from zipfile import ZipFile

from pipeline.download.crime import extract_csvs, parse_archives


def test_parse_archives_reads_monthly_zip_links_only():
    html = """
    <p><a href="/data/archive/latest.zip">latest.zip</a></p>
    <div class="archive crime">
      <div class="download">
        <i class="icon-file"></i> <span><a href="/data/archive/2026-03.zip">March 2026</a> (1.6&nbsp;GB)</span>
        <p class="contained-range">Contains data from Apr 2023 to Mar 2026</p>
        <p class="md5sum">6dde462489389445877f3988ef3f4f4b</p>
      </div>
      <div class="download">
        <i class="icon-file"></i> <span><a href="/data/archive/2019-06.zip">June 2019</a> (1.6&nbsp;GB)</span>
        <p class="contained-range">Contains data from Jul 2016 to Jun 2019</p>
        <p class="md5sum">d6494297b24c1434bdb2504e95261bf8-100</p>
      </div>
    </div>
    <div class="archive neighbourhood">
      <div class="download">
        <span><a href="/data/neighbourhood.zip">Neighbourhood crime</a> (2.2 MB)</span>
        <small class="md5sum">6b80e2b97d87f6668b7a45953924d191</small>
      </div>
    </div>
    """

    archives = parse_archives(html, "https://data.police.uk/data/archive/")

    assert [archive.filename for archive in archives] == [
        "2026-03.zip",
        "2019-06.zip",
    ]
    assert archives[0].url == "https://data.police.uk/data/archive/2026-03.zip"
    assert archives[0].md5 == "6dde462489389445877f3988ef3f4f4b"
    assert archives[1].md5 is None
    assert archives[1].raw_md5 == "d6494297b24c1434bdb2504e95261bf8-100"


def test_extract_csvs_preserves_existing_newer_files(tmp_path):
    zip_path = tmp_path / "older.zip"
    output = tmp_path / "crime"
    existing = output / "2023-01" / "2023-01-city-street.csv"
    existing.parent.mkdir(parents=True)
    existing.write_text("newer\n")

    with ZipFile(zip_path, "w") as archive:
        archive.writestr("2023-01/2023-01-city-street.csv", "older\n")
        archive.writestr("2022-12/2022-12-city-street.csv", "old\n")
        archive.writestr("../escape.csv", "bad\n")
        archive.writestr("notes.txt", "ignored\n")

    extracted, skipped = extract_csvs(zip_path, output)

    assert extracted == 1
    assert skipped == 1
    assert existing.read_text() == "newer\n"
    assert (output / "2022-12" / "2022-12-city-street.csv").read_text() == "old\n"
    assert not (tmp_path / "escape.csv").exists()