add diffrn_id to _pd_diffractogram and _pd_phase

[rowlesmr]

From COMCIFS/cif_core#344 (comment)

Can now link a diffractogram and phases to the conditions under which the data were collected.

[jamesrhester]

The _pd_diffractogram.diffrn_id part is fine. The part that I'm not sure about is linking phase with _diffrn.id through _pd_phase.diffrn_id. Presumably we could link a phase to a _structure.id, which would then point to a _diffrn.id so there's no need for an explicit link.

Also, I feel like a given phase can be associated either with a diffractogram and therefore with a _diffrn.id (so no need for a data name), or with multiple diffractograms and therefore with no particular _diffrn.id (so it makes no sense).

[rowlesmr]

The _pd_diffractogram.diffrn_id part is fine.

\o/

The part that I'm not sure about is linking phase with _diffrn.id through _pd_phase.diffrn_id. Presumably we could link a phase to a _structure.id, which would then point to a _diffrn.id so there's no need for an explicit link.

This is to allow for amorphous phases to have diffraction conditions.

Also, I feel like a given phase can be associated either with a diffractogram and therefore with a _diffrn.id (so no need for a data name), or with multiple diffractograms and therefore with no particular _diffrn.id (so it makes no sense).

This is more persuasive. Although, couldn't the same argument be made of _structure.diffrn_id (which apart from amorphous phases, is essentially the same thing)? For powder, I read DIFFRN as providing a set of conditions underwhich data were collected. Does it make sense to have the same phase being used in two different diffractograms at, for example, 10 K and 1000 K?