energy_data_retrieval.wind_catalogue.ons#

Bounded ONS wind unit observations; no farm capacity or lifecycle inference.

energy_data_retrieval.wind_catalogue.ons.wri_id(ceg)[source]#

Apply the documented WRI Brazil importer mapping, never a name-derived ID.

Parameters:

ceg (str)

Return type:

str

energy_data_retrieval.wind_catalogue.ons.parse_partition(data, *, sha256, offset, limit, retrieved_at)[source]#

Parse one sorted full-CEG partition of a pinned daily CSV.

A unit can represent multiple turbines. Nominal MW follows the dictionary, despite the legacy column name val_potenciaefetiva. Retrieval time is not an observation date; release-for-operation dates are not farm commissioning.

Parameters:
  • data (bytes)

  • sha256 (str)

  • offset (int)

  • limit (int)

  • retrieved_at (str)

Return type:

dict[str, Any]

energy_data_retrieval.wind_catalogue.ons.audit_partitions(partitions)[source]#

Verify a contiguous prefix of one pinned release before resuming its scan.

Report structural coverage only. Reviewed or accepted identities and complete turbine inventories cannot be inferred from a successfully parsed partition. Input order is immaterial; gaps, overlaps and daily-source drift are errors.

Parameters:

partitions (list[dict[str, Any]])

Return type:

dict[str, Any]

energy_data_retrieval.wind_catalogue.ons.reconcile(candidates, existing, *, source_id, identity_source_id, now)[source]#

Enrich unique registry-ID plus exact-name matches; preserve all scalar facts.

New/ambiguous identities remain in the ledger. Case and surrounding whitespace are normalized, but accents, words and phase suffixes are not approximated. Replaying the same source is idempotent; changed source content is rejected.

Parameters:
  • candidates (list[dict[str, Any]])

  • existing (list[dict[str, Any]])

  • source_id (str)

  • identity_source_id (str)

  • now (str)

Return type:

tuple[list[dict[str, Any]], list[dict[str, Any]]]