
                        max_f = float(max_f_conv.split()[0])
                        grad["Max Force"] = {}
                        grad["Max Force"]["value"] = max_f
                        grad["Max Force"]["converged"] = "*" in max_f_conv

                    if rms_f_tol is not None:
                        rms_f_conv = line[52:66]
                        rms_f = float(rms_f_conv.split()[0])
                        grad["RMS Force"] = {}
                        grad["RMS Force"]["value"] = rms_f
                        grad["RMS Force"]["converged"] = "*" in rms_f_conv

                    if max_d_tol is not None:
                        max_d_conv = line[66:80]
                        max_d = float(max_d_conv.split()[0])
                        grad["Max Disp"] = {}
                        grad["Max Disp"]["value"] = max_d
                        grad["Max Disp"]["converged"] = "*" in max_d_conv

                    if rms_d_tol is not None:
                        rms_d_conv = line[80:94]
                        rms_d = float(rms_d_conv.split()[0])
                        grad["RMS Disp"] = {}
                        grad["RMS Disp"]["value"] = rms_d
                        grad["RMS Disp"]["converged"] = "*" in max_d_conv

                    self.other["gradient"] = grad

                elif "Total Gradient" in line:
                    gradient = np.zeros((len(self.atoms), 3))
                    self.skip_lines(f, 2)
                    n += 2
                    for i in range(0, len(self.atoms)):
                        n += 1
                        line = f.readline()
                        info = line.split()
                        gradient[i] = np.array([float(x) for x in info[1:]])

                    self.other["forces"] = -gradient

                elif "SAPT Results" in line:
                    self.skip_lines(f, 1)
                    n += 1
                    while "Total sSAPT" not in line:
                        n += 1
                        line = f.readline()
                        if "---" in line:
                            break
                        if len(line.strip()) > 0:
                            if "Special recipe" in line:
                                continue
                            item = line[:26].strip()
                            val = 1e-3 * float(line[34:47])
                            self.other[item] = val

                elif "SCF energy" in line:
                    self.other["SCF energy"] = float(line.split()[-1])

                elif "correlation energy" in line and "=" in line:
                    item = line.split("=")[0].strip()
                    self.other[item] = float(line.split()[-1])

                elif "Full point group" in line:
                    self.other["full_point_group"] = line.split()[-1]

                elif "Molecular point group" in line:
                    self.other["molecular_point_group"] = line.split()[-1]

                elif (
                    "total energy" in line
                    and "=" in line
                    or re.search("\(.\) energy", line)
                ):
                    item = line.split("=")[0].strip().strip("*").strip()
                    self.other[item] = float(line.split()[-1])
                    # hopefully the highest level energy gets printed last
                    self.other["energy"] = self.other[item]

                elif "Total Energy" in line and "=" in line:
                    item = line.split("=")[0].strip().strip("*").strip()
                    self.other[item] = float(line.split()[-2])
                    # hopefully the highest level energy gets printed last
                    self.other["energy"] = self.other[item]

                elif "Correlation Energy" in line and "=" in line:
                    item = line.split("=")[0].strip().strip("*").strip()
                    if "DFT Exchange-Correlation" in item:
                        self.other[item] = float(line.split()[-1])
                    else:
                        self.other[item] = float(line.split()[-2])

                elif "Ground State -> Excited State Transitions" in line:
                    self.skip_lines(f, 3)
                    n += 3
                    line = f.readline()
                    s = ""
                    while line.strip():
                        s += line
                        n += 1
                        line = f.readline()
                    
                    self.other["uv_vis"] = ValenceExcitations(s, style="psi4")

                elif "Excitation Energy" in line and "Rotatory" in line:
                    self.skip_lines(f, 2)
                    n += 2
                    line = f.readline()
                    s = ""
                    while line.strip():
                        s += line
                        n += 1
                        line = f.readline()
                    
                    self.other["uv_vis"] = ValenceExcitations(s, style="psi4")

                elif re.search("\| State\s*\d+", line):
                    # read energies from property calculation
                    uv_vis += line

                elif "Excited state properties:" in line:
                    # read osc str or rotation from property calculation
                    while line.strip():
                        uv_vis += line
                        n += 1
                        line = f.readline()
                    
                    if "Oscillator" in uv_vis or "Rotation" in uv_vis:
                        self.other["uv_vis"] = ValenceExcitations(uv_vis, style="psi4")

                if "error" not in self.other:
                    for err in ERROR_PSI4:
                        if err in line:
                            self.other["error"] = ERROR_PSI4[err]
                            self.other["error_msg"] = line.strip()

                line = f.readline()
                n += 1

        if "error" not in self.other:
            self.other["error"] = None

    def read_orca_out(self, f, get_all=False, just_geom=True):
        """read orca output file"""

        nrg_regex = re.compile("(?:[A-Za-z]+\s+)?E\((.*)\)\s*\.\.\.\s*(.*)$")

        def add_grad(grad, name, line):
            grad[name] = {}
            grad[name]["value"] = line.split()[-3]
            grad[name]["converged"] = line.split()[-1] == "YES"

        def get_atoms(f, n):
            """parse atom info"""
            rv = []
            self.skip_lines(f, 1)
            n += 2
            line = f.readline()
            i = 0
            while line.strip():
                i += 1
                line = line.strip()
                atom_info = line.split()
                element = atom_info[0]
                coords = np.array([float(x) for x in atom_info[1:]])
                rv += [Atom(element=element, coords=coords, name=str(i))]

                line = f.readline()
                n += 1

            return rv, n

        line = f.readline()
        n = 1
        while line != "":
            if (
                "Psi4: An Open-Source Ab Initio Electronic Structure Package"
                in line
            ):
                self.file_type = "dat"
                return self.read_psi4_out(
                    f, get_all=get_all, just_geom=just_geom
                )
            
            if (
                "A Quantum Leap Into The Future Of Chemistry"
                in line
            ):
                self.file_type = "qout"
                return self.read_qchem_out(
                    f, get_all=get_all, just_geom=just_geom
                )

            if (
                "Entering Gaussian System"
                in line
            ):
                self.file_type = "log"
                return self.read_log(
                    f, get_all=get_all, just_geom=just_geom
                )
            
            if line.startswith("CARTESIAN COORDINATES (ANGSTROEM)"):
                if get_all and len(self.atoms) > 0:
                    if self.all_geom is None:
                        self.all_geom = []
                    self.all_geom += [
                        (deepcopy(self.atoms), deepcopy(self.other))
                    ]

                self.atoms, n = get_atoms(f, n)

            if just_geom:
                line = f.readline()
                n += 1
                continue
            else:
                nrg = nrg_regex.match(line)
                if nrg is not None:
                    nrg_type = nrg.group(1)
                    # for some reason, ORCA prints MP2 correlation energy
                    # as E(MP2) for CC jobs
                    if nrg_type == "MP2":
                        nrg_type = "MP2 CORR"
                    self.other["E(%s)" % nrg_type] = float(nrg.group(2))

                if line.startswith("FINAL SINGLE POINT ENERGY"):
                    # if the wavefunction doesn't converge, ORCA prints a message next
                    # to the energy so we can't use line.split()[-1]
                    self.other["energy"] = float(line.split()[4])

                if line.startswith("TOTAL SCF ENERGY"):
                    self.skip_lines(f, 2)
                    line = f.readline()
                    n += 3
                    self.other["SCF energy"] = float(line.split()[3])

                elif "TOTAL ENERGY:" in line:
                    item = line.split()[-5] + " energy"
                    self.other[item] = float(line.split()[-2])

                elif "CORRELATION ENERGY" in line and "Eh" in line:
                    item = line.split()[-6] + " correlation energy"
                    self.other[item] = float(line.split()[-2])
                
                elif re.match("E\(\S+\)\s+...\s+-?\d+\.\d+$", line):
                    nrg = re.match("(E\(\S+\))\s+...\s+(-?\d+\.\d+)$", line)
                    self.other["energy"] = float(nrg.group(2))
                    self.other[nrg.group(1)] = float(nrg.group(2))

                elif line.startswith("CARTESIAN GRADIENT"):
                    gradient = np.zeros((len(self.atoms), 3))
                    self.skip_lines(f, 2)
                    n += 2
                    for i in range(0, len(self.atoms)):
                        n += 1
                        line = f.readline()
                        # orca prints a warning before gradient if some
                        # coordinates are constrained
                        if line.startswith("WARNING:"):
                            continue
                        info = line.split()
                        gradient[i] = np.array([float(x) for x in info[3:]])

                    self.other["forces"] = -gradient

                elif line.startswith("VIBRATIONAL FREQUENCIES"):
                    stage = "frequencies"
                    freq_str = "VIBRATIONAL FREQUENCIES\n"
                    self.skip_lines(f, 4)
                    n += 5
                    line = f.readline()
                    while not (stage == "THERMO" and line == "\n") and line:
                        if "--" not in line and line != "\n":
                            freq_str += line

