view larch_feff.py @ 1:8ee2cc3374fe draft

planemo upload for repository https://github.com/MaterialsGalaxy/larch-tools/tree/main/larch_feff commit 0f66842e802430e887d1c6cb7be1cc5436408fd2

import json
import re
import shutil
import sys
from pathlib import Path

from larch.xafs.feffrunner import feff6l
from larch.xrd import cif2feff

from pymatgen.core import Species
from pymatgen.io.cif import CifParser


def get_path_labels(paths_file):
    is_meta = True
    count = 0
    a_path = {}
    all_paths = {}
    with open(paths_file) as datfile:
        dat_lines = datfile.readlines()
        for a_line in dat_lines:
            count += 1
            if re.match("-{15}", a_line.strip()) is not None:
                is_meta = False
            elif not is_meta:
                if re.match(r"\s*\d*\s{4}\d*\s{3}", a_line) is not None:
                    if a_path:
                        all_paths[a_path["index"]] = a_path
                    line_data = a_line.split()
                    a_path = {
                        "index": line_data[0],
                        "nleg": line_data[1],
                        "degeneracy": line_data[2],
                    }
                elif (
                    re.match(r"\s{6}x\s{11}y\s{5}", a_line) is None
                ):  # ignore the intermediate headings
                    line_data = a_line.split()
                    if "label" not in a_path:
                        a_path["label"] = line_data[4].replace("'", "")
                    else:
                        a_path["label"] += "." + line_data[4].replace("'", "")
    if a_path and "index" in a_path:
        all_paths[a_path["index"]] = a_path
    return all_paths


def main(structure_file: str, file_format: dict):
    crystal_f = Path(structure_file)
    feff_dir = "feff"
    feff_inp = "feff.inp"
    path = Path(feff_dir, feff_inp)
    path.parent.mkdir(parents=True, exist_ok=True)

    if file_format["format"] == "cif":
        print(f"Parsing {crystal_f.name} and saving to {path}")

        # Parse the cif file here... but only so that we can extract the
        # chemical symbols present in the crystal
        cif_parser = CifParser(crystal_f)
        structures = cif_parser.get_structures()
        length = len(structures)
        if length != 1:
            raise ValueError(
                f"Expected single structure in cif file but found {length}"
            )

        # cif2feffinp below will take the absorber to be a chemical symbol,
        # while this tool should support integer site index or chemical symbol
        # of site (string). Hence convert any integer input to the relevant
        # chemical symbol string.
        try:
            absorbing_atom_int = int(file_format["absorbing_atom"])
            specie = structures[0][absorbing_atom_int].specie
            if isinstance(specie, Species):
                absorbing_atom = str(specie.element)
            else:
                absorbing_atom = str(specie)
        except ValueError:
            absorbing_atom = file_format["absorbing_atom"]

        # NOTE: Here the first site listed in the cif file with the species
        # 'absorbing_atom' is selected as the absorbing atom.
        # NOTE: This generates output for FEFF6 via the 'version8' flag
        inp_data = cif2feff.cif2feffinp(
            crystal_f,
            absorber=absorbing_atom,
            edge=None,
            cluster_size=file_format["radius"],
            absorber_site=1,
            site_index=None,
            extra_titles=None,
            with_h=False,
            version8=False,
        )
        with open(path, "w") as feff_inp_file:
            feff_inp_file.write(inp_data + "\n")
            print(inp_data + "\n")

    else:
        print(f"Copying {crystal_f.name} to {path}")
        shutil.copy(crystal_f, path)

    feff6l(folder=feff_dir, feffinp=feff_inp)

    feff_files = "files.dat"
    input_file = Path(feff_dir, feff_files)
    # the .dat data is stored in fixed width strings
    comma_positions = [13, 21, 32, 41, 48, 61]
    paths_data = []
    # get the list of paths info to assign labels to paths
    try:
        paths_info = get_path_labels(Path(feff_dir, "paths.dat"))
    except FileNotFoundError as err:
        raise FileNotFoundError(
            "paths.dat does not exist, which implies FEFF failed to run"
        ) from err

    print("Reading from: " + str(input_file))
    with open(input_file) as datfile:
        # Read until we find the line at the end of the header
        line = datfile.readline()
        while not re.match("-+", line.strip()):
            line = datfile.readline()

        # Build headers
        line = datfile.readline()
        header = ""
        start = 0
        for end in comma_positions:
            header += line[start:end] + ","
            start = end
        header += f" {'label':30s}, {'select':6s}\n"
        paths_data.append(header)

        # Read data
        line = datfile.readline()
        while line:
            data = ""
            start = 0
            for end in comma_positions[:-1]:
                data += line[start:end] + ","
                start = end

            # Last column needs padding to align
            data += line[start:-1] + "     ,"

            # Add label and select column
            path_id = int(data[5:9])
            try:
                label = paths_info[str(path_id)]["label"] + f".{path_id}"
            except KeyError as err:
                msg = f"{path_id} not in {paths_info.keys()}"
                raise KeyError(msg) from err
            data += f" {label:30s}, {0:6d}\n"
            paths_data.append(data)
            line = datfile.readline()

    with open("out.csv", "w") as out:
        out.writelines(paths_data)


if __name__ == "__main__":
    structure_file = sys.argv[1]
    input_values = json.load(open(sys.argv[2], "r", encoding="utf-8"))
    main(structure_file, input_values["format"])