FabledCurator/tests/test_importer_attach_in_place.py

"""attach_in_place — Importer seam used by FC-3c (downloader)."""

from pathlib import Path

import pytest
from sqlalchemy import func, select

from backend.app.models import Artist, ImageRecord, ImportSettings
from backend.app.services.importer import Importer
from backend.app.services.thumbnailer import Thumbnailer
from backend.app.utils import safe_probe
from backend.app.utils.safe_probe import ProbeResult

pytestmark = pytest.mark.integration


def _make_jpg(path: Path, color=(255, 0, 0), size=(64, 64)) -> bytes:
    from PIL import Image
    path.parent.mkdir(parents=True, exist_ok=True)
    Image.new("RGB", size, color).save(path, "JPEG")
    return path.read_bytes()


@pytest.fixture
def importer(db_sync, tmp_path):
    settings = db_sync.execute(
        select(ImportSettings).where(ImportSettings.id == 1)
    ).scalar_one()
    images_root = tmp_path / "images"
    images_root.mkdir()
    return Importer(
        session=db_sync,
        images_root=images_root,
        import_root=tmp_path / "ignored",
        thumbnailer=Thumbnailer(images_root=images_root),
        settings=settings,
    )


def test_attach_in_place_imports_happy(importer, db_sync):
    images_root = importer.images_root
    artist = Artist(name="Alice", slug="alice")
    db_sync.add(artist)
    db_sync.flush()

    target = images_root / "alice" / "patreon" / "post1" / "img.jpg"
    _make_jpg(target)

    result = importer.attach_in_place(target, artist=artist)
    assert result.status == "imported"
    assert result.image_id is not None

    rec_path = db_sync.execute(
        select(ImageRecord.path, ImageRecord.origin, ImageRecord.artist_id)
        .where(ImageRecord.id == result.image_id)
    ).one()
    assert rec_path.path == str(target)
    assert rec_path.origin == "downloaded"
    assert rec_path.artist_id == artist.id


def test_attach_in_place_with_sidecar(importer, db_sync):
    images_root = importer.images_root
    artist = Artist(name="Bob", slug="bob")
    db_sync.add(artist)
    db_sync.flush()

    target = images_root / "bob" / "patreon" / "post1" / "img.jpg"
    _make_jpg(target)
    sidecar = target.with_suffix(target.suffix + ".json")
    sidecar.write_text(
        '{"category": "patreon", "id": "12345", "url": "https://patreon.com/post/12345", '
        '"title": "Test Post", "published_at": "2026-05-01T00:00:00+00:00"}'
    )

    result = importer.attach_in_place(target, artist=artist)
    assert result.status == "imported"

    primary_post_id = db_sync.execute(
        select(ImageRecord.primary_post_id).where(ImageRecord.id == result.image_id)
    ).scalar_one()
    assert primary_post_id is not None  # post created from sidecar


def test_attach_in_place_sha256_dedup_does_not_modify_disk(importer, db_sync):
    images_root = importer.images_root
    artist = Artist(name="Cara", slug="cara")
    db_sync.add(artist)
    db_sync.flush()

    first = images_root / "cara" / "p" / "a.jpg"
    bytes_ = _make_jpg(first)
    importer.attach_in_place(first, artist=artist)

    second = images_root / "cara" / "p" / "b.jpg"
    second.parent.mkdir(parents=True, exist_ok=True)
    second.write_bytes(bytes_)
    assert second.exists()

    result = importer.attach_in_place(second, artist=artist)
    assert result.status == "skipped"
    assert result.skip_reason.value == "duplicate_hash"
    # Caller is responsible for deleting the on-disk file; we don't touch it.
    assert second.exists()


def test_attach_in_place_supersede_keeps_file_at_new_path(importer, db_sync):
    """phash near-dup where the new file is LARGER → existing record
    swaps to the new path; old file deleted."""
    images_root = importer.images_root
    artist = Artist(name="Dee", slug="dee")
    db_sync.add(artist)
    db_sync.flush()

    smaller = images_root / "dee" / "p" / "small.jpg"
    _make_jpg(smaller, color=(50, 50, 200), size=(64, 64))
    first = importer.attach_in_place(smaller, artist=artist)
    assert first.status == "imported"

    bigger = images_root / "dee" / "p" / "big.jpg"
    _make_jpg(bigger, color=(50, 50, 200), size=(256, 256))

    result = importer.attach_in_place(bigger, artist=artist)
    assert result.status == "superseded"

    new_path = db_sync.execute(
        select(ImageRecord.path).where(ImageRecord.id == first.image_id)
    ).scalar_one()
    assert new_path == str(bigger)
    assert not smaller.exists()
    assert bigger.exists()


def test_attach_in_place_phash_smaller_skips_without_modifying_disk(
    importer, db_sync
):
    images_root = importer.images_root
    artist = Artist(name="Eve", slug="eve")
    db_sync.add(artist)
    db_sync.flush()