                        if "NORMAL MODES" in line:
                            stage = "modes"
                            self.skip_lines(f, 6)
                            n += 6

                        if "RAMAN SPECTRUM" in line:
                            stage = "RAMAN"
                            self.skip_lines(f, 2)
                            n += 2

                        if "IR SPECTRUM" in line:
                            stage = "IR"
                            self.skip_lines(f, 2)
                            n += 2

                        if "THERMOCHEMISTRY" in line:
                            stage = "THERMO"

                        n += 1
                        line = f.readline()

                    self.other["frequency"] = Frequency(
                        freq_str, hpmodes=False, style="orca"
                    )

                elif line.startswith("Temperature"):
                    self.other["temperature"] = float(line.split()[2])

                elif line.startswith("Total Mass"):
                    # this may only get printed for freq jobs
                    self.other["mass"] = float(line.split()[3])
                    self.other["mass"] *= UNIT.AMU_TO_KG

                elif line.startswith(" Total Charge"):
                    self.other["charge"] = int(line.split()[-1])

                elif line.startswith(" Multiplicity"):
                    self.other["multiplicity"] = int(line.split()[-1])

                elif "rotational symmetry number" in line:
                    # TODO: make this cleaner
                    self.other["rotational_symmetry_number"] = int(
                        line.split()[-2]
                    )

                elif "Symmetry Number:" in line:
                    self.other["rotational_symmetry_number"] = int(
                        line.split()[-1]
                    )

                elif line.startswith("Zero point energy"):
                    self.other["ZPVE"] = float(line.split()[4])

                elif line.startswith("Total Enthalpy"):
                    self.other["enthalpy"] = float(line.split()[3])

                elif line.startswith("Final Gibbs"):
                    # NOTE - Orca seems to only print Grimme's Quasi-RRHO free energy
                    # RRHO can be computed in AaronTool's CompOutput by setting the w0 to 0
                    self.other["free_energy"] = float(line.split()[5])

                elif line.startswith("Rotational constants in cm-1:"):
                    # orca doesn't seem to print rotational constants in older versions
                    self.other["rotational_temperature"] = [
                        float(x) for x in line.split()[-3:]
                    ]
                    self.other["rotational_temperature"] = [
                        x
                        * PHYSICAL.SPEED_OF_LIGHT
                        * PHYSICAL.PLANCK
                        / PHYSICAL.KB
                        for x in self.other["rotational_temperature"]
                    ]

                elif "Point Group:" in line:
                    self.other["full_point_group"] = line.split()[2][:-1]

                elif "Symmetry Number" in line:
                    self.other["rotational_symmetry_number"] = int(
                        line.split()[-1]
                    )

                elif "sn is the rotational symmetry number" in line:
                    # older versions of orca print this differently
                    self.other["rotational_symmetry_number"] = int(
                        line.split()[-2]
                    )

                elif "Geometry convergence" in line:
                    grad = {}
                    self.skip_lines(f, 2)
                    n += 3
                    line = f.readline()
                    while line and re.search("\w", line):
                        if re.search("Energy\schange", line):
                            add_grad(grad, "Delta E", line)
                        elif re.search("RMS\sgradient", line):
                            add_grad(grad, "RMS Force", line)
                        elif re.search("MAX\sgradient", line):
                            add_grad(grad, "Max Force", line)
                        elif re.search("RMS\sstep", line):
                            add_grad(grad, "RMS Disp", line)
                        elif re.search("MAX\sstep", line):
                            add_grad(grad, "Max Disp", line)

                        line = f.readline()
                        n += 1

                    self.other["gradient"] = grad

                elif "MAYER POPULATION ANALYSIS" in line:
                    self.skip_lines(f, 2)
                    n += 2
                    line = f.readline()
                    data = dict()
                    headers = []
                    while line.strip():
                        info = line.split()
                        header = info[0]
                        name = " ".join(info[2:])
                        headers.append(header)
                        data[header] = (name, [])
                        line = f.readline()
                    self.skip_lines(f, 1)
                    n += 1
                    for i in range(0, len(self.atoms)):
                        line = f.readline()
                        info = line.split()[2:]
                        for header, val in zip(headers, info):
                            data[header][1].append(float(val))

                    for header in headers:
                        self.other[data[header][0]] = np.array(data[header][1])

                elif line.startswith("LOEWDIN ATOMIC CHARGES"):
                    self.skip_lines(f, 1)
                    n += 1
                    charges = np.zeros(len(self.atoms))
                    for i in range(0, len(self.atoms)):
                        line = f.readline()
                        n += 1
                        charges[i] = float(line.split()[-1])
                    self.other["Löwdin Charges"] = charges

                elif line.startswith("BASIS SET IN INPUT FORMAT"):
                    # read basis set primitive info
                    self.skip_lines(f, 3)
                    n += 3
                    line = f.readline()
                    n += 1
                    self.other["basis_set_by_ele"] = dict()
                    while "--" not in line and line != "":
                        new_gto = re.search("NewGTO\s+(\S+)", line)
                        if new_gto:
                            ele = new_gto.group(1)
                            line = f.readline()
                            n += 1
                            primitives = []
                            while "end" not in line and line != "":
                                shell_type, n_prim = line.split()
                                n_prim = int(n_prim)
                                exponents = []
                                con_coeffs = []
                                for i in range(0, n_prim):
                                    line = f.readline()
                                    n += 1
                                    info = line.split()
                                    exponent = float(info[1])
                                    con_coeff = [float(x) for x in info[2:]]
                                    exponents.append(exponent)
                                    con_coeffs.extend(con_coeff)
                                primitives.append(
                                    (
                                        shell_type,
                                        n_prim,
                                        exponents,
                                        con_coeffs,
                                    )
                                )
                                line = f.readline()
                                n += 1
                            self.other["basis_set_by_ele"][ele] = primitives
                        line = f.readline()
                        n += 1

                elif "EXCITED STATES" in line or re.search("STEOM.* RESULTS", line) or line.startswith("APPROXIMATE EOM LHS"):
                    s = ""
                    done = False
                    while not done:
                        s += line
                        n += 1
                        line = f.readline()
                        if (
                            "ORCA-CIS/TD-DFT FINISHED WITHOUT ERROR" in line or
                            re.search("TDM done", line) or
                            "TIMINGS" in line or
                            line == ""
                        ):
                            done = True
                    self.other["uv_vis"] = ValenceExcitations(s, style="orca")

                elif line.startswith("MOLECULAR ORBITALS"):
                    # read molecular orbitals
                    self.skip_lines(f, 1)
                    n += 1
                    line = f.readline()
                    self.other["alpha_coefficients"] = []
                    self.other["beta_coefficients"] = []
                    self.other["alpha_nrgs"] = []
                    self.other["beta_nrgs"] = []
                    self.other["alpha_occupancies"] = []
                    self.other["beta_occupancies"] = []
                    at_info = re.compile(
                        "\s*(\d+)\S+\s+\d+(?:s|p[xyz]|d(?:z2|xz|yz|x2y2|xy)|[fghi][\+\-]?\d+)"
                    )
                    if self.other["multiplicity"] != 1:
                        args = [
                            ("alpha_coefficients", "beta_coefficients"),
                            ("alpha_nrgs", "beta_nrgs"),
                            ("alpha_occupancies", "beta_occupancies"),
                        ]
                    else:
                        args = [
                            ("alpha_coefficients",),
                            ("alpha_nrgs",),
                            ("alpha_occupancies",),
                        ]

                    for coeff_name, nrg_name, occ_name in zip(*args):
                        self.other["shell_to_atom"] = []
                        mo_coefficients = []
                        orbit_nrgs = []
                        occupancy = []
                        while line.strip() != "":
                            at_match = at_info.match(line)
                            if at_match:
                                ndx = int(at_match.group(1))
                                self.other["shell_to_atom"].append(ndx)
                                coeffs = []
                                # there might not always be a space between the coefficients
                                # so we can't just split(), but they are formatted(-ish)
                                for coeff in re.findall("-?\d+\.\d\d\d\d\d\d", line[16:]):
                                    coeffs.append(float(coeff))
                                for coeff, mo in zip(coeffs, mo_coefficients):
                                    mo.append(coeff)
                            elif "--" not in line:
                                orbit_nrgs = occupancy
                                occupancy = [float(x) for x in line.split()]
                            elif "--" in line:
                                self.other[nrg_name].extend(orbit_nrgs)
                                self.other[occ_name].extend(occupancy)
                                if mo_coefficients:
                                    self.other[coeff_name].extend(
                                        mo_coefficients
                                    )
                                    if not all(
                                        len(coeff) == len(mo_coefficients[0])
                                        for coeff in mo_coefficients
                                    ):
                                        self.LOG.warning(
                                            "orbital coefficients may not "
                                            "have been parsed correctly"
                                        )
                                mo_coefficients = [[] for x in orbit_nrgs]
                                orbit_nrgs = []
                            line = f.readline()
                            n += 1
                        self.other[coeff_name].extend(mo_coefficients)
                        line = f.readline()

