from pathlib import Path

from rf4_research.media_assets import extract_media_candidates, merge_manifest, store_asset


def test_extracts_and_classifies_unique_https_media() -> None:
    html = '''<article><a href="/fish/pike"><img src="/media/fish/pike.png" alt="Щука"></a></article>
    <figure><img data-src="https://cdn.example/maps/kuori.webp" alt="Карта Куори"></figure>
    <img src="http://unsafe.example/bait.png"><img src="/media/fish/pike.png">'''
    items = extract_media_candidates(html, source_page="https://example.test/guide")
    assert [(item.entity_type, item.asset_url) for item in items] == [
        ("fish", "https://example.test/media/fish/pike.png"),
        ("waterbody", "https://cdn.example/maps/kuori.webp"),
    ]
    assert items[0].external_id == "pike"


def test_does_not_mistake_seafishing_screenshot_for_fish_entity() -> None:
    items = extract_media_candidates('<img src="/res/userguide/seafishing-1.jpeg">', source_page="https://rf4game.de/userguide/")
    assert items[0].entity_type == "reference"


def test_manifest_merges_and_binary_store_is_content_addressed(tmp_path: Path) -> None:
    item = extract_media_candidates('<img src="/bait/worm.png" alt="Bait worm">', source_page="https://example.test")[0]
    first = merge_manifest(tmp_path / "manifest.json", [item])
    second = merge_manifest(tmp_path / "manifest.json", [item])
    assert len(first["assets"]) == len(second["assets"]) == 1
    digest, relative = store_asset(tmp_path, b"image", content_type="image/png", source_url=item.asset_url)
    assert len(digest) == 64
    assert (tmp_path / relative).read_bytes() == b"image"