"""Accept explicitly reviewed owner-identified plants without invented registry IDs."""
import json
from pathlib import Path
from typing import Any
from energy_data_retrieval.pv_catalogue.catalogue import canonical_json, load_records
from .catalogue import stable_id, upsert_records, validate_catalogue
from .gppd import evidence_record
from .kengen import SOURCE_ID
from .ons_new import name_key
[docs]
def new_record(entry: dict[str, Any], *, now: str) -> dict[str, Any]:
"""Use a frozen catalogue accession, not a purported national licence ID."""
review, item = entry["review"], entry["inventory"]
if (
review.get("decision") != "reviewed_new_identity"
or review.get("source_name") != item["source_name"]
or not review.get("identity_basis")
or not review.get("accession")
):
raise ValueError("Explicit owner identity review required")
return evidence_record(
{
"record_id": stable_id("projects", "reviewed-accession", review["accession"]),
"entity_type": "projects",
"name": item["source_name"],
"country_code": "KEN",
"record_scope": "plant",
"capacity_scope": "Owner station; unresolved physical overlap",
"aggregation_role": "unresolved",
"status": "unknown",
"status_as_of": now[:10],
"status_date_basis": (
"Review date; historical statistics do not establish current status"
),
"technology": "unknown",
"owner_names": ["Kenya Electricity Generating Company PLC"],
"external_ids_json": "{}",
"observations_json": canonical_json(item["observations"]),
"extensions_json": canonical_json(
{
"kengen_new_identity_review": review,
"owner_identity_source": "kengen-portfolio-20261106",
"registry_id_status": "Unknown; EPRA download returned HTTP 503",
}
),
},
source_id=SOURCE_ID,
now=now,
locator="KenGen report pp. 246/249; per-cell observation locators",
)
[docs]
def accept_new(root: Path, batch: dict[str, Any], output: Path) -> int:
"""Preflight every identity and name collision before writing one new shard."""
existing = load_records(root)
lookup = {r["record_id"]: r for r in existing}
incoming = [new_record(e, now=batch["reviewed_at"]) for e in batch["accepted"]]
if len({r["record_id"] for r in incoming}) != len(incoming):
raise ValueError("Duplicate incoming accession")
new = []
for row in incoming:
# The portfolio corroborates identity only, not the report's numeric cells.
fs, ev = json.loads(row["field_sources_json"]), json.loads(row["evidence_json"])
fs["extensions_json"].append("kengen-portfolio-20261106")
ev["extensions_json"]["locator"] = "Named station sections; review basis in extensions_json"
row.update(
field_sources_json=canonical_json(fs),
evidence_json=canonical_json(ev),
sources=sorted(set(row["sources"] + ["kengen-portfolio-20261106"])),
)
if row["record_id"] in lookup:
if lookup[row["record_id"]] != row:
raise ValueError("Existing accession differs; conflict review required")
continue
for prior in existing + new:
if prior["country_code"] == "KEN" and name_key(row["name"]) in {
name_key(n) for n in [prior["name"], *prior.get("aliases", [])]
}:
raise ValueError("Name collision requires identity review")
new.append(row)
combined, conflicts = upsert_records(existing, new, now=batch["reviewed_at"])
if conflicts:
raise ValueError("Insertion conflict")
validate_catalogue(combined, json.loads((root / "sources.json").read_text()))
if new:
if output.exists():
raise ValueError("Refuse to overwrite existing shard")
output.write_text(
"".join(canonical_json(r) + "\n" for r in sorted(new, key=lambda r: r["record_id"]))
)
return len(new)