                elif line.startswith("N(Alpha)  "):
                    self.other["n_alpha"] = int(
                        np.rint(float(line.split()[2]))
                    )

                elif line.startswith("N(Beta)  "):
                    self.other["n_beta"] = int(np.rint(float(line.split()[2])))

                elif ORCA_NORM_FINISH in line:
                    self.other["finished"] = True

                # TODO E_ZPVE
                if "error" not in self.other or not self["error"]:
                    for err in ERROR_ORCA:
                        if err in line:
                            self.other["error"] = ERROR_ORCA[err]
                            self.other["error_msg"] = line.strip()
                            break
                    else:
                        self.other["error"] = False
                        if "!!!!!!!!" in line:
                            line = f.readline()
                            n += 1
                            self.other["error"] = line.strip()
                            line = f.readline()
                            n += 1
                            self.other["error_msg"] = ""
                            found_error = False
                            while "!!!!!!!" not in line:
                                for err in ERROR_ORCA:
                                    if err in line:
                                        self.other["error"] = ERROR_ORCA[err]
                                self.other["error_msg"] += line.strip() + "\n"
                                line = f.readline()
                                n += 1

                line = f.readline()
                n += 1

        if not just_geom:
            if "finished" not in self.other:
                self.other["finished"] = False

            if (
                "alpha_coefficients" in self.other
                and "basis_set_by_ele" in self.other
            ):
                self.other["orbitals"] = Orbitals(self)
        
        if "error" not in self.other:
            self.other["error"] = None

    def read_qchem_out(self, f, get_all=False, just_geom=True):
        """read qchem output file"""
        def get_atoms(f, n):
            """parse atom info"""
            rv = []
            self.skip_lines(f, 2)
            n += 1
            line = f.readline()
            i = 0
            while "--" not in line:
                i += 1
                line = line.strip()
                atom_info = line.split()
                element = atom_info[1]
                coords = np.array([float(x) for x in atom_info[2:5]])
                rv += [Atom(element=element, coords=coords, name=str(i))]

                line = f.readline()
                n += 1

            return rv, n

        def add_grad(grad, name, line):
            grad[name] = {}
            grad[name]["value"] = line.split()[-3]
            grad[name]["converged"] = line.split()[-1] == "YES"

        line = f.readline()
        n = 1
        while line != "":
            if (
                "Psi4: An Open-Source Ab Initio Electronic Structure Package"
                in line
            ):
                self.file_type = "dat"
                return self.read_psi4_out(
                    f, get_all=get_all, just_geom=just_geom
                )
            
            if "* O   R   C   A *" in line:
                self.file_type = "out"
                return self.read_orca_out(
                    f, get_all=get_all, just_geom=just_geom
                )
            
            
            if (
                "A Quantum Leap Into The Future Of Chemistry"
                in line
            ):
                self.file_type = "qout"
                return self.read_qchem_out(
                    f, get_all=get_all, just_geom=just_geom
                )
            
            if "Standard Nuclear Orientation (Angstroms)" in line:
                if get_all and len(self.atoms) > 0:
                    if self.all_geom is None:
                        self.all_geom = []
                    self.all_geom += [
                        (deepcopy(self.atoms), deepcopy(self.other))
                    ]

                self.atoms, n = get_atoms(f, n)

            if just_geom:
                line = f.readline()
                n += 1
                continue
            else:
                if "energy in the final basis set" in line:
                    self.other["energy"] = float(line.split()[-1])
                    if "SCF" in line:
                        self.other["scf_energy"] = self.other["energy"]

                if re.search(r"energy\s+=\s+-?\d+\.\d+", line):
                    info = re.search(r"\s*([\S\s]+)\s+energy\s+=\s+(-?\d+\.\d+)", line)
                    kind = info.group(1)
                    if len(kind.split()) <= 2:
                        val = float(info.group(2))
                        if "correlation" not in kind and len(kind.split()) <= 2:
                            self.other["E(%s)" % kind.split()[0]] = val
                            self.other["energy"] = val
                        else:
                            self.other["E(corr)(%s)" % kind.split()[0]] = val

                if "Total energy:" in line:
                    self.other["energy"] = float(line.split()[-2])

                #MPn energy is printed as EMPn(SDQ)
                if re.search("EMP\d(?:[A-Z]+)?\s+=\s*-?\d+.\d+$", line):
                    self.other["energy"] = float(line.split()[-1])
                    self.other["E(%s)" % line.split()[0][1:]] = self.other["energy"]

                if "Molecular Point Group" in line:
                    self.other["full_point_group"] = line.split()[3]
                
                if "Largest Abelian Subgroup" in line:
                    self.other["abelian_subgroup"] = line.split()[3]
                
                if "Ground-State Mulliken Net Atomic Charges" in line:
                    charges = []
                    self.skip_lines(f, 3)
                    n += 2
                    line = f.readline()
                    while "--" not in line:
                        charge = float(line.split()[-1])
                        charges.append(charge)
                        line = f.readline()
                        n += 1
                    
                    self.other["Mulliken Charges"] = charges

                if "Cnvgd?" in line:
                    grad = {}
                    line = f.readline()
                    while line and re.search("\w", line):
                        if re.search("Energy\schange", line):
                            add_grad(grad, "Delta E", line)
                        elif re.search("Displacement", line):
                            add_grad(grad, "Disp", line)
                        elif re.search("Gradient", line):
                            add_grad(grad, "Max Disp", line)

                        line = f.readline()
                        n += 1

                    self.other["gradient"] = grad
            
                if "VIBRATIONAL ANALYSIS" in line:
                    freq_str = ""
                    self.skip_lines(f, 10)
                    n += 9
                    line = f.readline()
                    while "STANDARD THERMODYNAMIC QUANTITIES" not in line:
                        n += 1
                        freq_str += line
                        line = f.readline()
                    self.other["frequency"] = Frequency(
                        freq_str, style="qchem",
                    )
                    self.other["temperature"] = float(line.split()[4])
    
                if "Rotational Symmetry Number is" in line:
                    self.other["rotational_symmetry_number"] = int(line.split()[-1])
    
                if "Molecular Mass:" in line:
                    self.other["mass"] = float(line.split()[-2]) * UNIT.AMU_TO_KG
    
                if "$molecule" in line.lower():
                    line = f.readline()
                    while "$end" not in line.lower() and line:
                        if re.search("\d+\s+\d+", line):
                            match = re.search("^\s*(\d+)\s+(\d+)\s*$", line)
                            self.other["charge"] = int(match.group(1))
                            self.other["multiplicity"] = int(match.group(2))
                            break
                        line = f.readline()
    
                if "Principal axes and moments of inertia" in line:
                    self.skip_lines(f, 1)
                    line = f.readline()
                    rot_consts = np.array([
                        float(x) for x in line.split()[2:]
                    ])
                    rot_consts *= UNIT.AMU_TO_KG
                    rot_consts *= UNIT.A0_TO_BOHR ** 2
                    rot_consts *= 1e-20
                    rot_consts = PHYSICAL.PLANCK ** 2 / (8 * np.pi ** 2 * rot_consts * PHYSICAL.KB)
            
                    self.other["rotational_temperature"] = rot_consts
                
                if line.startswith("Mult"):
                    self.other["multiplicity"] = int(line.split()[1])
                
                # TD-DFT excitations
                if re.search("TDDFT.* Excitation Energies", line):
                    excite_s = ""
                    self.skip_lines(f, 2)
                    line = f.readline()
                    n += 3
                    while "---" not in line and line:
                        excite_s += line
                        line = f.readline()
                        n += 1
                    
                    self.other["uv_vis"] = ValenceExcitations(
                        excite_s, style="qchem",
                    )

                # ADC excitations
                if re.search("Excited State Summary", line):
                    excite_s = ""
                    self.skip_lines(f, 2)
                    line = f.readline()
                    n += 3
                    while "===" not in line and line:
                        excite_s += line
                        line = f.readline()
                        n += 1
                    
                    self.other["uv_vis"] = ValenceExcitations(
                        excite_s, style="qchem",
                    )
                
