templ_enum.cif: add _units.code to save frames

[vaitkus]

It would be useful to have the units of measure in the same save frame as the enumerated values since the same save frame may potentially be imported by multiple definitions. This is probably most useful in the atomic_mass, radius_bond and length_neutron save frames (and potentially in electron_count to make it clear that is unitless and not measured in electrons).

The _units.code can still be retained in the importing save frames and simply ignored or replaced depending on the import mode.

[jamesrhester]

I don't understand the suggestion. Perhaps you could give an example?

[vaitkus]

@jamesrhester sorry for being unclear.

I suggest adding the _units.code attribute directly to some of the save frames in the templ_enum.cif file. An example of this can be seen in PR #396 where femtometres are explicitly added as units in the length_neutron save frame.

[jamesrhester]

Yes, I see now. I agree that this is a good idea.

[rowlesmr]

Questions re units.

Shouldn't all the b coefficients for Cromer-Mann have units of angstrom_squared?

Shouldn't the Fox coefficients all have various powers of angstroms as their units?