Aims bug fixes (#3466)

* Add default species dir option for inputs

Restore the use of an environment variable to set the
aims species directory

* Fix errors in parsers

1) Remove the filter for deciding if line is in a header or calc chunk
   Done since some lines appear in both
2) Remove stray call to an atoms object made from the atomate2 parsers

Modify the AimsOutput test for Si to use phonon calc file and not a relaxation
to check part 2

* Fix error to use f-strings

had an fstring but did not specify it was

* Add header information to all source files

Add myself and @ansobolev as authors

* Reduce test of default species dir

No need to retest everything, just the one functionality

pymatgen/io/aims/inputs.py

@@ -6,6 +6,7 @@
 from __future__ import annotations
 
 import gzip
+import os
 import time
 from copy import deepcopy
 from dataclasses import dataclass, field
@@ -20,6 +21,11 @@
 if TYPE_CHECKING:
     from collections.abc import Sequence
 
+__author__ = "Thomas A. R. Purcell"
+__version__ = "1.0"
+__email__ = "[email protected]"
+__date__ = "November 2023"
+
 
 @dataclass
 class AimsGeometryIn(MSONable):
@@ -566,7 +572,8 @@ def write_file(
                     fd.write(cube.control_block)
 
             fd.write(lim + "\n\n")
-            fd.write(self.get_species_block(structure, self._parameters["species_dir"]))
+            species_dir = self._parameters.get("species_dir", os.environ.get("AIMS_SPECIES_DIR"))
+            fd.write(self.get_species_block(structure, species_dir))
 
     def get_species_block(self, structure: Structure | Molecule, species_dir: str | Path) -> str:
         """Get the basis set information for a structure
@@ -592,7 +599,7 @@ def get_species_block(self, structure: Structure | Molecule, species_dir: str |
                 with gzip.open(f"{filename}.gz", "rt") as sf:
                     sb += "".join(sf.readlines())
             else:
-                raise ValueError("Species file for {sp.symbol} not found.")
+                raise ValueError(f"Species file for {sp.symbol} not found.")
 
         return sb
 

pymatgen/io/aims/outputs.py

@@ -21,6 +21,11 @@
 
     from pymatgen.core import Molecule, Structure
 
+__author__ = "Andrey Sobolev and Thomas A. R. Purcell"
+__version__ = "1.0"
+__email__ = "[email protected] and [email protected]"
+__date__ = "November 2023"
+
 
 class AimsOutput(MSONable):
     """The main output file for FHI-aims."""

pymatgen/io/aims/parsers.py

@@ -16,6 +16,11 @@
 
     from emmet.core.math import Matrix3D, Vector3D
 
+__author__ = "Thomas A. R. Purcell and Andrey Sobolev"
+__version__ = "1.0"
+__email__ = "[email protected] and [email protected]"
+__date__ = "November 2023"
+
 # TARP: Originally an object, but type hinting needs this to be an int
 LINE_NOT_FOUND = -1000
 EV_PER_A3_TO_KBAR = 1.60217653e-19 * 1e22
@@ -997,7 +1002,7 @@ def get_aims_out_chunks(content: str | TextIOWrapper, header_chunk: AimsOutHeade
     Yields:
         The next AimsOutChunk
     """
-    lines = list(filter(lambda x: x not in header_chunk.lines, get_lines(content)))
+    lines = get_lines(content)[len(header_chunk.lines) :]
     if len(lines) == 0:
         return
 
@@ -1122,7 +1127,9 @@ def read_aims_output_from_content(
     check_convergence(chunks, non_convergence_ok)
     # Relaxations have an additional footer chunk due to how it is split
     images = (
-        [chunk.structure for chunk in chunks[:-1]] if header_chunk.is_relaxation else [chunk.atoms for chunk in chunks]
+        [chunk.structure for chunk in chunks[:-1]]
+        if header_chunk.is_relaxation
+        else [chunk.structure for chunk in chunks]
     )
     return images[index]
 

tests/io/aims/test_aims_inputs.py

@@ -2,6 +2,7 @@
 
 import gzip
 import json
+import os
 from pathlib import Path
 
 import numpy as np
@@ -29,7 +30,6 @@ def compare_files(ref_file, test_file):
     for test_line, ref_line in zip(test_lines, ref_lines):
         if "species_dir" in ref_line:
             continue
-
         assert test_line.strip() == ref_line.strip()
 
 
@@ -228,3 +228,38 @@ def test_aims_control_in(tmp_path):
 
     aims_control_from_dict.write_file(si, directory=tmp_path, verbose_header=True, overwrite=True)
     compare_files(infile_dir / "control.in.si", f"{tmp_path}/control.in")
+
+
+def test_aims_control_in_default_species_dir(tmp_path):
+    original_sd = os.environ.get("AIMS_SPECIES_DIR", None)
+    os.environ["AIMS_SPECIES_DIR"] = str(infile_dir.parent / "species_directory/light")
+
+    parameters = {
+        "cubes": [
+            AimsCube(type="eigenstate 1", points=[10, 10, 10]),
+            AimsCube(type="total_density", points=[10, 10, 10]),
+        ],
+        "xc": "LDA",
+        "smearing": ["fermi-dirac", 0.01],
+        "vdw_correction_hirshfeld": True,
+        "compute_forces": True,
+        "relax_geometry": ["trm", "1e-3"],
+        "batch_size_limit": 200,
+        "output": ["band 0 0 0 0.5 0 0.5 10 G X", "band 0 0 0 0.5 0.5 0.5 10 G L"],
+        "k_grid": [1, 1, 1],
+    }
+
+    aims_control = AimsControlIn(parameters.copy())
+
+    for key, val in parameters.items():
+        assert aims_control[key] == val
+
+    si = AimsGeometryIn.from_file(infile_dir / "geometry.in.si.gz").structure
+
+    aims_control.write_file(si, directory=tmp_path, verbose_header=True, overwrite=True)
+    compare_files(infile_dir / "control.in.si.no_sd", f"{tmp_path}/control.in")
+
+    if original_sd is not None:
+        os.environ["AIMS_SPECIES_DIR"] = original_sd
+    else:
+        del os.environ["AIMS_SPECIES_DIR"]