                # EOM excitations
                if re.search("Start computing the transition properties", line):
                    excite_s = ""
                    line = f.readline()
                    n += 1
                    while "All requested transition properties have been computed" not in line and line:
                        excite_s += line
                        line = f.readline()
                        n += 1
                    
                    self.other["uv_vis"] = ValenceExcitations(
                        excite_s, style="qchem",
                    )
                
                if "Thank you very much for using Q-Chem" in line:
                    self.other["finished"] = True
                
                line = f.readline()
                n += 1
        
        if not just_geom and "finished" not in self.other:
            self.other["finished"] = False

        if "error" not in self.other:
            self.other["error"] = None

    def read_log(self, f, get_all=False, just_geom=True):
        def get_atoms(f, n):
            rv = self.atoms
            self.skip_lines(f, 4)
            line = f.readline()
            n += 5
            atnum = 0
            while "--" not in line:
                line = line.strip()
                line = line.split()
                for l in line:
                    try:
                        float(l)
                    except ValueError:
                        msg = "Error detected with log file on line {}"
                        raise IOError(msg.format(n))
                try:
                    rv[atnum].coords = np.array(line[3:], dtype=float)
                except IndexError:
                    pass
                    rv.append(Atom(
                        element=ELEMENTS[int(line[1])],
                        name=str(atnum + 1),
                    ))
                    rv[atnum].coords = np.array(line[3:], dtype=float)

                atnum += 1
                line = f.readline()
                n += 1
            return rv, n

        def get_input(f, n):
            rv = []
            line = f.readline()
            n += 1
            match = re.search(
                "Charge\s*=\s*(-?\d+)\s*Multiplicity\s*=\s*(\d+)", line
            )
            if match is not None:
                self.other["charge"] = int(match.group(1))
                self.other["multiplicity"] = int(match.group(2))
            line = f.readline()
            n += 1
            a = 0
            if len(line.split()) == 1:
                # internal coordinate z-matrix
                # not parsed
                while line.split():
                    line = line.split()
                    rv += [Atom(element=line[0], name=str(a), coords=np.zeros(3))]
                    a += 1
                    line = f.readline()
                return rv, n

            while len(line.split()) > 1:
                line  = line.split()
                if len(line) == 5:
                    flag = not(bool(line[1]))
                    a += 1
                    rv += [Atom(element=line[0], flag=flag, coords=line[2:], name=str(a))]
                elif len(line) == 4:
                    a += 1
                    rv += [Atom(element=line[0], coords=line[1:], name=str(a))]
                line = f.readline()
                n += 1
            return rv, n

        def get_params(f, n):
            rv = []
            self.skip_lines(f, 2)
            n += 3
            line = f.readline()
            if "Definition" in line:
                definition = True
            else:
                definition = False
            self.skip_lines(f, 1)
            n += 2
            line = f.readline()
            while "--" not in line:
                line = line.split()
                param = line[1]
                if definition:
                    val = float(line[3])
                else:
                    val = float(line[2])
                rv.append((param, val))
                line = f.readline()
                n += 1
            return rv, n

        def get_modredundant(f, n):
            """read constraints for modredundant section"""
            rv = {}
            line = f.readline()
            n += 1
            while line.strip():
                atom_match = re.search("X\s+(\d+)\s+F", line)
                bond_match = re.search("B\s+(\d+)\s+(\d+)\s+F", line)
                angle_match = re.search("A\s+(\d+)\s+(\d+)\s+(\d+)\s+F", line)
                torsion_match = re.search(
                    "D\s+(\d+)\s+(\d+)\s+(\d+)\s+(\d+)\s+F", line
                )
                if atom_match:
                    if "atoms" not in rv:
                        rv["atoms"] = ""
                    else:
                        rv["atoms"] += ","
                    rv["atoms"] += atom_match.group(1)
                elif bond_match:
                    if "bonds" not in rv:
                        rv["bonds"] = []
                    rv["bonds"].append(
                        ",".join([bond_match.group(1), bond_match.group(2)])
                    )
                elif angle_match:
                    if "angles" not in rv:
                        rv["angles"] = []
                    rv["angles"].append(
                        ",".join(
                            [
                                angle_match.group(1),
                                angle_match.group(2),
                                angle_match.group(3),
                            ]
                        )
                    )
                elif torsion_match:
                    if "torsions" not in rv:
                        rv["torsions"] = []
                    rv["torsions"].append(
                        ",".join(
                            [
                                torsion_match.group(1),
                                torsion_match.group(2),
                                torsion_match.group(3),
                                torsion_match.group(4),
                            ]
                        )
                    )

                line = f.readline()
                n += 1

            return rv, n

        self.all_geom = []
        line = f.readline()
        self.other["archive"] = ""
        constraints = {}
        self.other["opt_steps"] = 0
        found_archive = False
        n = 1
        route = None
        while line != "":
            # route
            # we need to grab the route b/c sometimes 'hpmodes' can get split onto multiple lines:
            # B3LYP/genecp EmpiricalDispersion=GD3 int=(grid=superfinegrid) freq=(h
            # pmodes,noraman,temperature=313.15)
            if line.strip().startswith("#") and route is None:
                route = ""
                while "------" not in line:
                    route += line.strip() + " "
                    n += 1
                    line = f.readline()
            # archive entry
            if line.strip().startswith("1\\1\\"):
                found_archive = True
                line = "@" + line.strip()[4:]
            if found_archive and line.strip().endswith("@"):
                self.other["archive"] = self.other["archive"][:-2] + "\\\\"
                found_archive = False
            elif found_archive:
                self.other["archive"] += line.strip()

            # input atom specs and charge/mult
            if "Symbolic Z-matrix:" in line:
                self.atoms, n = get_input(f, n)

            #Pseudopotential info
            if "Pseudopotential Parameters" in line:
                self.other["ECP"] = []
                self.skip_lines(f, 4)
                n += 5
                line = f.readline()
                while "=====" not in line:
                    line = line.split()
                    if line[0].isdigit() and line[1].isdigit():
                        ele = line[1]
                        n += 1
                        line = f.readline().split()
                        if line[0] != "No":
                            self.other["ECP"].append(ELEMENTS[int(ele)])
                    n += 1
                    line = f.readline()

            # geometry
            if re.search("(Standard|Input) orientation:", line):
                if get_all and self.atoms:
                    self.all_geom += [
                        (deepcopy(self.atoms), deepcopy(self.other))
                    ]
                self.atoms, n = get_atoms(f, n)
                self.other["opt_steps"] += 1

            if re.search(
                "The following ModRedundant input section has been read:", line
            ):
                constraints, n = get_modredundant(f, n)

            if just_geom:
                line = f.readline()
                n += 1
                continue
                # z-matrix parameters
            if re.search("Optimized Parameters", line):
                self.other["params"], n = get_params(f, n)

            # status
            if NORM_FINISH in line:
                self.other["finished"] = True
            # read energies from different methods
            if "SCF Done" in line:
                tmp = [word.strip() for word in line.split()]
                idx = tmp.index("=")
                self.other["energy"] = float(tmp[idx + 1])
                self.other["scf_energy"] = float(tmp[idx + 1])

            else:
                nrg_match = re.search("\s+(E\(\S+\))\s*=\s*(\S+)", line)
                # ^ matches many methods
                # will also match the SCF line (hence the else here)
                # the match in the SCF line could be confusing b/c
                # the SCF line could be
                # SCF Done:  E(RB2PLYPD3) =  -76.2887108570     A.U. after   10 cycles
                # and later on, there will be a line...
                #  E2(B2PLYPD3) =    -0.6465105880D-01 E(B2PLYPD3) =    -0.76353361915801D+02
                # this will give:
                # * E(RB2PLYPD3) = -76.2887108570 
                # * E(B2PLYPD3) = -76.353361915801
                # very similar names for very different energies...
                if nrg_match:
                    self.other["energy"] = float(nrg_match.group(2).replace("D", "E"))
                    self.other[nrg_match.group(1)] = self.other["energy"]
            
            # CC energy
            if line.startswith(" CCSD(T)= "):
                self.other["energy"] = float(line.split()[-1].replace("D", "E"))
                self.other["E(CCSD(T))"] = self.other["energy"]
            
            # MP energies
            mp_match = re.search("([RU]MP\d+(?:\(\S+\))?)\s*=\s*(\S+)", line)
            if mp_match:
                self.other["energy"] = float(mp_match.group(2).replace("D", "E"))
                self.other["E(%s)" % mp_match.group(1)] = self.other["energy"]

            if "Molecular mass:" in line:
                self.other["mass"] = float(float_num.search(line).group(0))
                self.other["mass"] *= UNIT.AMU_TO_KG

            # Frequencies
            if route is not None and "hpmodes" in route.lower():
                self.other["hpmodes"] = True
            if "Harmonic frequencies" in line:
                freq_str = line
                line = f.readline()
                while line.strip():
                    n += 1
                    freq_str += line
                    line = f.readline()
                if "hpmodes" not in self.other:
                    self.other["hpmodes"] = False
                self.other["frequency"] = Frequency(
                    freq_str, hpmodes=self.other["hpmodes"]
                )

            if "Anharmonic Infrared Spectroscopy" in line:
                self.skip_lines(f, 5)
                n += 5
                anharm_str = ""
                combinations_read = False
                combinations = False
                line = f.readline()
                while not combinations_read:
                    n += 1
                    anharm_str += line
                    if "Combination Bands" in line:
                        combinations = True
                    line = f.readline()
                    if combinations and line == "\n":
                        combinations_read = True

                self.other["frequency"].parse_gaussian_lines(
                    anharm_str.splitlines(), harmonic=False,
                )

