cadrumo.domain.iva.identification module¶

Resolve which Member State a party is IVA-identified in — never where it is established.

Separate from establishment because the two answer legally distinct questions, and conflating them is the defect this module exists to end. A document printing DE811234567 states two things that look like one: that the party operates under a German IVA identification, and — the reader’s inference — that the party is in Germany. Only the first is evidence.

Registration IS the identification fact, so a printed prefix is decisive here. Where establishment refuses to let the same prefix settle a place on its own – the establishment ladder reaches a territory only once a printed country name or postal code corroborates one – this module treats it as what it actually is: the Member State under whose identification the party operates, stated by the party itself on its own invoice. There is nothing to corroborate, because nothing further is being claimed.

And it establishes NOTHING about establishment, symmetrically on both sides. Every Member State registers non-residents on the same terms Spain does, so a German number is exactly as compatible with a sede de actividad económica in Spain as a Spanish N-leader CIF is with one in Germany (Ley 37/1992 arts. 69–70). The asymmetry that made a foreign prefix decisive for place while a Spanish one was correctly refused is closed by splitting the fact, not by demanding corroboration on one side of it.

Spain cannot arise from a printed IVA number here, and that is a property of the prefix vocabulary rather than a judgement made in this module: Spanish identifiers are checksum identifiers routed through the Spanish tax-id validator and ES is absent from NifIvaPrefix. A Spanish identification is therefore established from the Spanish identifier authority or declared, never inferred here.

That absence is an implementation boundary, not a statement that the evidence does not exist. This path recognises a prefix by matching the number’s body against the structural pattern its own prefix claims, and a Spanish identifier is validated by the AEAT control-letter checksum instead – so ES could not join the vocabulary as a pattern the way its siblings did. Whether it SHOULD join, through the checksum validator this codebase already ships, is an open scope question about the axis rather than a settled refusal, and the earlier wording here read as the latter.