fix(outputs.py): handle missing trailing newline in ICOHPLIST.lobster (#4350)

* fix(outputs.py): handle missing trailing newline in ICOHPLIST.lobster

LOBSTER 5.1.1 sometimes produces an ICOHPLIST.lobster file without a trailing
newline, which causes the Icohplist parser to miscount bond entries (e.g., 1079 vs. 1080).
This commit updates the file-reading logic in Icohplist.__init__ to use splitlines()
and filters out any blank lines, ensuring that only valid, non-empty lines are parsed.
This change prevents the ValueError ("COHPCAR and ICOHPLIST do not fit together")
and aligns the bond count with COHPCAR.lobster.

Ref: See related issue in lobsterpy (#389) for additional context.

Signed-off-by: Ali Hussain Umar Bhatti <[email protected]>

* pre-commit auto-fixes

* Update outputs.py

* Add test_missing_trailing_newline

* Robust header detection and version parsing in Icohplist

- Strip only trailing blank lines so a missing final newline still works
- Compute header length dynamically (skip title and optional spin‐line)
- Determine LOBSTER version by column count (6→2.2.1, 8→3.1.1, 9→5.1.0)
- Preserve non‐orbitalwise LCFO entries (fixes length mismatch for non-orbitalwise LCFO lists)

---------

Signed-off-by: Ali Hussain Umar Bhatti <[email protected]>
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Shyue Ping Ong <[email protected]>

Author: 3 people

src/pymatgen/io/lobster/outputs.py

@@ -405,24 +405,35 @@ def __init__(
         # and we don't need the header.
         if self._icohpcollection is None:
             with zopen(self._filename, mode="rt", encoding="utf-8") as file:
-                all_lines = file.read().split("\n")
-                lines = all_lines[1:-1] if "spin" not in all_lines[1] else all_lines[2:-1]
-            if len(lines) == 0:
-                raise RuntimeError("ICOHPLIST file contains no data.")
-
-            # Determine LOBSTER version
-            if len(lines[0].split()) == 8 and "spin" not in all_lines[1]:
-                version = "3.1.1"
-            elif (len(lines[0].split()) == 8 or len(lines[0].split()) == 9) and "spin" in all_lines[1]:
-                version = "5.1.0"
-            elif len(lines[0].split()) == 6:
-                version = "2.2.1"
-                warnings.warn(
-                    "Please consider using a newer LOBSTER version. See www.cohp.de.",
-                    stacklevel=2,
-                )
-            else:
-                raise ValueError("Unsupported LOBSTER version.")
+                all_lines = file.read().splitlines()
+
+                # strip *trailing* blank lines only
+                all_lines = [line for line in all_lines if line.strip()]
+                # --- detect header length robustly ---
+                header_len = 0
+                try:
+                    int(all_lines[0].split()[0])
+                except ValueError:
+                    header_len += 1
+                if header_len < len(all_lines) and "spin" in all_lines[header_len].lower():
+                    header_len += 1
+                lines = all_lines[header_len:]
+                if not lines:
+                    raise RuntimeError("ICOHPLIST file contains no data.")
+                # --- version by column count only ---
+                ncol = len(lines[0].split())
+                if ncol == 6:
+                    version = "2.2.1"
+                    warnings.warn(
+                        "Please consider using a newer LOBSTER version. See www.cohp.de.",
+                        stacklevel=2,
+                    )
+                elif ncol == 8:
+                    version = "3.1.1"
+                elif ncol == 9:
+                    version = "5.1.0"
+                else:
+                    raise ValueError(f"Unsupported LOBSTER version ({ncol} columns).")
 
             # If the calculation is spin polarized, the line in the middle
             # of the file will be another header line.
@@ -587,6 +598,10 @@ def icohplist(self) -> dict[Any, dict[str, Any]]:
                 "translation": value._translation,
                 "orbitals": value._orbitals,
             }
+
+        # for LCFO only files drop the single orbital resolved entry when not in orbitalwise mode
+        if self.is_lcfo and not self.orbitalwise:
+            icohp_dict = {k: d for k, d in icohp_dict.items() if d.get("orbitals") is None}
         return icohp_dict
 
     @property
@@ -1720,7 +1735,12 @@ def has_good_quality_check_occupied_bands(
             raise ValueError("number_occ_bands_spin_down has to be specified")
 
         for spin in (Spin.up, Spin.down) if spin_polarized else (Spin.up,):
-            num_occ_bands = number_occ_bands_spin_up if spin is Spin.up else number_occ_bands_spin_down
+            if spin is Spin.up:
+                num_occ_bands = number_occ_bands_spin_up
+            else:
+                if number_occ_bands_spin_down is None:
+                    raise ValueError("number_occ_bands_spin_down has to be specified")
+                num_occ_bands = number_occ_bands_spin_down
 
             for overlap_matrix in self.band_overlaps_dict[spin]["matrices"]:
                 sub_array = np.asarray(overlap_matrix)[:num_occ_bands, :num_occ_bands]
@@ -2333,7 +2353,7 @@ def _parse_matrix(
         file_data: list[str],
         pattern: str,
         e_fermi: float,
-    ) -> tuple[list[float], dict, dict]:
+    ) -> tuple[list[np.ndarray], dict[Any, Any], dict[Any, Any]]:
         complex_matrices: dict = {}
         matrix_diagonal_values = []
         start_inxs_real = []

tests/io/lobster/test_outputs.py

@@ -3,6 +3,7 @@
 import copy
 import json
 import os
+import tempfile
 
 import numpy as np
 import pytest
@@ -1955,7 +1956,7 @@ def test_attributes(self):
         assert self.icohp_lcfo.is_spin_polarized
         assert len(self.icohp_lcfo.icohplist) == 28
         assert not self.icohp_lcfo_non_orbitalwise.orbitalwise
-        assert len(self.icohp_lcfo_non_orbitalwise.icohplist) == 27
+        assert len(self.icohp_lcfo_non_orbitalwise.icohplist) == 28
 
     def test_values(self):
         icohplist_bise = {
@@ -2171,6 +2172,23 @@ def test_msonable(self):
             else:
                 assert getattr(icohplist_from_dict, attr_name) == attr_value
 
+    def test_missing_trailing_newline(self):
+        content = (
+            "1   Co1   O1   1.00000   0   0   0   -0.50000   -1.00000\n"
+            "2   Co2   O2   1.10000   0   0   0   -0.60000   -1.10000"
+        )
+
+        with tempfile.NamedTemporaryFile("w+", delete=False) as tmp:
+            tmp.write(content)
+            tmp.flush()
+            fname = tmp.name
+        try:
+            ip = Icohplist(filename=fname)
+            assert len(ip.icohplist) == 2
+            assert ip.icohplist["1"]["icohp"][Spin.up] == approx(-0.5)
+        finally:
+            os.remove(fname)
+
 
 class TestNciCobiList:
     def setup_method(self):