            # X matrix for anharmonic
            if "Total Anharmonic X Matrix" in line:
                self.skip_lines(f, 1)
                n += 1
                n_freq = len(self.other["frequency"].data)
                n_sections = int(np.ceil(n_freq / 5))
                x_matrix = np.zeros((n_freq, n_freq))
                for section in range(0, n_sections):
                    header = f.readline()
                    n += 1
                    for j in range(5 * section, n_freq):
                        line = f.readline()
                        n += 1
                        ll = 5 * section
                        ul = 5 * section + min(j - ll + 1, 5)
                        x_matrix[j, ll:ul] = [
                            float(x.replace("D", "e"))
                            for x in line.split()[1:]
                        ]
                x_matrix += np.tril(x_matrix, k=-1).T
                self.other["X_matrix"] = x_matrix

            if "Total X0" in line:
                self.other["X0"] = float(line.split()[5])

            # TD-DFT output
            if line.strip().startswith("Ground to excited state"):
                uv_vis = ""
                highest_state = 0
                done = False
                read_states = False
                while not done:
                    n += 1
                    uv_vis += line
                    if not read_states and line.strip() and line.split()[0].isdigit():
                        state = int(line.split()[0])
                        if state > highest_state:
                            highest_state = state
                    if line.strip().startswith("Ground to excited state transition velocity"):
                        read_states = True
                    if re.search("Excited State\s*%i:" % highest_state, line):
                        done = True
                    if line.strip().startswith("Total Energy, E"):
                        nrg = re.search(
                            r"Total Energy, E\((\S+)\)\s*=\s*(-?\d+\.\d+)", line
                        )
                        self.other["E(%s)" % nrg.group(1)] = float(nrg.group(2))
                        self.other["energy"] = float(nrg.group(2))

                    line = f.readline()
                self.other["uv_vis"] = ValenceExcitations(
                    uv_vis, style="gaussian"
                )

            # Thermo
            if re.search("Temperature\s*\d+\.\d+", line):
                self.other["temperature"] = float(
                    float_num.search(line).group(0)
                )
            if "Rotational constants (GHZ):" in line:
                rot = float_num.findall(line)
                rot = [
                    float(r) * PHYSICAL.PLANCK * (10 ** 9) / PHYSICAL.KB
                    for r in rot
                ]
                self.other["rotational_temperature"] = rot

            # rotational constants from anharmonic frequency jobs
            if "Rotational Constants (in MHz)" in line:
                self.skip_lines(f, 2)
                n += 2
                equilibrium_rotational_temperature = np.zeros(3)
                ground_rotational_temperature = np.zeros(3)
                centr_rotational_temperature = np.zeros(3)
                for i in range(0, 3):
                    line = f.readline()
                    n += 1
                    info = line.split()
                    Be = float(info[1])
                    B00 = float(info[3])
                    B0 = float(info[5])
                    equilibrium_rotational_temperature[i] = Be
                    ground_rotational_temperature[i] = B00
                    centr_rotational_temperature[i] = B0
                equilibrium_rotational_temperature *= (
                    PHYSICAL.PLANCK * 1e6 / PHYSICAL.KB
                )
                ground_rotational_temperature *= (
                    PHYSICAL.PLANCK * 1e6 / PHYSICAL.KB
                )
                centr_rotational_temperature *= (
                    PHYSICAL.PLANCK * 1e6 / PHYSICAL.KB
                )
                self.other[
                    "equilibrium_rotational_temperature"
                ] = equilibrium_rotational_temperature
                self.other[
                    "ground_rotational_temperature"
                ] = ground_rotational_temperature
                self.other[
                    "centr_rotational_temperature"
                ] = centr_rotational_temperature

            if "Sum of electronic and zero-point Energies=" in line:
                self.other["E_ZPVE"] = float(float_num.search(line).group(0))
            if "Sum of electronic and thermal Enthalpies=" in line:
                self.other["enthalpy"] = float(float_num.search(line).group(0))
            if "Sum of electronic and thermal Free Energies=" in line:
                self.other["free_energy"] = float(
                    float_num.search(line).group(0)
                )
            if "Zero-point correction=" in line:
                self.other["ZPVE"] = float(float_num.search(line).group(0))
            if "Rotational symmetry number" in line:
                self.other["rotational_symmetry_number"] = int(
                    re.search("\d+", line).group(0)
                )

            # Gradient
            if re.search("Threshold\s+Converged", line) is not None:
                line = f.readline()
                n += 1
                grad = {}

                def add_grad(line, name, grad):
                    line = line.split()
                    grad[name] = {
                        "value": line[2],
                        "threshold": line[3],
                        "converged": True if line[4] == "YES" else False,
                    }
                    return grad

                while line != "":
                    if "Predicted change in Energy" in line:
                        break
                    if re.search("Maximum\s+Force", line) is not None:
                        grad = add_grad(line, "Max Force", grad)
                    if re.search("RMS\s+Force", line) is not None:
                        grad = add_grad(line, "RMS Force", grad)
                    if re.search("Maximum\s+Displacement", line) is not None:
                        grad = add_grad(line, "Max Disp", grad)
                    if re.search("RMS\s+Displacement", line) is not None:
                        grad = add_grad(line, "RMS Disp", grad)
                    line = f.readline()
                    n += 1
                self.other["gradient"] = grad

            # electronic properties
            if "Electrostatic Properties (Atomic Units)" in line:
                self.skip_lines(f, 5)
                n += 5
                self.other["electric_potential"] = []
                self.other["electric_field"] = []
                line = f.readline()
                while "--" not in line:
                    info = line.split()
                    self.other["electric_potential"].append(float(info[2]))
                    self.other["electric_field"].append([float(x) for x in info[3:]])
                    line = f.readline()
                    n += 1
                self.other["electric_potential"] = np.array(self.other["electric_potential"])
                self.other["electric_field"] = np.array(self.other["electric_field"])

            # optical features
            if "[Alpha]" in line:
                alpha_match = re.search("\[Alpha\].*\(\s*(.*\s?.*)\)\s*=\s*(-?\d+\.\d+)", line)
                self.other["optical_rotation_(%s)" % alpha_match.group(1)] = \
                float(alpha_match.group(2))

            # symmetry
            if "Full point group" in line:
                self.other["full_point_group"] = line.split()[-3]

            if "Largest Abelian subgroup" in line:
                self.other["abelian_subgroup"] = line.split()[-3]

            if "Largest concise Abelian subgroup" in line:
                self.other["concise_abelian_subgroup"] = line.split()[-3]

            # forces
            if "Forces (Hartrees/Bohr)" in line:
                gradient = np.zeros((len(self.atoms), 3))
                self.skip_lines(f, 2)
                n += 2
                for i in range(0, len(self.atoms)):
                    n += 1
                    line = f.readline()
                    info = line.split()
                    gradient[i] = np.array([float(x) for x in info[2:]])

                self.other["forces"] = gradient

            # nbo stuff
            if "N A T U R A L   A T O M I C   O R B I T A L   A N D" in line:
                self.read_nbo(f)

            # atomic charges
            charge_match = re.search("(\S+) charges:\s*$", line)
            if charge_match:
                self.skip_lines(f, 1)
                n += 1
                charges = []
                for i in range(0, len(self.atoms)):
                    line = f.readline()
                    n += 1
                    charges.append(float(line.split()[2]))
                    self.atoms[i].charge = float(line.split()[2])
                self.other[charge_match.group(1) + " Charges"] = charges

            # capture errors
            # only keep first error, want to fix one at a time
            if "error" not in self.other:
                for err in ERROR:
                    if re.search(err, line):
                        self.other["error"] = ERROR[err]
                        self.other["error_msg"] = line.strip()
                        break

            line = f.readline()
            n += 1

        if not just_geom:
            if route is not None:
                other_kwargs = {GAUSSIAN_ROUTE: {}}
                route_spec = re.compile("(\w+)=?\((.*)\)")
                method_and_basis = re.search(
                    "#(?:[NnPpTt]\s+?)(\S+)|#\s*?(\S+)", route
                )
                if method_and_basis is not None:
                    if method_and_basis.group(2):
                        method_info = method_and_basis.group(2).split("/")
                    else:
                        method_info = method_and_basis.group(1).split("/")

                    method = method_info[0]
                    if len(method_info) > 1:
                        basis = method_info[1]
                    else:
                        basis = None

                    route_options = route.split()
                    job_type = []
                    grid = None
                    solvent = None
                    for option in route_options:
                        if option.startswith("#"):
                            continue
                        elif option.startswith(method):
                            continue

                        option_lower = option.lower()
                        if option_lower.startswith("opt"):
                            ts = False
                            match = route_spec.search(option)
                            if match:
                                options = match.group(2).split(",")
                            elif option_lower.startswith("opt="):
                                options = ["".join(option.split("=")[1:])]
                            else:
                                if not constraints:
                                    # if we didn't read constraints, try using flagged atoms instead
                                    from AaronTools.finders import FlaggedAtoms

