_space_group_magn_transforms.source & _space_group_magn_ssg_transforms.source

[brantonc]

It would be helpful to have the _enumeration_set.state options for _space_group_magn_transforms.source & _space_group_magn_ssg_transforms.source be more similar. Specifically, we need the 'data_block_CURRENT' option to exist for both of them. Also, we need a way to specify another data block within the same file (.e.g. 'datablock') or something like that.

[jamesrhester]

I will add data_block_CURRENT where it is missing, however datablock is not drawn from a specific list of values and so can't be put in an enumerated list, which is why it was left out. I will need to think about how we can make this work. Note that cif_core defines _audit.block_id, which is a unique identifier for a datablock, and _audit_link.block_description, where you can describe how other datablocks are linked to the current datablock, perhaps we can use this somehow.

[brantonc]

Has anything changed lately that might resolve this non-urgent issue?

[jamesrhester]

Perhaps the main thing that has been clarified is that data blocks do not exist in the abstract information space defined by the dictionaries. The dictionaries define no more and no less than a collection of tables, which data blocks populate portions of, so after the notional population has been performed there are no data blocks.

The "proper" way to refer to information in another data block is to associate that information with some identifier, and then to state that identifier in order to refer to it.

[brantonc]

When the _structure.id concept is worked out, feel free to suggest a revision to the descriptions of items in the save_space_group_magn_transforms category.

[jamesrhester]

OK, I have raised an issue in CIF core (https://github.com/COMCIFS/cif_core/issues/442).