Source code for energy_data_retrieval.wind_catalogue.connectors

"""Offline exact-part OEM tables and dated wind-assembly crosswalks.

Only public product classifications are selected from the rendered page state.
Unrelated site configuration, prices and inferred mating combinations are omitted.
"""

from __future__ import annotations

import hashlib
import json
import re
from html.parser import HTMLParser
from typing import Any

from energy_data_retrieval.pv_catalogue.catalogue import canonical_json

from .catalogue import stable_id
from .gppd import evidence_record

REFERENCE_IDS = ("harting_09120053101_20261110",)


class _State(HTMLParser):
    def __init__(self) -> None:
        super().__init__()
        self.active = False
        self.parts: list[str] = []
        self.count = 0

    def handle_starttag(self, tag, attrs):
        if tag == "script" and dict(attrs).get("id") == "ng-state":
            self.count += 1
            self.active = True

    def handle_data(self, data):
        if self.active:
            self.parts.append(data)

    def handle_endtag(self, tag):
        if tag == "script":
            self.active = False


class _Text(HTMLParser):
    def __init__(self) -> None:
        super().__init__()
        self.parts: list[str] = []

    def handle_starttag(self, tag, attrs):
        if tag == "sup":
            self.parts.append("^(")
        elif tag == "sub":
            self.parts.append("_(")
        elif tag == "br":
            self.parts.append("; ")

    def handle_endtag(self, tag):
        if tag in {"sup", "sub"}:
            self.parts.append(")")

    def handle_data(self, data):
        self.parts.append(data)


def _text(value: str) -> str:
    parser = _Text()
    parser.feed(value)
    return " ".join("".join(parser.parts).replace("\u200c", "").split())


[docs] def parse_product(original: bytes, part_number: str) -> dict[str, Any]: """Require the exact OEM part; retain units, inequalities and superscripts. No numerical coercion is applied: UL/CSA voltage, impulse voltage, limiting temperatures and conditional IP ratings remain separate named observations. Missing or duplicate product/feature identities fail closed. """ if not re.fullmatch(r"\d{11}", part_number): raise ValueError("Exact eleven-digit part number required") parser = _State() parser.feed(original.decode("utf-8")) if parser.count != 1 or parser.active: raise ValueError("Missing or incomplete public product state") state = json.loads("".join(parser.parts)) product = state.get(f'CS|D|CommerceProductService#getProduct?["{part_number}"]') if not product or product.get("code") != part_number or not product.get("name"): raise ValueError("Exact product identity absent; redirects are not evidence") groups = product.get("classifications") or [] if not groups or not any(g.get("name") == "Identification" for g in groups): raise ValueError("Missing product classifications") attributes = [] seen = set() for group in groups: for feature in group["features"]: key = (group["code"], feature["code"]) if key in seen: raise ValueError("Duplicate product attribute") seen.add(key) attributes.append( { "section": group["name"], "attribute_id": feature["code"], "name": feature["name"], "values": [_text(v["value"]) for v in feature["featureValues"]], "normalization": ( "HTML to text; superscripts ^(), subscripts _(); units retained" ), "locator": f"Product details / {group['name']} / {feature['name']}", } ) return {"part_number": part_number, "name": product["name"], "attributes": attributes}
[docs] def stage_connectors( documents: list[dict[str, Any]], originals: dict[str, bytes], counterparts: list[dict[str, Any]], existing: list[dict[str, Any]], *, now: str, ) -> tuple[list[dict[str, Any]], list[dict[str, Any]]]: """Review a complete numbered assembly partition without claiming fitment. Exact part-number equality supports a reference crosswalk, not equivalence between current OEM specifications and the installed historical revision. Existing cross-sector IDs must be supplied and are held for explicit reuse. """ expected = {r["part_number_as_printed"].replace(" ", "") for r in counterparts} parts = [d["part_number"] for d in documents] if len(set(parts)) != len(parts) or set(parts) != expected: raise ValueError("Incomplete or duplicate counterpart partition") records, ledger = [], [] for doc in sorted(documents, key=lambda d: d["part_number"]): part, sid = doc["part_number"], doc["source_id"] raw = originals[sid] if hashlib.sha256(raw).hexdigest() != doc["sha256"]: raise ValueError("Original product checksum mismatch") product = parse_product(raw, part) matches = [ r for r in existing if r.get("manufacturer", "").casefold() == "harting" and str(r.get("model", "")).replace(" ", "") == part ] if matches: ledger.append( { "part_number": part, "decision": "existing_identity_review", "record_ids": [r["record_id"] for r in matches], } ) continue applications = [ r for r in counterparts if r["part_number_as_printed"].replace(" ", "") == part ] if any(not r.get("source_id") or not r.get("locator") for r in applications): raise ValueError("Counterpart provenance required") extension = { "connector_specifications": { **product, "source_id": sid, "sha256": doc["sha256"], "revision": None, "published_at": None, "retrieved_at": now, "region": "United Kingdom / English public product page", "assembly_references": applications, "installed_revision_verified": False, "specification_scope": "Dated OEM claims, not independently validated performance", "derating_curves": "Not imported; curve axes/conditions require separate review", } } row = evidence_record( { "record_id": stable_id("equipment", "oem-electrical", "HARTING", part), "entity_type": "equipment", "record_scope": "configuration", "name": f"HARTING {part} — {product['name']}", "manufacturer": "HARTING", "model": part, "variant": "OEM part " + part + "; UK public specification", "equipment_category": "connector", "capacity_scope": "Connector electrical ratings; no turbine active power", "external_ids_json": canonical_json({"harting_part_number": part}), "extensions_json": canonical_json(extension), "notes": "Exact numbered connector, insert, contact or hood. Assembly references " "identify dated counterpart requirements only; current OEM revision, installed " "revision, turbine/project fitment and warranty are unresolved. " "No active power inferred.", }, source_id=sid, now=now, locator="Product details / exact part number and classifications", ) sources = sorted({sid, *(r["source_id"] for r in applications)}) row["sources"] = sources fs = json.loads(row["field_sources_json"]) fs["extensions_json"] = sources row["field_sources_json"] = canonical_json(fs) ev = json.loads(row["evidence_json"]) ev["extensions_json"]["supporting_evidence"] = applications row["evidence_json"] = canonical_json(ev) records.append(row) ledger.append( { "part_number": part, "record_id": row["record_id"], "decision": "addition", "source_id": sid, } ) return records, ledger