                                    constraints = {"atoms": FlaggedAtoms}
                                    if not any(atom.flag for atom in self.atoms):
                                        constraints = None
                                job_type.append(
                                    OptimizationJob(constraints=constraints)
                                )
                                continue

                            other_kwargs[GAUSSIAN_ROUTE]["opt"] = []

                            for opt in options:
                                if opt.lower() == "ts":
                                    ts = True
                                else:
                                    other_kwargs[GAUSSIAN_ROUTE]["opt"].append(
                                        opt
                                    )

                            job_type.append(
                                OptimizationJob(
                                    transition_state=ts,
                                    constraints=constraints,
                                )
                            )

                        elif option_lower.startswith("freq"):
                            temp = 298.15
                            match = route_spec.search(option)
                            if match:
                                options = match.group(2).split(",")
                            elif option_lower.startswith("freq="):
                                options = "".join(option.split("=")[1:])
                            else:
                                job_type.append(FrequencyJob())
                                continue

                            other_kwargs[GAUSSIAN_ROUTE]["freq"] = []

                            for opt in options:
                                if opt.lower().startswith("temp"):
                                    temp = float(opt.split("=")[1])
                                else:
                                    other_kwargs[GAUSSIAN_ROUTE][
                                        "freq"
                                    ].append(opt)

                            job_type.append(FrequencyJob(temperature=temp))

                        elif option_lower == "sp":
                            job_type.append(SinglePointJob())

                        elif option_lower.startswith("int"):
                            match = route_spec.search(option)
                            if match:
                                options = match.group(2).split(",")
                            elif option_lower.startswith("freq="):
                                options = "".join(option.split("=")[1:])
                            else:
                                job_type.append(FrequencyJob())
                                continue

                            for opt in options:
                                if opt.lower().startswith("grid"):
                                    grid_name = opt.split("=")[1]
                                    grid = IntegrationGrid(grid_name)
                                else:
                                    if (
                                        "Integral"
                                        not in other_kwargs[GAUSSIAN_ROUTE]
                                    ):
                                        other_kwargs[GAUSSIAN_ROUTE][
                                            "Integral"
                                        ] = []
                                    other_kwargs[GAUSSIAN_ROUTE][
                                        "Integral"
                                    ].append(opt)

                        else:
                            # TODO: parse solvent
                            match = route_spec.search(option)
                            if match:
                                keyword = match.group(1)
                                options = match.group(2).split(",")
                                other_kwargs[GAUSSIAN_ROUTE][keyword] = options
                            elif "=" in option:
                                keyword = option.split("=")[0]
                                options = "".join(option.split("=")[1:])
                                other_kwargs[GAUSSIAN_ROUTE][keyword] = [
                                    options
                                ]
                            else:
                                other_kwargs[GAUSSIAN_ROUTE][option] = []
                                continue

                    self.other["other_kwargs"] = other_kwargs
                    try:
                        theory = Theory(
                            charge=self.other["charge"],
                            multiplicity=self.other["multiplicity"],
                            job_type=job_type,
                            basis=basis,
                            method=method,
                            grid=grid,
                            solvent=solvent,
                        )
                        theory.kwargs = self.other["other_kwargs"]
                        self.other["theory"] = theory
                    except KeyError:
                        # if there is a serious error, too little info may be available
                        # to properly create the theory object
                        pass

        for i, a in enumerate(self.atoms):
            a.name = str(i + 1)

        if "finished" not in self.other:
            self.other["finished"] = False
        if "error" not in self.other:
            self.other["error"] = None
        return

    def read_com(self, f):
        found_atoms = False
        found_constraint = False
        atoms = []
        other = {}
        for line in f:
            # header
            if line.startswith("%"):
                continue
            if line.startswith("#"):
                method = re.search("^#([NnPpTt]\s+?)(\S+)|^#\s*?(\S+)", line)
                # route can be #n functional/basis ...
                # or #functional/basis ...
                # or # functional/basis ...
                if method.group(3):
                    other["method"] = method.group(3)
                else:
                    other["method"] = method.group(2)
                if "temperature=" in line:
                    other["temperature"] = float(
                        re.search("temperature=(\d+\.?\d*)", line).group(1)
                    )
                if "solvent=" in line:
                    other["solvent"] = re.search(
                        "solvent=(\S+)\)", line
                    ).group(1)
                if "scrf=" in line:
                    # solvent model should be non-greedy b/c solvent name can have commas
                    other["solvent_model"] = re.search(
                        "scrf=\((\S+?),", line
                    ).group(1)
                if "EmpiricalDispersion=" in line:
                    other["emp_dispersion"] = re.search(
                        "EmpiricalDispersion=(\S+)", line
                    ).group(1)
                if "int=(grid" in line or "integral=(grid" in line.lower():
                    other["grid"] = re.search(
                        "(?:int||Integral)=\(grid[(=](\S+?)\)", line
                    ).group(1)
                # comments can be multiple lines long
                # but there should be a blank line between the route and the comment
                # and another between the comment and the charge+mult
                blank_lines = 0
                while blank_lines < 2:
                    line = f.readline().strip()
                    if len(line) == 0:
                        blank_lines += 1
                    else:
                        if "comment" not in other:
                            other["comment"] = ""
                        other["comment"] += "%s\n" % line
                other["comment"] = (
                    other["comment"].strip() if "comment" in other else ""
                )
                line = f.readline()
                if len(line.split()) > 1:
                    line = line.split()
                else:
                    line = line.split(",")
                other["charge"] = int(line[0])
                other["multiplicity"] = int(line[1])
                found_atoms = True
                continue
            # constraints
            if found_atoms and line.startswith("B") and line.endswith("F"):
                found_constraint = True
                if "constraint" not in other:
                    other["constraint"] = []
                other["constraint"] += [float_num.findall(line)]
                continue
            # footer
            if found_constraint:
                if "footer" not in other:
                    other["footer"] = ""
                other["footer"] += line
                continue
            # atom coords
            nums = float_num.findall(line)
            line = line.split()
            if len(line) == 5 and is_alpha(line[0]) and len(nums) == 4:
                if not is_int(line[1]):
                    continue
                a = Atom(element=line[0], coords=nums[1:], flag=nums[0])
                atoms += [a]
            elif len(line) == 4 and is_alpha(line[0]) and len(nums) == 3:
                a = Atom(element=line[0], coords=nums)
                atoms += [a]
            else:
                continue
        for i, a in enumerate(atoms):
            a.name = str(i + 1)
        self.atoms = atoms
        self.other = other
        return

    def read_fchk(self, f, just_geom=True, max_length=10000000):
        def parse_to_list(
            i, lines, length, data_type, debug=False, max_length=max_length,
        ):
            """takes a block in an fchk file and turns it into an array
            block headers all end with N=   <int>
            the length of the array will be <int>
            the data type is specified by data_type"""
            i += 1
            line = f.readline()
            # print("first line", line)
            items_per_line = len(line.split())
            # print("items per line", items_per_line)
            total_items = items_per_line
            num_lines = ceil(length / items_per_line)

            # print("lines in block", num_lines)

            block = [line]
            for k in range(0, num_lines - 1):
                line = f.readline()
                if max_length < length:
                    continue
                block.append(line)
            
            if max_length < length:
                return length, i + num_lines
            block = " ".join(block)

            if debug:
                print("full block")
                print(block)

            return (
                np.fromstring(block, count=length, dtype=data_type, sep=" "),
                i + num_lines,
            )

        self.atoms = []
        atom_numbers = []
        atom_coords = []

        other = {}

        int_info = re.compile("([\S\s]+?)\s*I\s*(N=)?\s*(-?\d+)")
        real_info = re.compile(
            "([\S\s]+?)\s*R\s*(N=)\s*(-?\d+\.?\d*[Ee]?[+-]?\d*)"
        )
        char_info = re.compile(
            "([\S\s]+?)\s*C\s*(N=)?\s*(-?\d+\.?\d*[Ee]?[+-]?\d*)"
        )

        theory = Theory()

        line = f.readline()

        i = 0
        while line != "":
            if i == 0:
                other["comment"] = line.strip()
            elif i == 1:
                i += 1
                line = f.readline()
                job_info = line.split()
                if job_info[0] == "SP":
                    theory.job_type = [SinglePointJob()]
                elif job_info[0] == "FOPT":
                    theory.job_type[OptimizationJob()]
                elif job_info[0] == "FTS":
                    theory.job_type = [OptimizationJob(transition_state=True)]
                elif job_info[0] == "FORCE":
                    theory.job_type = [ForceJob()]
                elif job_info[0] == "FREQ":
                    theory.job_type = [FrequencyJob()]

                theory.method = job_info[1]
                if len(job_info) > 2:
                    theory.basis = job_info[2]

                i += 1
                line = f.readline()
                continue

            int_match = int_info.match(line)
            real_match = real_info.match(line)
            char_match = char_info.match(line)
            if int_match is not None:
                data = int_match.group(1)
                # print("int", data)
                value = int_match.group(3)
                if data == "Charge" and not just_geom:
                    theory.charge = int(value)
                elif data == "Multiplicity" and not just_geom:
                    theory.multiplicity = int(value)
                elif data == "Atomic numbers":
                    atom_numbers, i = parse_to_list(i, f, int(value), int)
                elif not just_geom:
                    if int_match.group(2):
                        other[data], i = parse_to_list(
                            i, f, int(value), int
                        )
                    else:
                        other[data] = int(value)

            elif real_match is not None:
                data = real_match.group(1)
                # print("real", data)
                value = real_match.group(3)
                if data == "Current cartesian coordinates":
                    atom_coords, i = parse_to_list(i, f, int(value), float)
                elif data == "Total Energy":
                    other["energy"] = float(value)
                elif not just_geom:
                    if real_match.group(2):
                        other[data], i = parse_to_list(
                            i, f, int(value), float
                        )
                    else:
                        other[data] = float(value)

