# Python 3. No packages required. Optional key: DATASCOOP_API_KEY.
import json
import os
import sys
from datetime import datetime
from urllib.error import HTTPError, URLError
from urllib.parse import urlencode
from urllib.request import Request, urlopen

ENDPOINT = "https://datascoop.io/v1/datasets/cqc/query"
DATASET = "cqc"
FILTERS = json.loads("{\"careHome\":\"true\",\"localAuthorityExact\":\"Devon\",\"serviceType\":\"Nursing homes\"}")
PAGE_SIZE, MAX_ROWS, MAX_PAGES = 25, 2500, 100
HEADERS = {"Accept": "application/json"}
key = os.environ.get("DATASCOOP_API_KEY", "").strip()
if key:
    HEADERS["Authorization"] = "Bearer " + key

def text_value(value):
    return "true" if value is True else "false" if value is False else str(value)

def retrieve():
    rows, total, imported_at, first = [], None, None, None
    for _ in range(MAX_PAGES):
        params = dict(FILTERS, strict_filters="1", limit=PAGE_SIZE, offset=len(rows))
        request = Request(ENDPOINT + "?" + urlencode(params), headers=HEADERS)
        with urlopen(request, timeout=30) as response:
            result = json.load(response)
        if (not isinstance(result, dict) or result.get("dataset") != DATASET or
            not isinstance(result.get("rows"), list) or
            any(not isinstance(row, dict) for row in result["rows"]) or
            not isinstance(result.get("ignored"), list) or result["ignored"] or
            type(result.get("total")) is not int or result["total"] < 0 or
            type(result.get("offset")) is not int or result["offset"] != len(rows) or
            type(result.get("returned")) is not int or
            result.get("returned") != len(result["rows"]) or
            result["returned"] > PAGE_SIZE or type(result.get("truncated")) is not bool):
            raise ValueError("Unexpected query response; no partial result was printed.")
        if result["total"] > MAX_ROWS:
            raise ValueError("More than 2500 rows match. Narrow filters or review the bound.")
        applied = result.get("applied_filters")
        if not isinstance(applied, dict):
            raise ValueError("Applied filters are missing.")
        for name, value in FILTERS.items():
            if name not in applied or text_value(applied[name]) != text_value(value):
                raise ValueError("A requested filter was not applied: " + name)
        provenance = result.get("provenance")
        imported = provenance.get("imported_at") if isinstance(provenance, dict) else None
        if not isinstance(imported, str):
            raise ValueError("Import date is missing; no partial result was printed.")
        datetime.fromisoformat(imported.replace("Z", "+00:00"))
        if first is None:
            first, total, imported_at = result, result["total"], imported
        if result["total"] != total or imported != imported_at:
            raise ValueError("The import or match count changed. Run the request again.")
        next_offset = len(rows) + result["returned"]
        if (next_offset > total or result["truncated"] != (next_offset < total) or
            (result["returned"] == 0 and next_offset < total)):
            raise ValueError("Pagination stopped making sense; no partial result was printed.")
        rows.extend(result["rows"])
        if len(rows) == total:
            return dict(dataset=DATASET, filters=FILTERS, total=total, rows=rows,
                        provenance=first["provenance"], licence=first.get("licence"),
                        attribution=first.get("attribution"))
    raise ValueError("Page limit reached. Narrow filters; no partial result was printed.")

try:
    result = retrieve()
    if result["total"] == 0:
        print("No matching records. Check the selected area and filters.", file=sys.stderr)
    # Keep provenance and attribution with cached or republished data.
    print(json.dumps(result, indent=2))
except HTTPError as error:
    print("Query failed: HTTP " + str(error.code) + ". Check filters and access; retry later for rate limits or server errors.", file=sys.stderr)
    sys.exit(1)
except (URLError, TimeoutError, ValueError, TypeError, KeyError) as error:
    print(str(error), file=sys.stderr)
    sys.exit(1)