Open e-kwsm opened 1 year ago
Is F 4/m -3 2/m
invalid and silently converted to P 1
in Open Babel?
I am afraid you must go to the source code of the CIF import in OB and investigate yourself.
@e-kwsm By the lattice vectors (a,b,c) -- accounting both for lengths and enclosed angles \alpha, \beta, \gamma -- the model is in the cubic class, i.e. a = b = c and \alpha = \beta = \gamma. The single x, y, z
in the loop of symmetry operators were suitable for the monoclinic space group P1
, but not sufficient to describe higher symmetry.
For a contemporary crystallographic model about NaCl in symmetry F m -3 m
and lattice vectors not too far away from your example, see the public .cif file attached below about COD 2108652. The loop in question starts on line 106 and ends with line 229. With this in hand, obabel (version 3.1.1 as packaged by Linux Debian 13/trixie) does not report a problem, i.e.
$ obabel 2108652.cif -ocif > /dev/null
1 molecule converted
What was the source of the .cif file you used in your example? Curators of ICSD, CCDC/CSD, COD, etc. usually welcome if users' reports allow to improve their database further.
@e-kwsm An addition: for NaCl, this page by TU Graz/Austria compiles all 192 symmetry operators of space group 225.
OB use an database of HM symbols. This database cover only a subset of possible symbols (all standard, some non standard). It does not for sure cover all possible symbols combination. You can find this database in OB source code. I am not sure OB is able to ignore the HM symbols and interpret the symetry operations list (this should be the correct behaviour) ... Anyway I suggest to study crystallography and hava a look on IUCR symmetry operaton tables book ....
Environment Information
Open Babel version: 2b211d6acfc0f9e1c6746a984da3483b729aa32b Operating system and version: EndeavourOS
Expected Behavior
CIF has a Hermann–Mauguin symbol entry,
space_group_name_H-M_alt
:(emphasis mine)
Usually short forms of HM symbols are used instead of full ones, e.g., #225
F m -3 m
vsF 4/m -3 2/m
.CIF also has symmetry operation entry,
_space_group_symop_operation_xyz
:(emphasis mine)
obabel behaviour must be independent from the Hermann–Mauguin notation form and the symmetry operartions.
Actual Behavior
The geometry is converted to
P 1
when the symmetry operations are insufficient, but no messages are issued for the full HM symbol:space_group_name_H-M_alt
in input CIF_space_group_symop_operation_xyz
in input CIF_space_group_name_H-M_alt
in generated CIFF m -3 m
(short)F m 3 m
F m -3 m
(short)P 1
F m -3 m
(short)F m 3 m
F 4/m -3 2/m
(full)F m 3 m
F 4/m -3 2/m
(full)P 1
F 4/m -3 2/m
(full)F m 3 m
Steps to Reproduce
Prepare the following file named as NaCl.cif:
Here,
_space_group_symop_operation_xyz
is apparently insufficient.all of the operations for Fm-3m
Then
If
F m -3 m
is replaced withF 4/m -3 2/m
, the messages are goneBut the spacegroup of the output is
P 1
.