cadrumo.domain.calculations.registry.restated_families module¶
An edition’s authored statement that it restates a family in full.
A delta edition states only what changed and inherits the rest from its predecessor, family by family. That inheritance is the right reading whenever the successor’s silence about a member means “unchanged”. It is the wrong reading when the official structure the successor was drawn from states the family differently end to end: the members the successor does not repeat are not unchanged, they are gone, and inheriting them would carry into the successor rows nobody established for it.
A restatement declaration is the edition saying so. It names the family, the cause, and the reason somebody wrote, and it means: this edition states this family in full here; the merge does not inherit it on this edge. Absence cannot carry that claim – an edition that inherits and an edition that meant to restate look identical from the successor’s files alone – so the claim is authored rather than inferred from a member count or a diff.
The family is closed to INHERITED_FAMILIES, because restating a family
the merge never inherits withdraws nothing and would read as though it did.
- INHERITED_FAMILIES: Final[frozenset[str]]¶
Every family the predecessor merge carries forward, and so the only families a successor can meaningfully decline to inherit.
casillasinherit through the loader’s own casilla pass; the rest are the loader’s keyed families, which inherit by member identity. The set is projected from the domain-owned policy table so the authored claim and the merge are closed on one vocabulary.A family absent here is not inherited, so it has nothing to decline; whether a family is inherited is read from the policy table and never restated here, because a second list of examples goes stale the moment one is enrolled. Enrolling a family is an inheritance judgement made on that table, not a consequence of wanting to restate it.
- class RestatedFamilyCause(*values)[source]¶
Bases:
StrEnumWhy an edition states a family in full instead of inheriting it.
One member today.
official_structure_differsis the only cause that makes inheritance wrong rather than merely redundant: the successor’s governing document lays the family out end to end, so the predecessor’s unrepeated members are withdrawn by that document and not by an editorial choice. Spelled as an enumerated field rather than assumed, so a second, differently-grounded cause is an added member reviewed on its own instead of a reinterpretation of every reason string already written.- OFFICIAL_STRUCTURE_DIFFERS¶