Allow standard uncertainty values for multiple data items

[vaitkus]

Multiple data items in the DDLm version of the CIF_CORE dictionary have been allowed to have associated standard uncertainty (s.u.) values (see issue COMCIFS/cif_core#107). This change should also be reflected in the DDL1 version. Affected data items:

_diffrn_orient_refln_angle_[chi, kappa, omega, phi, psi, theta]
_diffrn_refln_counts_[bg_1, bg_2, net, peak, total] †
_diffrn_refln_intensity_net
_diffrn_scale_group_i_net
_diffrn_standards_scale_[sigma, u]
_refln_[a, b, f, f_squared]_[calc, meas]
_refln_intensity_[calc, meas]
_refln_phase_calc
_cell_measurement_refln_theta
_exptl_crystal_density_diffrn
_exptl_crystal_size_[length, min, mid, max, rad]
_exptl_crystal_face_diffr_[chi, kappa, phi, psi]
_exptl_crystal_face_perp_dist
_atom_type_analytical_mass_%
_chemical_formula_weight_meas
_geom_bond_valence
_atom_sites_cartn_tran_matrix_[1, 2, 3][1, 2, 3]
_atom_sites_cartn_tran_vector_[1, 2, 3]
_atom_sites_fract_tran_matrix_[1, 2, 3][1, 2, 3]
_atom_sites_fract_tran_vector_[1, 2, 3]
_refine_ls_restrained_s_[all, obs, gt]
_refine_ls_F_calc_precision

† Note, that the SU eligibility of items _diffrn_refln_counts_[bg_1, bg_2, net, peak, total] may change depending on the resolution of issue COMCIFS/cif_core#278.

[jcbollinger]

It may be that some of these items instead should not be afforded the possibility of an SU in the DDLm core. In addition to the _diffrn_refln_counts_* items, I am looking at _diffrn_standards_scale_[sigma, u] and _refine_ls_restrained_s_[all, obs, gt].

Also, although _atom_sites_cartn_tran_matrix_[1, 2, 3][1, 2, 3], _atom_sites_cartn_tran_vector_[1, 2, 3], _atom_sites_fract_tran_matrix_[1, 2, 3][1, 2, 3], and _atom_sites_fract_tran_vector_[1, 2, 3] are derived from cell parameters, many of which do have standard uncertainties, I think there are circumstances under which the transformations described are meant to be or should be treated as exact. At minimum, one needs to recognize that if they are treated as inexact transformations then they are irreversible with respect to their effects on SUs.

[vaitkus]

Acknowledged, I will not go forward with these changes on DDL1 until this is resolved in DDLm. I have reopened the relevant DDLm issue [1].

[1] COMCIFS/cif_core#107