SOF-7915: refactor espresso structure parser, support for non-zero ibrav

.github/workflows/cicd.yml

@@ -16,21 +16,22 @@ jobs:
 
     steps:
       - name: Checkout this repository
-        uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # v6.0.2
+        uses: actions/checkout@9c091bb21b7c1c1d1991bb908d89e4e9dddfe3e0 # v7.0.0
         with:
           lfs: true
 
       - name: Checkout actions repository
-        uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # v6.0.2
+        uses: actions/checkout@9c091bb21b7c1c1d1991bb908d89e4e9dddfe3e0 # v7.0.0
         with:
-          repository: Exabyte-io/actions
+          repository: mat3ra/actions
           token: ${{ secrets.BOT_GITHUB_TOKEN }}
           path: actions
 
       - name: Run ruff linter
         uses: ./actions/py/lint
         with:
           python-version: ${{ matrix.python-version }}
+          ruff-version: 0.15.20
 
   run-tests:
     runs-on: ubuntu-24.04
@@ -43,14 +44,14 @@ jobs:
 
     steps:
       - name: Checkout this repository
-        uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # v6.0.2
+        uses: actions/checkout@9c091bb21b7c1c1d1991bb908d89e4e9dddfe3e0 # v7.0.0
         with:
           lfs: true
 
       - name: Checkout actions repository
-        uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # v6.0.2
+        uses: actions/checkout@9c091bb21b7c1c1d1991bb908d89e4e9dddfe3e0 # v7.0.0
         with:
-          repository: Exabyte-io/actions
+          repository: mat3ra/actions
           token: ${{ secrets.BOT_GITHUB_TOKEN }}
           path: actions
 
@@ -71,15 +72,15 @@ jobs:
 
     steps:
       - name: Checkout this repository
-        uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # v6.0.2
+        uses: actions/checkout@9c091bb21b7c1c1d1991bb908d89e4e9dddfe3e0 # v7.0.0
         with:
           lfs: true
           fetch-tags: true
 
       - name: Checkout actions repository
-        uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # v6.0.2
+        uses: actions/checkout@9c091bb21b7c1c1d1991bb908d89e4e9dddfe3e0 # v7.0.0
         with:
-          repository: Exabyte-io/actions
+          repository: mat3ra/actions
           token: ${{ secrets.BOT_GITHUB_TOKEN }}
           path: actions
 

express/parsers/apps/espresso/formats/txt.py

@@ -827,9 +827,9 @@ def average_quantity(self, stdout_file: str) -> np.ndarray:
         """
         average_file = find_file(settings.AVERAGE_FILE, self.work_dir)
         # backup in case avg.dat doesn't exist
-        if type(average_file) != str:
+        if type(average_file) is not str:
             average_file = find_file(stdout_file, self.work_dir)
-        if type(average_file) == str and os.path.isfile(average_file):
+        if type(average_file) is str and os.path.isfile(average_file):
             dtype = np.dtype([("x", float), ("planar_average", float), ("macroscopic_average", float)])
             data = np.fromregex(average_file, settings.REGEX["average_quantity"]["regex"], dtype)
             return data

express/parsers/apps/nwchem/parser.py

@@ -45,7 +45,7 @@ def total_energy_contributions(self):
         energy_contributions = self.txt_parser.total_energy_contributions(self._get_file_content(self.stdout_file))
         for key1, value1 in energy_contributions.items():
             for key2, value2 in value1.items():
-                if type(value2) == float:
+                if type(value2) is float:
                     value1[key2] = value2 * Constant.HARTREE
         return energy_contributions
 

express/parsers/structure.py

@@ -1,9 +1,7 @@
-import io
 import json
 
 import pymatgen as mg
 from pymatgen.core.structure import Structure
-from ase.io import read, write
 from jarvis.core.atoms import Atoms
 from jarvis.io.vasp.inputs import Poscar
 
@@ -38,22 +36,40 @@ def __init__(self, *args, **kwargs):
         super().__init__(*args, **kwargs)
         self.structure_string = kwargs.get("structure_string")
         self.structure_format = kwargs.get("structure_format")
+        self.cell_type = kwargs.get("cell_type", "original")  # original, primitive or conventional
 
-        # convert espresso input into poscar
+        # convert espresso input into pymatgen.core.structure
         if self.structure_format == "espresso-in":
-            self.structure_format = "poscar"
-            self.structure_string = self.espresso_input_to_poscar(self.structure_string)
+            from mat3ra.parsers.applications.espresso.pw_x.stdin.material import EspressoPwxStdinMaterial
+
+            material_parser = EspressoPwxStdinMaterial(self.structure_string)
+
+            # Cache the parsed results to avoid multiple property evaluations
+            parsed_lattice = material_parser.lattice
+            parsed_basis = material_parser.basis
+
+            # Reconstruct the PyMatGen Structure object safely from the standardized MADE config outputs
+            lattice_matrix = [
+                parsed_lattice["vectors"]["a"],
+                parsed_lattice["vectors"]["b"],
+                parsed_lattice["vectors"]["c"],
+            ]
+            species = [element["value"] for element in parsed_basis["elements"]]
+            coords = [coord["value"] for coord in parsed_basis["coordinates"]]
+            coords_are_cartesian = (parsed_basis["units"] == "cartesian")
+
+            self.structure = Structure(
+                lattice=lattice_matrix,
+                species=species,
+                coords=coords,
+                coords_are_cartesian=coords_are_cartesian,
+            )
 
