CIF parsing does not work when pdb flags are used

[Andrew-S-Rosen]

Python version

3.12

Pymatgen version

2025.10.7

Operating system version

No response

Current behavior

The Pymatgen CIF parser cannot properly parse CIFs where the pdb formatting flags are used. Most parsers can't, although the official CCDC Mercury GUI can. This is a horrible format, and I'm not suggesting someone waste their time to fix it, but I'm reporting it anyway.

demo.cif.txt

from pymatgen.core import Structure

s = Structure.from_file('demo.cif')

Expected Behavior

CIF should be parsed.

Minimal example

Relevant files to reproduce this bug

No response

[CompRhys]

It would bring in an extra dep but I have found https://parsnip-cif.readthedocs.io/en/v0.5.0/ nice to use and it works here.

# %%
import sys
import os

def test_parsnip_parsing(filepath):
    print(f"--- Testing parsnip-cif on {filepath} ---")

    try:
        import parsnip
    except ImportError:
        print("Error: 'parsnip' library not found. Please install it with 'pip install parsnip-cif'.")
        return

    print(f"Using parsnip version: {getattr(parsnip, '__version__', 'unknown')}")

    try:
        # According to documentation/search, parsnip.parse() is a common entry point
        # or CifFile class. Let's try to detect the API.

        if hasattr(parsnip, 'parse'):
            print("Detected 'parsnip.parse' API...")
            with open(filepath, 'r') as f:
                content = f.read()
                blocks = parsnip.parse(content)
                print(f"✓ Successfully parsed {len(blocks)} blocks.")
                for i, block in enumerate(blocks):
                    print(f"  Block {i}: {getattr(block, 'name', 'unnamed')}")
                    # Try to extract formula or space group if possible
                    # This depends on the specific structure of the parsed blocks
                    if hasattr(block, 'items'):
                        items = dict(block.items())
                        formula = items.get('_chem_formula_sum') or items.get('_chemical_formula_sum')
                        sg = items.get('_symmetry_space_group_name_H-M') or items.get('_space_group_name_H-M_alt')
                        if formula: print(f"    Formula: {formula}")
                        if sg: print(f"    Space Group: {sg}")

        elif hasattr(parsnip, 'CifFile'):
            print("Detected 'parsnip.CifFile' API...")
            cif = parsnip.CifFile(filepath)
            print("✓ Successfully loaded CifFile.")

            print("\n--- Pairs (Key-Value Data) ---")
            for key, value in list(cif.pairs.items())[:10]:
                print(f"  {key}: {value}")
            if len(cif.pairs) > 10:
                print(f"  ... and {len(cif.pairs) - 10} more pairs.")

            print("\n--- Loops (Tabular Data) ---")
            for i, loop in enumerate(cif.loops):
                print(f"  Loop {i}:")
                if hasattr(loop, 'keys'):
                    print(f"    Columns: {list(loop.keys())}")
                    # Try to get number of rows from first column
                    first_col = next(iter(loop.values()))
                    print(f"    Rows: {len(first_col)}")
                elif isinstance(loop, dict):
                    print(f"    Columns: {list(loop.keys())}")
                    first_col = next(iter(loop.values()))
                    print(f"    Rows: {len(first_col)}")
                else:
                    print(f"    Type: {type(loop)}")

        else:
            print("Unknown parsnip API. Available attributes:")
            print([attr for attr in dir(parsnip) if not attr.startswith('_')])

    except Exception as e:
        print(f"✗ Failed to parse file: {e}")
        import traceback
        traceback.print_exc()

    return cif

# %%
cif_path = "/Users/radical-rhys/Downloads/demo.cif.txt"
if not os.path.exists(cif_path):
    print(f"Errcor: File not found at {cif_path}")
else:
    cif = test_parsnip_parsing(cif_path)
    print(cif.loop_labels)

>>> [('_chem_comp.id', '_chem_comp.type'),
 ('_atom_type.symbol',),
 ('_atom_site.group_PDB',
  '_atom_site.id',
  '_atom_site.type_symbol',
  '_atom_site.label_atom_id',
  '_atom_site.label_alt_id',
  '_atom_site.label_comp_id',
  '_atom_site.label_asym_id',
  '_atom_site.label_entity_id',
  '_atom_site.label_seq_id',
  '_atom_site.pdbx_PDB_ins_code',
  '_atom_site.Cartn_x',
  '_atom_site.Cartn_y',
  '_atom_site.Cartn_z',
  '_atom_site.occupancy',
  '_atom_site.B_iso_or_equiv',
  '_atom_site.pdbx_formal_charge',
  '_atom_site.auth_seq_id',
  '_atom_site.auth_asym_id',
  '_atom_site.pdbx_PDB_model_num')]