noncollinear attribute of Outcar class has wrong value for VASP calculations with LNONCOLLINEAR=.TRUE.

[KylinGuo]

Python version

Python 3.11.0

Pymatgen version

2024.7.18

Operating system version

Ubuntu 22.04 LTS

Current behavior

I expect that outcar.noncollinear is True for noncollinear calculations. But I got False.

Expected Behavior

I expect that outcar.noncollinear is True for noncollinear calculations.

Minimal example

from pymatgen.io.vasp.outputs import Outcar
outcar = Outcar(filename='OUTCAR')
print(outcar.noncollinear)  # 'True' for LNONCOLLINEAR=.TRUE.

### Relevant files to reproduce this bug

_No response_

[janosh]

could you share the OUTCAR file so we can repro?

[KylinGuo]

```python
print(outcar.noncollinear)  # 'True' for LNONCOLLINEAR=.TRUE.

Yes. Please check the attached file (Please remove the suffix .txt).
OUTCAR.txt

[DanielYang59]

Thanks. I was able to reproduce this with the OUTCAR attached. Looks like the pattern match failed:

pymatgen/src/pymatgen/io/vasp/outputs.py

Lines 2114 to 2118 in edcd465

	# Check if calculation is non-collinear
	self.noncollinear = False
	self.read_pattern({"noncollinear": "LNONCOLLINEAR = T"})
	if self.data.get("noncollinear", []):
	self.noncollinear = False

I haven't looked through this part of source code before, wondering why was this re pattern compiled like this (number of spaces hard coded?

[KylinGuo]

Thanks. I was able to reproduce this with the OUTCAR attached. Looks like the pattern match failed:

pymatgen/src/pymatgen/io/vasp/outputs.py

Lines 2114 to 2118 in edcd465

	# Check if calculation is non-collinear
	self.noncollinear = False
	self.read_pattern({"noncollinear": "LNONCOLLINEAR = T"})
	if self.data.get("noncollinear", []):
	self.noncollinear = False

I haven't looked through this part of source code before, wondering why was this re pattern compiled like this (number of spaces hard coded?

Though the number of white spaces are hard-coded, this should work. The problem lies at line 2118. Here is the correct code.

        self.read_pattern({"noncollinear": r"LNONCOLLINEAR\s*=\s*T"})
        if self.data.get("noncollinear", []):
            self.noncollinear = True  # !!!

[DanielYang59]

Yep the logic should be reversed. would fix this. Thanks!