         # convert jarvis-db-entry JSON into poscar
-        if self.structure_format == "jarvis-db-entry":
-            self.structure_format = "poscar"
+        elif self.structure_format == "jarvis-db-entry":
             self.structure_string = self.jarvis_db_entry_json_to_poscar(self.structure_string)
-
-        # cell_type is either original, primitive or conventional
-        self.cell_type = kwargs.get("cell_type", "original")
-
-        # Initialize structure class
-        if self.structure_format == "pymatgen.core.structure":
+            self.structure = Structure.from_str(self.structure_string, "poscar")
+        elif self.structure_format == "pymatgen.core.structure":
             structure_as_dict = json.loads(self.structure_string)
             self.structure = Structure.from_dict(structure_as_dict)
         else:
@@ -63,7 +79,7 @@ def __init__(self, *args, **kwargs):
             self.structure = STRUCTURE_MAP[self.cell_type](self.structure)
 
         # keep only one atom inside the basis in order to have the original lattice type
-        self.lattice_only_structure = self.structure.copy() # deepcopy
+        self.lattice_only_structure = self.structure.copy()  # deepcopy
         self.lattice_only_structure.remove_sites(range(1, len(self.structure.sites)))
 
     def lattice_vectors(self):
@@ -178,16 +194,15 @@ def basis(self):
         for i, site in enumerate(self.structure.sites):
             if not site.is_ordered:
                 raise ValueError(
-                   f"Disordered site at {site.frac_coords.tolist()} with "
-                   f"occupancy {site.species} is not supported."
+                    f"Disordered site at {site.frac_coords.tolist()} with "
+                    f"occupancy {site.species} is not supported."
                 )
 
             # Use specie.symbol to strip oxidation state (e.g. "Li0+" → "Li", "O2-" → "O")
             elements.append({"id": i, "value": site.specie.symbol})
-            coordinates.append({
-                "id": i,
-                "value": self._round(site.frac_coords.tolist(), PRECISION_MAP["coordinates_crystal"])
-            })
+            coordinates.append(
+                {"id": i, "value": self._round(site.frac_coords.tolist(), PRECISION_MAP["coordinates_crystal"])}
+            )
         return {"units": "crystal", "elements": elements, "coordinates": coordinates}
 
     def space_group_symbol(self):
@@ -240,27 +255,6 @@ def atomic_constraints(self):
         """
         return self.structure.site_properties.get("selective_dynamics")
 
-    def espresso_input_to_poscar(self, espresso_input):
-        """
-        Extracts structure from espresso input file and returns in poscar format.
-
-        Args:
-            espresso_input (str): input file content
-
-        Returns:
-            str: poscar
-        """
-        input_ = io.StringIO()
-        input_.write(espresso_input)
-        input_.seek(0)
-        atoms = read(input_, format="espresso-in")
-        output_ = io.StringIO()
-        write(output_, atoms, format="vasp", vasp5=True)
-        content = output_.getvalue()
-        input_.close()
-        output_.close()
-        return content
-
     def jarvis_db_entry_json_to_poscar(self, jarvis_db_entry_json_str):
         """
         Extracts structure from jarvis atoms dictionary and returns in poscar format.

pyproject.toml

@@ -15,10 +15,10 @@ classifiers = [
     "Topic :: Software Development",
 ]
 dependencies = [
-    "pymatgen>=2023.8.10",
-    "ase>=3.17.0",
-    "mat3ra-esse>=2026.6.12",
     "jarvis-tools>=2023.12.12",
+    "mat3ra-esse>=2026.6.30.post0",
+    "mat3ra-made[tools]>=2026.7.4.post0",
+    "mat3ra-parsers>=2026.7.4.post0",
     # To avoid module 'numpy.linalg._umath_linalg' has no attribute '_ilp64' in Colab
     "numpy>=1.24.4,<2",
 ]

tests/.gitattributes

@@ -5,3 +5,4 @@ fixtures/vasp/** filter=lfs diff=lfs merge=lfs -text
 fixtures/data.py filter=lfs diff=lfs merge=lfs -text
 fixtures/pyML/test-001/*.JSON filter=lfs diff=lfs merge=lfs -text
 *.cif filter=lfs diff=lfs merge=lfs -text
+fixtures/structural/test-005/*.in filter=lfs diff=lfs merge=lfs -text