            # elif char_match is not None:
            #     data = char_match.group(1)
            #     value = char_match.group(3)
            #     if not just_geom:
            #         other[data] = lines[i + 1]
            #         i += 1

            line = f.readline()
            i += 1

        self.other = other
        self.other["theory"] = theory

        if isinstance(atom_coords, int):
            raise RuntimeError(
                "max. array size is insufficient to parse atom data\n"
                "must be at least %i" % atom_coords
            )
        coords = np.reshape(atom_coords, (len(atom_numbers), 3))
        for n, (atnum, coord) in enumerate(zip(atom_numbers, coords)):
            atom = Atom(
                element=ELEMENTS[atnum],
                coords=UNIT.A0_TO_BOHR * coord,
                name=str(n + 1),
            )
            self.atoms.append(atom)

        try:
            self.other["orbitals"] = Orbitals(self)
        except (NotImplementedError, KeyError):
            pass
        except (TypeError, ValueError) as err:
            self.LOG.warning(
                "could not create Orbitals, try increasing the max.\n"
                "array size to read from FCHK files\n\n"
                "%s" % err
            )
            for key in [
                "Alpha MO coefficients", "Beta MO coefficients",
                "Shell types", "Shell to atom map", "Contraction coefficients",
                "Primitive exponents", "Number of primitives per shell",
                "Coordinates of each shell",
            ]:
                if key in self.other and isinstance(self.other[key], int):
                    self.LOG.warning(
                        "size of %s is > %i: %i" % (key, max_length, self.other[key])
                    )

    def read_nbo(self, f):
        """
        read nbo data
        """
        line = f.readline()
        while line:
            if "natural bond orbitals (summary):" in line.lower():
                break

            if "NATURAL POPULATIONS:" in line:
                self.skip_lines(f, 3)
                ao_types = []
                ao_atom_ndx = []
                nao_types = []
                occ = []
                nrg = []
                blank_lines = 0
                while blank_lines <= 1:
                    match = re.search(
                        "\d+\s+[A-Z][a-z]?\s+(\d+)\s+(\S+)\s+([\S\s]+?)(-?\d+\.\d+)\s+(-?\d+\.\d+)",
                        line
                    )
                    if match:
                        ao_atom_ndx.append(int(match.group(1)) - 1)
                        ao_types.append(match.group(2))
                        nao_types.append(match.group(3))
                        occ.append(float(match.group(4)))
                        nrg.append(float(match.group(5)))
                        blank_lines = 0
                    else:
                        blank_lines += 1
                    line = f.readline()
                self.other["ao_types"] = ao_types
                self.other["ao_atom_ndx"] = ao_atom_ndx
                self.other["nao_type"] = nao_types
                self.other["ao_occ"] = occ
                self.other["ao_nrg"] = nrg

            if "Summary of Natural Population Analysis:" in line:
                self.skip_lines(f, 5)
                core_occ = []
                val_occ = []
                rydberg_occ = []
                nat_q = []
                line = f.readline()
                while "==" not in line:
                    info = line.split()
                    core_occ.append(float(info[3]))
                    val_occ.append(float(info[4]))
                    rydberg_occ.append(float(info[5]))
                    nat_q.append(float(info[2]))
                    line = f.readline()
                self.other["Natural Charges"] = nat_q
                self.other["core_occ"] = core_occ
                self.other["valence_occ"] = val_occ
                self.other["rydberg_occ"] = rydberg_occ

            if "Wiberg bond index matrix in the NAO basis" in line:
                dim = len(self.other["Natural Charges"])
                bond_orders = np.zeros((dim, dim))
                done = False
                j = 0
                for block in range(0, ceil(dim / 9)):
                    offset = 9 * j
                    self.skip_lines(f, 3)
                    for i in range(0, dim):
                        line = f.readline()
                        for k, bo in enumerate(line.split()[2:]):
                            bo = float(bo)
                            bond_orders[i][offset + k] = bo
                    j += 1
                self.other["wiberg_nao"] = bond_orders

            line = f.readline()

    def read_crest(self, f, conf_name=None):
        """
        conf_name = False to skip conformer loading (doesn't get written until crest job is done)
        """
        if conf_name is None:
            conf_name = os.path.join(
                os.path.dirname(self.name), "crest_conformers.xyz"
            )
        line = True
        self.other["finished"] = False
        self.other["error"] = None
        while line:
            line = f.readline()
            if "terminated normally" in line:
                self.other["finished"] = True
            elif "population of lowest" in line:
                self.other["best_pop"] = float(float_num.findall(line)[0])
            elif "ensemble free energy" in line:
                self.other["free_energy"] = (
                    float(float_num.findall(line)[0]) / UNIT.HART_TO_KCAL
                )
            elif "ensemble entropy" in line:
                self.other["entropy"] = (
                    float(float_num.findall(line)[1]) / UNIT.HART_TO_KCAL
                )
            elif "ensemble average energy" in line:
                self.other["avg_energy"] = (
                    float(float_num.findall(line)[0]) / UNIT.HART_TO_KCAL
                )
            elif "E lowest" in line:
                self.other["energy"] = float(float_num.findall(line)[0])
            elif "T /K" in line:
                self.other["temperature"] = float(float_num.findall(line)[0])
            elif (
                line.strip()
                .lower()
                .startswith(("forrtl", "warning", "*warning"))
            ):
                self.other["error"] = "UNKNOWN"
                if "error_msg" not in self.other:
                    self.other["error_msg"] = ""
                self.other["error_msg"] += line
            elif "-chrg" in line:
                self.other["charge"] = int(float_num.findall(line)[0])
            elif "-uhf" in line:
                self.other["multiplicity"] = (
                    int(float_num.findall(line)[0]) + 1
                )

        if self.other["finished"] and conf_name:
            self.other["conformers"] = FileReader(
                conf_name,
                get_all=True,
            ).all_geom
            self.comment, self.atoms = self.other["conformers"][0]
            self.other["conformers"] = self.other["conformers"][1:]

    def read_xtb(self, f, freq_name=None):
        line = True
        self.other["finished"] = False
        self.other["error"] = None
        self.atoms = []
        self.comment = ""
        while line:
            line = f.readline()
            if "Optimized Geometry" in line:
                line = f.readline()
                n_atoms = int(line.strip())
                line = f.readline()
                self.comment = " ".join(line.strip().split()[2:])
                for i in range(n_atoms):
                    line = f.readline()
                    elem, x, y, z = line.split()
                    self.atoms.append(Atom(element=elem, coords=[x, y, z]))
            if "normal termination" in line:
                self.other["finished"] = True
            if "abnormal termination" in line:
                self.other["error"] = "UNKNOWN"
            if line.strip().startswith("#ERROR"):
                if "error_msg" not in self.other:
                    self.other["error_msg"] = ""
                self.other["error_msg"] += line
            if "charge" in line and ":" in line:
                self.other["charge"] = int(float_num.findall(line)[0])
            if "spin" in line and ":" in line:
                self.other["multiplicity"] = (
                    2 * float(float_num.findall(line)[0]) + 1
                )
            if "total energy" in line:
                self.other["energy"] = (
                    float(float_num.findall(line)[0]) * UNIT.HART_TO_KCAL
                )
            if "zero point energy" in line:
                self.other["ZPVE"] = (
                    float(float_num.findall(line)[0]) * UNIT.HART_TO_KCAL
                )
            if "total free energy" in line:
                self.other["free_energy"] = (
                    float(float_num.findall(line)[0]) * UNIT.HART_TO_KCAL
                )
            if "electronic temp." in line:
                self.other["temperature"] = float(float_num.findall(line)[0])
        if freq_name is not None:
            with open(freq_name) as f_freq:
                self.other["frequency"] = Frequency(f_freq.read())

    def read_sqm(self, f):
        lines = f.readlines()