    bigger = images_root / "eve" / "p" / "big.jpg"
    _make_jpg(bigger, color=(0, 200, 0), size=(256, 256))
    importer.attach_in_place(bigger, artist=artist)

    smaller = images_root / "eve" / "p" / "small.jpg"
    _make_jpg(smaller, color=(0, 200, 0), size=(64, 64))

    result = importer.attach_in_place(smaller, artist=artist)
    assert result.status == "skipped"
    assert result.skip_reason.value == "duplicate_phash"
    assert smaller.exists()


def test_attach_in_place_duplicate_image_links_both_posts(importer, db_sync):
    """A byte-identical image appearing in a SECOND post must link to BOTH posts
    (enrich-on-duplicate) so it shows on every post — not be silently dropped,
    which left the second post a bare shell. Operator-flagged 2026-06-08."""
    from backend.app.models import ImageProvenance, Post

    images_root = importer.images_root
    artist = Artist(name="Hana", slug="hana")
    db_sync.add(artist)
    db_sync.flush()

    first = images_root / "hana" / "patreon" / "post_a" / "img.jpg"
    bytes_ = _make_jpg(first, color=(120, 40, 90))
    first.with_suffix(first.suffix + ".json").write_text(
        '{"category": "patreon", "id": "111", "title": "Post 111"}'
    )
    r1 = importer.attach_in_place(first, artist=artist)
    assert r1.status == "imported"

    second = images_root / "hana" / "patreon" / "post_b" / "img.jpg"
    second.parent.mkdir(parents=True, exist_ok=True)
    second.write_bytes(bytes_)  # identical bytes → sha256 duplicate
    second.with_suffix(second.suffix + ".json").write_text(
        '{"category": "patreon", "id": "222", "title": "Post 222"}'
    )
    r2 = importer.attach_in_place(second, artist=artist)
    assert r2.status == "skipped"
    assert r2.skip_reason.value == "duplicate_hash"
    assert r2.image_id == r1.image_id

    linked = db_sync.execute(
        select(Post.external_post_id)
        .join(ImageProvenance, ImageProvenance.post_id == Post.id)
        .where(ImageProvenance.image_record_id == r1.image_id)
    ).scalars().all()
    assert set(linked) == {"111", "222"}  # shows on BOTH posts

    primary = db_sync.execute(
        select(Post.external_post_id)
        .join(ImageRecord, ImageRecord.primary_post_id == Post.id)
        .where(ImageRecord.id == r1.image_id)
    ).scalar_one()
    assert primary == "111"  # primary stays on the first post


def test_attach_in_place_duplicate_attachment_links_both_posts(importer, db_sync):
    """The same non-art file attached to two posts gets a row on EACH post over a
    single sha-addressed blob — neither post is left bare. Operator-flagged
    2026-06-08 (1589 empty Anduo shells from globally-unique attachment sha)."""
    from backend.app.models import Post, PostAttachment

    images_root = importer.images_root
    artist = Artist(name="Iris", slug="iris")
    db_sync.add(artist)
    db_sync.flush()

    payload = b"the same readme creator attaches to every post"
    a = images_root / "iris" / "patreon" / "post_a" / "readme.txt"
    a.parent.mkdir(parents=True, exist_ok=True)
    a.write_bytes(payload)
    a.with_suffix(a.suffix + ".json").write_text(
        '{"category": "patreon", "id": "301", "title": "Post 301"}'
    )
    assert importer.attach_in_place(a, artist=artist).status == "attached"

    b = images_root / "iris" / "patreon" / "post_b" / "readme.txt"
    b.parent.mkdir(parents=True, exist_ok=True)
    b.write_bytes(payload)
    b.with_suffix(b.suffix + ".json").write_text(
        '{"category": "patreon", "id": "302", "title": "Post 302"}'
    )
    assert importer.attach_in_place(b, artist=artist).status == "attached"

    linked = db_sync.execute(
        select(Post.external_post_id)
        .join(PostAttachment, PostAttachment.post_id == Post.id)
        .where(PostAttachment.original_filename == "readme.txt")
    ).scalars().all()
    assert set(linked) == {"301", "302"}  # one row per post, both present


def test_attach_in_place_post_first_skips_post_fields(importer, db_sync):
    """Post-first (#856, milestone #67): with post_first set (native path), the
    per-media import links provenance + localization (source_filehash) but does
    NOT write the post body/title — the post-record owns those. Guards the import
    side even if a body sneaks into the sidecar."""
    from backend.app.models import ImageProvenance, Post

    importer.post_first = True
    images_root = importer.images_root
    artist = Artist(name="Pat", slug="pat")
    db_sync.add(artist)
    db_sync.flush()

    target = images_root / "pat" / "patreon" / "post1" / "img.jpg"
    _make_jpg(target)
    cdn = "https://cdn.patreon.com/" + "a" * 32 + "/x.jpg"
    target.with_suffix(target.suffix + ".json").write_text(
        '{"category": "patreon", "id": "900", "title": "Should NOT apply", '
        '"content": "<p>body should NOT apply</p>", '
        f'"source_url": "{cdn}"}}'
    )

    result = importer.attach_in_place(target, artist=artist)
    assert result.status == "imported"

