Glob logic with bader_analysis_from_path is error-prone

[Andrew-S-Rosen]

Python version

3.11.9

Pymatgen version

2024.9.17.1

Operating system version

Linux

Current behavior

When calling bader_analysis_from_path, I get an indexing error due to poor glob logic.

Expected Behavior

No indexing error

Minimal example

The following files are taken directly from MP.

from monty.serialization import loadfn
from pymatgen.command_line.bader_caller import bader_analysis_from_path

chgcar = loadfn('CHGCAR.json.gz')['data']
aeccar0 = loadfn('AECCAR0.json.gz')['data']
aeccar2 = loadfn('AECCAR2.json.gz')['data']

chgcar.write_file('CHGCAR')
aeccar0.write_file('AECCAR0')
aeccar2.write_file('AECCAR2')
bader_analysis_from_path('.')

Traceback:

File /scratch/gpfs/asrosen/software/miniconda/envs/quacc/lib/python3.11/site-packages/pymatgen/command_line/bader_caller.py:537, in bader_analysis_from_path(path, suffix)
    535 if not aeccar0_path:
    536     warnings.warn("Could not find AECCAR0, interpret Bader results with severe caution!")
--> 537 aeccar0 = Chgcar.from_file(aeccar0_path) if aeccar0_path else None
    539 aeccar2_path = _get_filepath("AECCAR2")
    540 if not aeccar2_path:

File /scratch/gpfs/asrosen/software/miniconda/envs/quacc/lib/python3.11/site-packages/pymatgen/io/vasp/outputs.py:3711, in Chgcar.from_file(cls, filename)
   3701 @classmethod
   3702 def from_file(cls, filename: str) -> Self:
   3703     """Read a CHGCAR file.
   3704
   3705     Args:
   (...)
   3709         Chgcar
   3710     """
-> 3711     poscar, data, data_aug = VolumetricData.parse_file(filename)
   3712     return cls(poscar, data, data_aug=data_aug)

File /scratch/gpfs/asrosen/software/miniconda/envs/quacc/lib/python3.11/site-packages/pymatgen/io/vasp/outputs.py:3562, in VolumetricData.parse_file(filename)
   3557     data_aug = {
   3558         "total": all_dataset_aug.get(0),
   3559         "diff": all_dataset_aug.get(1),
   3560     }
   3561 else:
-> 3562     data = {"total": all_dataset[0]}
   3563     data_aug = {"total": all_dataset_aug.get(0)}
   3564 return poscar, data, data_aug

IndexError: list index out of range

This is because of the following code, which is looking for the .json.gz files instead of the newly written out files:

pymatgen/src/pymatgen/command_line/bader_caller.py

Lines 515 to 527 in fa62ff4

	def _get_filepath(filename: str, msg: str = "") -> str | None:
	paths = glob(glob_pattern := f"{path}/{filename}{suffix}*")
	if len(paths) == 0:
	warnings.warn(msg or f"no matches for {glob_pattern=}")
	return None
	if len(paths) > 1:
	# using reverse=True because, if multiple files are present,
	# they likely have suffixes 'static', 'relax', 'relax2', etc.
	# and this would give 'static' over 'relax2' over 'relax'
	# however, better to use 'suffix' kwarg to avoid this!
	paths.sort(reverse=True)
	warnings.warn(f"Multiple files detected, using {os.path.basename(path)}")
	return paths[0]

Relevant files to reproduce this bug

CHGCAR.json.gz
AECCAR0.json.gz
AECCAR2.json.gz

CC @aryannsaha