tests/fixtures/structural/references.py

@@ -29,3 +29,14 @@
         {"id": 8, "value": [0.333333330, 0.666666670, 0.666666670]},
     ],
 }
+
+SI_IBRAV_BASIS = {
+    "units": "crystal",
+    "elements": [{"id": 0, "value": "Si"}, {"id": 1, "value": "Si"}],
+    "coordinates": [
+        {"id": 0, "value": [0.0, 0.0, 0.0]},
+        {"id": 1, "value": [0.25, 0.25, 0.25]},
+    ],
+}
+
+SI_PRIMITIVE_LATTICE_A = 3.867  # Angstrom

tests/unit/parsers/test_structure_parser.py

@@ -1,7 +1,7 @@
 import os
 
 from tests.unit import UnitTestBase
-from tests.fixtures.structural.references import LI_CIF_BASIS
+from tests.fixtures.structural.references import LI_CIF_BASIS, SI_IBRAV_BASIS, SI_PRIMITIVE_LATTICE_A
 from express.parsers.structure import StructureParser
 
 LI_CIF_PATH = os.path.join(os.path.dirname(__file__), "..", "..", "fixtures", "structural", "test-004", "Li.cif")
@@ -10,6 +10,14 @@
     os.path.dirname(__file__), "..", "..", "fixtures", "structural", "test-004", "SrLaCoO4.cif"
 )
 
+SI_IBRAV0_PATH = os.path.join(
+    os.path.dirname(__file__), "..", "..", "fixtures", "structural", "test-005", "pw_si_ibrav0.in"
+)
+
+SI_IBRAV2_PATH = os.path.join(
+    os.path.dirname(__file__), "..", "..", "fixtures", "structural", "test-005", "pw_si_ibrav2.in"
+)
+
 
 def _read_file(path):
     with open(path) as f:
@@ -46,12 +54,12 @@ class DisorderedStructureParserTest(UnitTestBase):
     when basis() is called on a structure with disordered (mixed-occupancy) sites.
 
     The SrLaCoO4 CIF has Sr2+ and La3+ sharing the same Wyckoff 4e site
-    with occupancy 0.5 each — a canonical disordered case.
+    with occupancy 0.5 each, a disordered case.
     """
 
     def setUp(self):
         super().setUp()
-        # Parsing itself succeeds — pymatgen can load disordered structures.
+        # Parsing itself succeeds as pymatgen can load disordered structures.
         # The error is raised lazily when basis() is called.
         self.parser = StructureParser(
             structure_string=_read_file(DISORDERED_CIF_PATH),
@@ -72,3 +80,59 @@ def test_basis_raises_for_disordered_site(self):
         self.assertIn("is not supported", error)
         self.assertIn("occupancy", error)
         self.assertIn("0.361", error)
+
+
+class EspressoInIbrav0StructureParserTest(UnitTestBase):
+    """Tests that StructureParser correctly parses espresso-in files with ibrav=0
+    and explicit CELL_PARAMETERS.
+    The primitive lattice constant is 3.867 Angstrom."""
+
+    def setUp(self):
+        super().setUp()
+        self.parser = StructureParser(
+            structure_string=_read_file(SI_IBRAV0_PATH),
+            structure_format="espresso-in",
+        )
+
+    def tearDown(self):
+        super().tearDown()
+
+    def test_basis(self):
+        self.assertDeepAlmostEqual(self.parser.basis(), SI_IBRAV_BASIS, places=5)
+
+    def test_formula(self):
+        self.assertEqual(self.parser.formula(), "Si2")
+
+    def test_lattice_bravais_type(self):
+        self.assertEqual(self.parser.lattice_bravais()["type"], "FCC")
+
+    def test_lattice_parameter_a(self):
+        self.assertAlmostEqual(self.parser.lattice_bravais()["a"], SI_PRIMITIVE_LATTICE_A, places=3)
+
+
+class EspressoInIbrav2StructureParserTest(UnitTestBase):
+    """Tests that StructureParser correctly parses espresso-in files with ibrav=2
+    (FCC) defined via celldm(1), without explicit CELL_PARAMETERS.
+    The primitive lattice constant is 3.867 Angstrom."""
+
+    def setUp(self):
+        super().setUp()
+        self.parser = StructureParser(
+            structure_string=_read_file(SI_IBRAV2_PATH),
+            structure_format="espresso-in",
+        )
+
+    def tearDown(self):
+        super().tearDown()
+
+    def test_basis(self):
+        self.assertDeepAlmostEqual(self.parser.basis(), SI_IBRAV_BASIS, places=5)
+
+    def test_formula(self):
+        self.assertEqual(self.parser.formula(), "Si2")
+
+    def test_lattice_bravais_type(self):
+        self.assertEqual(self.parser.lattice_bravais()["type"], "FCC")
+
+    def test_lattice_parameter_a(self):
+        self.assertAlmostEqual(self.parser.lattice_bravais()["a"], SI_PRIMITIVE_LATTICE_A, places=3)