    # Image-specific work STILL happens: localization filehash + provenance link.
    rec = db_sync.execute(
        select(ImageRecord).where(ImageRecord.id == result.image_id)
    ).scalar_one()
    assert rec.source_filehash == "a" * 32
    assert rec.primary_post_id is not None

    post = db_sync.get(Post, rec.primary_post_id)
    assert post.external_post_id == "900"
    # Post-first: body/title NOT written by the media path (post-record owns them).
    assert post.description is None
    assert post.post_title is None
    # Provenance link exists regardless of post_first.
    prov = db_sync.execute(
        select(ImageProvenance).where(
            ImageProvenance.image_record_id == result.image_id
        )
    ).scalars().all()
    assert len(prov) == 1


def test_attach_in_place_post_first_false_applies_post_fields(importer, db_sync):
    """Default (gallery-dl path): the sidecar IS the body source, so post fields
    ARE applied. Guards that the post_first flag actually gates this — not a no-op."""
    from backend.app.models import Post

    assert importer.post_first is False  # default
    images_root = importer.images_root
    artist = Artist(name="Gal", slug="gal")
    db_sync.add(artist)
    db_sync.flush()

    target = images_root / "gal" / "patreon" / "post1" / "img.jpg"
    _make_jpg(target)
    target.with_suffix(target.suffix + ".json").write_text(
        '{"category": "patreon", "id": "901", "title": "Applies", '
        '"content": "<p>body applies</p>"}'
    )

    result = importer.attach_in_place(target, artist=artist)
    assert result.status == "imported"
    rec = db_sync.get(ImageRecord, result.image_id)
    post = db_sync.get(Post, rec.primary_post_id)
    assert post.description == "<p>body applies</p>"
    assert post.post_title == "Applies"


def _fake_video_probe(monkeypatch, dims):
    """Patch safe_probe.probe_video to return controlled (w,h,duration) keyed by
    filename — so video dedup (#871) can be tested without real mp4s/ffprobe."""
    def _probe(path, **_kw):
        w, h, d = dims[Path(path).name]
        return ProbeResult(ok=True, width=w, height=h, duration=d)
    monkeypatch.setattr(safe_probe, "probe_video", _probe)


def _mkvideo(images_root, rel, content):
    p = images_root / rel
    p.parent.mkdir(parents=True, exist_ok=True)
    p.write_bytes(content)
    return p


def test_attach_in_place_video_dedup_skips_smaller(importer, db_sync, monkeypatch):
    """#871 Tier-1: a same-artist video with matching duration+aspect but smaller
    resolution is a near-dup of the existing larger video → skipped + linked, not
    re-imported (the #859 'same video from multiple sources' case)."""
    images_root = importer.images_root
    artist = Artist(name="Vid", slug="vid")
    db_sync.add(artist)
    db_sync.flush()
    _fake_video_probe(monkeypatch, {
        "a.mp4": (1920, 1080, 300.0),
        "b.mp4": (1280, 720, 300.4),  # within 1s, same 16:9 aspect, smaller
    })

    big = _mkvideo(images_root, "vid/patreon/p/a.mp4", b"AAAA")
    r1 = importer.attach_in_place(big, artist=artist)
    assert r1.status == "imported"

    small = _mkvideo(images_root, "vid/patreon/p2/b.mp4", b"BBBB")
    r2 = importer.attach_in_place(small, artist=artist)
    assert r2.status == "skipped"
    assert r2.skip_reason.value == "duplicate_phash"
    assert r2.image_id == r1.image_id
    assert db_sync.execute(
        select(func.count()).select_from(ImageRecord)
    ).scalar_one() == 1


def test_attach_in_place_video_dedup_supersedes_larger(importer, db_sync, monkeypatch):
    """A higher-res re-download supersedes the smaller existing video, keeping the
    row id + adopting the new duration."""
    images_root = importer.images_root
    artist = Artist(name="Vid2", slug="vid2")
    db_sync.add(artist)
    db_sync.flush()
    _fake_video_probe(monkeypatch, {
        "s.mp4": (640, 360, 120.0),
        "l.mp4": (1920, 1080, 120.2),  # within 1s, same aspect, larger
    })

    small = _mkvideo(images_root, "vid2/p/s.mp4", b"SSSS")
    r1 = importer.attach_in_place(small, artist=artist)
    assert r1.status == "imported"

