Discussion about layer group support is raised at #199.
Let's continue the discussion here.

Possible issues

• Mapping between layer group and space group.
• Hard coded periodic_axes.

Test and refactoring

By @lan496,

For testing, we may need another repository to check layer-group implementations. For example, I've created and checked magnetic-space-group implementations for MAGNDATA in a private repo. For refactoring, wrapper functions to hide comparison of cell->aperiodic_axis may be helpful, like inline int is_aperiodic_axis(const *Cell cell, const int axis).

Future plan

Proposed by @site-g

• Systematic test (Almost done by @mikaelkuisma1)
• Something in datasets to indicate group type (now negative hall number means layer group)
• C-APIs (now only spg_get_layer_dataset)
• Python, etc. interfaces
• Documentation for layer group
• Add group_class to SpglibDataset