        self.other["finished"] = False

        self.atoms = []
        i = 0
        while i < len(lines):
            line = lines[i]
            if "Atomic Charges for Step" in line:
                elements = []
                for info in lines[i + 2 :]:
                    if not info.strip() or not info.split()[0].isdigit():
                        break
                    ele = info.split()[1]
                    elements.append(ele)
                i += len(elements) + 2

            if "Final Structure" in line:
                k = 0
                for info in lines[i + 4 :]:
                    data = info.split()
                    coords = np.array([x for x in data[4:7]])
                    self.atoms.append(
                        Atom(
                            name=str(k + 1),
                            coords=coords,
                            element=elements[k],
                        )
                    )
                    k += 1
                    if k == len(elements):
                        break
                i += k + 4

            if "Calculation Completed" in line:
                self.other["finished"] = True

            if "Total SCF energy" in line:
                self.other["energy"] = (
                    float(line.split()[4]) / UNIT.HART_TO_KCAL
                )

            i += 1

        if not self.atoms:
            # there's no atoms if there's an error
            # error is probably on the last line
            self.other["error"] = "UNKNOWN"
            self.other["error_msg"] = line

    def read_nbo_47(self, f, nbo_name=None):
        lines = f.readlines()
        bohr = False
        i = 0
        while i < len(lines):
            line = lines[i]
            if line.startswith(" $"):
                section = line.split()[0]
                if section.startswith("$COORD"):
                    i += 1
                    self.atoms = []
                    line = lines[i]
                    while not line.startswith(" $END"):
                        if re.search("\d+\s+\d+(?:\s+-?\d+\.\d+\s){3}", line):
                            info = line.split()
                            ndx = int(info[0])
                            coords = [float(x) for x in info[2:5]]
                            self.atoms.append(
                                Atom(
                                    element=ELEMENTS[ndx],
                                    name=str(len(self.atoms) + 1),
                                    coords=np.array(coords),
                                )
                            )
                        i += 1
                        line = lines[i]
                elif section.startswith("$BASIS"):
                    reading_centers = False
                    reading_labels = False
                    i += 1
                    line = lines[i]
                    while not line.startswith(" $END"):
                        if "CENTER" in line.upper():
                            self.other["shell_to_atom"] = [
                                int(x) for x in line.split()[2:]
                            ]
                            reading_centers = True
                            reading_labels = False
                        elif "LABEL" in line.upper():
                            self.other["momentum_label"] = [
                                int(x) for x in line.split()[2:]
                            ]
                            reading_labels = True
                            reading_centers = False
                        elif reading_centers:
                            self.other["shell_to_atom"].extend(
                                [int(x) for x in line.split()]
                            )
                        elif reading_labels:
                            self.other["momentum_label"].extend(
                                [int(x) for x in line.split()]
                            )
                        i += 1
                        line = lines[i]
                elif section.startswith("$CONTRACT"):
                    int_sections = {
                        "NCOMP": "funcs_per_shell",
                        "NPRIM": "n_prim_per_shell",
                        "NPTR": "start_ndx",
                    }
                    float_sections = {
                        "EXP": "exponents",
                        "CS": "s_coeff",
                        "CP": "p_coeff",
                        "CD": "d_coeff",
                        "CF": "f_coeff",
                    }
                    i += 1
                    line = lines[i]
                    while not line.startswith(" $END"):
                        if any(line.strip().startswith(section) for section in int_sections):
                            section = line.split()[0]
                            self.other[int_sections[section]] = [
                                int(x) for x in line.split()[2:]
                            ]
                            i += 1
                            line = lines[i]
                            while "=" not in line and "$" not in line:
                                self.other[int_sections[section]].extend([
                                    int(x) for x in line.split()
                                ])
                                i += 1
                                line = lines[i]
                        elif any(line.strip().startswith(section) for section in float_sections):
                            section = line.split()[0]
                            self.other[float_sections[section]] = [
                                float(x) for x in line.split()[2:]
                            ]
                            i += 1
                            line = lines[i]
                            while "=" not in line and "$" not in line:
                                self.other[float_sections[section]].extend([
                                    float(x) for x in line.split()
                                ])
                                i += 1
                                line = lines[i]
                        else:
                            i += 1
                            line = lines[i]
                            
                elif section.startswith("$GENNBO"):
                    if "BOHR" in section.upper():
                        bohr = True
                    nbas = re.search("NBAS=(\d+)", line)
                    n_funcs = int(nbas.group(1))
                    if "CUBICF" in section.upper():
                        self.LOG.warning("cubic F shell will not be handled correctly")
            i += 1
        
        if nbo_name is not None:
            self._read_nbo_coeffs(nbo_name)
    
    def _read_nbo_coeffs(self, nbo_name):
        """
        read coefficients in AO basis for NBO's/NLHO's/NAO's/etc.
        called by methods that read NBO input (.47) or output files (.31)
        """
        with open(nbo_name, "r") as f2:
            lines = f2.readlines()
        kind = re.search("P?(\S+)s", lines[1]).group(1)
        desc_file = os.path.splitext(nbo_name)[0] + ".46"
        if os.path.exists(desc_file):
            with open(desc_file, "r") as f3:
                desc_lines = f3.readlines()
                for k, line in enumerate(desc_lines):
                    if kind in line:
                        self.other["orbit_kinds"] = []
                        n_orbits = int(line.split()[1])
                        k += 1
                        while len(self.other["orbit_kinds"]) < n_orbits:
                            self.other["orbit_kinds"].extend([
                                desc_lines[k][i: i + 10]
                                for i in range(1, len(desc_lines[k]) - 1, 10)
                            ])
                            k += 1
        else:
            self.LOG.warning(
                "no .46 file found - orbital descriptions will be unavialable"
            )
                
        j = 3
        self.other["alpha_coefficients"] = []
        while len(self.other["alpha_coefficients"]) < sum(self.other["funcs_per_shell"]):
            mo_coeff = []
            while len(mo_coeff) < sum(self.other["funcs_per_shell"]):
                mo_coeff.extend([float(x) for x in lines[j].split()])
                j += 1
            self.other["alpha_coefficients"].append(mo_coeff)
        self.other["orbitals"] = Orbitals(self)

    def read_nbo_31(self, f, nbo_name=None):
        lines = f.readlines()
        comment = lines[0].strip()
        info = lines[3].split()
        n_atoms = int(info[0])
        self.atoms = []
        for i in range(5, 5 + n_atoms):
            atom_info = lines[i].split()
            ele = ELEMENTS[int(atom_info[0])]
            coords = np.array([float(x) for x in atom_info[1:4]])
            self.atoms.append(
                Atom(
                    element=ele,
                    coords=coords,
                    name=str(i-4),
                )
            )
        
        i = n_atoms + 6
        line = lines[i]
        self.other["shell_to_atom"] = []
        self.other["momentum_label"] = []
        self.other["funcs_per_shell"] = []
        self.other["start_ndx"] = []
        self.other["n_prim_per_shell"] = []
        while "---" not in line:
            info = line.split()
            ndx = int(info[0])
            funcs = int(info[1])
            start_ndx = int(info[2])
            n_prim = int(info[3])
            self.other["shell_to_atom"].extend([ndx for j in range(0, funcs)])
            self.other["funcs_per_shell"].append(funcs)
            self.other["start_ndx"].append(start_ndx)
            self.other["n_prim_per_shell"].append(n_prim)
            i += 1
            line = lines[i]
            momentum_labels = [int(x) for x in line.split()]
            self.other["momentum_label"].extend(momentum_labels)
            i += 1
            line = lines[i]
        
        i += 1
        self.other["exponents"] = []
        line = lines[i]
        while line.strip() != "":
            exponents = [float(x) for x in line.split()]
            self.other["exponents"].extend(exponents)
            i += 1
            line = lines[i]

        i += 1
        self.other["s_coeff"] = []
        line = lines[i]
        while line.strip() != "":
            coeff = [float(x) for x in line.split()]
            self.other["s_coeff"].extend(coeff)
            i += 1
            line = lines[i]
        
        i += 1
        self.other["p_coeff"] = []
        line = lines[i]
        while line.strip() != "":
            coeff = [float(x) for x in line.split()]
            self.other["p_coeff"].extend(coeff)
            i += 1
            line = lines[i]

        i += 1
        self.other["d_coeff"] = []
        line = lines[i]
        while line.strip() != "":
            coeff = [float(x) for x in line.split()]
            self.other["d_coeff"].extend(coeff)
            i += 1
            line = lines[i]

        i += 1
        self.other["f_coeff"] = []
        line = lines[i]
        while line.strip() != "":
            coeff = [float(x) for x in line.split()]
            self.other["f_coeff"].extend(coeff)
            i += 1
            line = lines[i]

        i += 1
        self.other["g_coeff"] = []
        line = lines[i]
        while line.strip() != "":
            coeff = [float(x) for x in line.split()]
            self.other["g_coeff"].extend(coeff)
            i += 1
            line = lines[i]

        if nbo_name is not None:
            self._read_nbo_coeffs(nbo_name)