Model/backend/address2UPRN/tests/test_csv.py

# tests/test_address_to_uprn_csv.py

import csv
import json
import pytest
from datetime import date
from pathlib import Path
from backend.address2UPRN.main import get_uprn

FIXTURE_PATH = Path(__file__).parent / "test_data.csv"
SIDECAR_PATH = Path(__file__).parent / "test_lodgement_dates.json"
NEW_API_CUTOFF = date(2012, 1, 1)


def _load_sidecar() -> dict:
    if SIDECAR_PATH.exists():
        return json.loads(SIDECAR_PATH.read_text())
    return {}


def load_test_cases():
    sidecar = _load_sidecar()
    with open(FIXTURE_PATH, newline="", encoding="utf-8") as f:
        reader = csv.DictReader(f)
        cases = []
        for row in reader:
            key = f"{row['User Input']}|{row['Postcode']}"
            entry = sidecar.get(key, {})
            lodgement_date = entry.get("lodgement_date")

            marks = []
            if lodgement_date:
                parsed = date.fromisoformat(lodgement_date[:10])
                if parsed < NEW_API_CUTOFF:
                    marks.append(
                        pytest.mark.xfail(
                            reason=f"EPC lodged {lodgement_date} — predates new API coverage (Jan 2012)",
                            strict=False,
                        )
                    )

            cases.append(
                pytest.param(
                    row["User Input"],
                    row["Postcode"],
                    row["Manual UPRN Code"],
                    id=f'{row["User Input"]} [{row["Postcode"]}]',
                    marks=marks,
                )
            )
        return cases


@pytest.mark.parametrize(
    "user_input,postcode,expected_uprn",
    load_test_cases(),
)
def test_uprn_resolution_matches_manual(
    user_input: str,
    postcode: str,
    expected_uprn: str,
):
    uprn = get_uprn(user_input, postcode)
    if uprn:
        assert uprn == expected_uprn
    else:
        assert str(uprn) == expected_uprn