import csv
import io
import zipfile

from pipeline.download.gias import _CSV_COLUMNS, transform


def _zip_with_rows(rows: list[dict[str, str]]) -> bytes:
    text = io.StringIO()
    writer = csv.DictWriter(text, fieldnames=_CSV_COLUMNS)
    writer.writeheader()
    for row in rows:
        writer.writerow({col: row.get(col, "") for col in _CSV_COLUMNS})
    buffer = io.BytesIO()
    with zipfile.ZipFile(buffer, "w") as archive:
        archive.writestr(
            "edubasealldata20260611.csv",
            text.getvalue().encode("cp1252"),
        )
    return buffer.getvalue()


def _school(name: str, status: str) -> dict[str, str]:
    return {
        "URN": "100000",
        "EstablishmentName": name,
        "TypeOfEstablishment (name)": "Community school",
        "EstablishmentTypeGroup (name)": "Local authority maintained schools",
        "EstablishmentStatus (name)": status,
        "PhaseOfEducation (name)": "Primary",
        "StatutoryLowAge": "4",
        "StatutoryHighAge": "11",
        "Easting": "530000",
        "Northing": "180000",
        "Postcode": "SW1A 1AA",
        "Street": "1 School Lane",
        "Town": "London",
        "LA (name)": "Westminster",
    }


def test_transform_keeps_open_but_proposed_to_close_schools() -> None:
    # "Open, but proposed to close" establishments are operating schools (GIAS
    # can keep the status for years); only closed and proposed-to-open rows are
    # out of scope for the map.
    rows = [
        _school("Open School", "Open"),
        _school("Closing School", "Open, but proposed to close"),
        _school("Closed School", "Closed"),
        _school("Future School", "Proposed to open"),
    ]
    result = transform(_zip_with_rows(rows))

    assert sorted(result["name"].to_list()) == ["Closing School", "Open School"]