    big = _mkvideo(images_root, "vid2/p/l.mp4", b"LLLL")
    r2 = importer.attach_in_place(big, artist=artist)
    assert r2.status == "superseded"
    assert r2.image_id == r1.image_id

    db_sync.expire_all()
    rec = db_sync.get(ImageRecord, r1.image_id)
    assert rec.width == 1920
    assert rec.duration_seconds == 120.2
    assert db_sync.execute(
        select(func.count()).select_from(ImageRecord)
    ).scalar_one() == 1


def test_attach_in_place_video_distinct_durations_not_merged(importer, db_sync, monkeypatch):
    """Guard against false-merge: two videos with clearly different durations stay
    separate even at the same artist + resolution."""
    images_root = importer.images_root
    artist = Artist(name="Vid3", slug="vid3")
    db_sync.add(artist)
    db_sync.flush()
    _fake_video_probe(monkeypatch, {
        "a.mp4": (1920, 1080, 60.0),
        "b.mp4": (1920, 1080, 200.0),  # 140s apart — different content
    })

    a = _mkvideo(images_root, "vid3/p/a.mp4", b"AAAA")
    b = _mkvideo(images_root, "vid3/p/b.mp4", b"BBBB")
    assert importer.attach_in_place(a, artist=artist).status == "imported"
    assert importer.attach_in_place(b, artist=artist).status == "imported"
    assert db_sync.execute(
        select(func.count()).select_from(ImageRecord)
    ).scalar_one() == 2


def test_attach_in_place_invalid_image_returns_skipped(importer):
    images_root = importer.images_root
    bad = images_root / "fred" / "bad.jpg"
    bad.parent.mkdir(parents=True, exist_ok=True)
    bad.write_bytes(b"not a jpeg, just some bytes")

    result = importer.attach_in_place(bad)
    assert result.status == "skipped"
    assert result.skip_reason.value == "invalid_image"


def test_attach_in_place_non_media_routes_to_attachment(importer, db_sync):
    """FC-2d-iii dispatch parity for the download path. A non-media file
    (.pdf, .txt etc.) downloaded by gallery-dl must become a PostAttachment,
    not bounce back as `skipped+invalid_image` — the latter flipped
    otherwise-successful runs to status="error" downstream. Operator-flagged
    2026-06-02 (Lustria patreon OST zip)."""
    from backend.app.models import PostAttachment

    images_root = importer.images_root
    artist = Artist(name="Gus", slug="gus")
    db_sync.add(artist)
    db_sync.flush()

    txt = images_root / "gus" / "patreon" / "post" / "notes.txt"
    txt.parent.mkdir(parents=True, exist_ok=True)
    txt.write_bytes(b"some non-media payload that the importer should preserve")

    result = importer.attach_in_place(txt, artist=artist)
    assert result.status == "attached"

    row = db_sync.execute(
        select(PostAttachment).where(PostAttachment.original_filename == "notes.txt")
    ).scalar_one()
    assert row.ext == ".txt"
    assert row.artist_id == artist.id


def test_a_non_media_file_lands_on_the_source_being_downloaded(importer, db_sync):
    """#4435: a Discord artist has one source per channel. A non-media file
    downloaded for the SECOND channel was filed under the artist's first
    Discord source (the (artist, platform) lookup takes the lowest id), as an
    undated shell post; the real post record then made a second post under
    the right source. The attachment must follow the source it was fetched for."""
    from backend.app.models import Post, PostAttachment, Source

    images_root = importer.images_root
    artist = Artist(name="Yara", slug="yara")
    db_sync.add(artist)
    db_sync.flush()
    first = Source(
        artist_id=artist.id, platform="discord",
        url="https://discord.com/channels/1/100",
    )
    second = Source(
        artist_id=artist.id, platform="discord",
        url="https://discord.com/channels/1/200",
    )
    db_sync.add_all([first, second])
    db_sync.flush()

    rar = images_root / "yara" / "discord" / "rewards" / "20241226_555_01_pack.rar"
    rar.parent.mkdir(parents=True, exist_ok=True)
    rar.write_bytes(b"not a real archive, so it is kept as an attachment")
    rar.with_suffix(rar.suffix + ".json").write_text(
        '{"category": "discord", "id": "555", "message_id": "555"}'
    )

    result = importer.attach_in_place(rar, artist=artist, source=second)
    assert result.status == "attached"

    owner = db_sync.execute(
        select(Post.source_id, Post.external_post_id)
        .join(PostAttachment, PostAttachment.post_id == Post.id)
        .where(PostAttachment.original_filename == rar.name)
    ).one()
    assert owner.source_id == second.id
    assert owner.external_post_id == "555"
    assert db_sync.execute(
        select(func.count()).select_from(Post).where(Post.source_id == first.id)
    ).scalar_one() == 0