feat(sidecar): importer creates Source/Post/ImageProvenance from sidecars (idempotent, fail-soft)
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
This commit is contained in:
@@ -0,0 +1,164 @@
|
||||
"""FC-2d-v: sidecar → Source/Post/ImageProvenance (sync importer)."""
|
||||
|
||||
import json
|
||||
from pathlib import Path
|
||||
|
||||
import pytest
|
||||
from PIL import Image
|
||||
from sqlalchemy import func, select
|
||||
|
||||
from backend.app.models import (
|
||||
ImageProvenance,
|
||||
ImageRecord,
|
||||
ImportSettings,
|
||||
Post,
|
||||
Source,
|
||||
)
|
||||
from backend.app.services.importer import Importer
|
||||
from backend.app.services.thumbnailer import Thumbnailer
|
||||
|
||||
pytestmark = pytest.mark.integration
|
||||
|
||||
|
||||
@pytest.fixture
|
||||
def import_layout(tmp_path):
|
||||
import_root = tmp_path / "import"
|
||||
images_root = tmp_path / "images"
|
||||
import_root.mkdir()
|
||||
images_root.mkdir()
|
||||
return import_root, images_root
|
||||
|
||||
|
||||
@pytest.fixture
|
||||
def importer(db_sync, import_layout):
|
||||
import_root, images_root = import_layout
|
||||
settings = db_sync.execute(
|
||||
select(ImportSettings).where(ImportSettings.id == 1)
|
||||
).scalar_one()
|
||||
return Importer(
|
||||
session=db_sync,
|
||||
images_root=images_root,
|
||||
import_root=import_root,
|
||||
thumbnailer=Thumbnailer(images_root=images_root),
|
||||
settings=settings,
|
||||
)
|
||||
|
||||
|
||||
def _split(path: Path, orient, size=(256, 256)):
|
||||
"""Structured image (half/half) — solid colors phash-collapse, which
|
||||
would route a 2nd image to the supersede path (no sidecar applied)."""
|
||||
path.parent.mkdir(parents=True, exist_ok=True)
|
||||
w, h = size
|
||||
im = Image.new("L", size, 0)
|
||||
px = im.load()
|
||||
for y in range(h):
|
||||
for x in range(w):
|
||||
if (x / w if orient == "v" else y / h) >= 0.5:
|
||||
px[x, y] = 255
|
||||
im.convert("RGB").save(path, "JPEG")
|
||||
|
||||
|
||||
def _sidecar(media: Path, payload: dict):
|
||||
media.with_suffix(".json").write_text(json.dumps(payload))
|
||||
|
||||
|
||||
def test_sidecar_creates_provenance(importer, import_layout):
|
||||
import_root, _ = import_layout
|
||||
m = import_root / "Alice" / "a.jpg"
|
||||
_split(m, "v")
|
||||
_sidecar(m, {
|
||||
"category": "patreon", "id": 555,
|
||||
"url": "https://patreon.com/posts/555", "title": "Set 1",
|
||||
"content": "<p>hi</p>", "page_count": 2,
|
||||
"published_at": "2023-08-01T00:00:00Z",
|
||||
})
|
||||
r = importer.import_one(m)
|
||||
assert r.status == "imported"
|
||||
rec = importer.session.get(ImageRecord, r.image_id)
|
||||
src = importer.session.execute(select(Source)).scalar_one()
|
||||
post = importer.session.execute(select(Post)).scalar_one()
|
||||
assert src.platform == "patreon"
|
||||
assert post.external_post_id == "555"
|
||||
assert post.post_url == "https://patreon.com/posts/555"
|
||||
assert post.post_title == "Set 1"
|
||||
assert post.description == "<p>hi</p>"
|
||||
assert post.attachment_count == 2
|
||||
assert post.post_date is not None
|
||||
assert post.raw_metadata["id"] == 555
|
||||
prov = importer.session.execute(select(ImageProvenance)).scalar_one()
|
||||
assert prov.image_record_id == rec.id and prov.post_id == post.id
|
||||
assert rec.primary_post_id == post.id
|
||||
|
||||
|
||||
def test_reimport_same_post_idempotent(importer, import_layout):
|
||||
import_root, _ = import_layout
|
||||
payload = {"category": "patreon", "id": 777, "title": "P"}
|
||||
m1 = import_root / "Bob" / "p1.jpg"
|
||||
_split(m1, "v")
|
||||
_sidecar(m1, payload)
|
||||
importer.import_one(m1)
|
||||
m2 = import_root / "Bob" / "p2.jpg"
|
||||
_split(m2, "h") # structurally distinct → not a phash dup
|
||||
_sidecar(m2, payload)
|
||||
r2 = importer.import_one(m2)
|
||||
assert r2.status == "imported"
|
||||
assert importer.session.execute(
|
||||
select(func.count()).select_from(Source)
|
||||
).scalar_one() == 1
|
||||
assert importer.session.execute(
|
||||
select(func.count()).select_from(Post)
|
||||
).scalar_one() == 1
|
||||
assert importer.session.execute(
|
||||
select(func.count()).select_from(ImageProvenance)
|
||||
).scalar_one() == 2
|
||||
|
||||
|
||||
def test_garbage_sidecar_still_imports(importer, import_layout):
|
||||
import_root, _ = import_layout
|
||||
m = import_root / "Carol" / "c.jpg"
|
||||
_split(m, "v")
|
||||
m.with_suffix(".json").write_text("{ not json")
|
||||
r = importer.import_one(m)
|
||||
assert r.status == "imported"
|
||||
assert importer.session.execute(
|
||||
select(func.count()).select_from(Post)
|
||||
).scalar_one() == 0
|
||||
|
||||
|
||||
def test_no_sidecar_unchanged(importer, import_layout):
|
||||
import_root, _ = import_layout
|
||||
m = import_root / "Dave" / "d.jpg"
|
||||
_split(m, "v")
|
||||
r = importer.import_one(m)
|
||||
assert r.status == "imported"
|
||||
assert importer.session.execute(
|
||||
select(func.count()).select_from(Post)
|
||||
).scalar_one() == 0
|
||||
|
||||
|
||||
def test_no_artist_anywhere_skips_provenance(importer, import_layout):
|
||||
import_root, _ = import_layout
|
||||
m = import_root / "rootfile.jpg" # no top-level artist folder
|
||||
_split(m, "v")
|
||||
_sidecar(m, {"category": "x", "id": 1}) # no artist key
|
||||
r = importer.import_one(m)
|
||||
assert r.status == "imported"
|
||||
assert importer.session.execute(
|
||||
select(func.count()).select_from(Source)
|
||||
).scalar_one() == 0
|
||||
|
||||
|
||||
def test_sidecar_artist_used_when_no_folder_artist(importer, import_layout):
|
||||
from backend.app.models import Artist
|
||||
|
||||
import_root, _ = import_layout
|
||||
m = import_root / "e.jpg" # root → no folder artist
|
||||
_split(m, "v")
|
||||
_sidecar(m, {"category": "pixiv", "id": 9, "artist": "Yuki"})
|
||||
r = importer.import_one(m)
|
||||
assert r.status == "imported"
|
||||
a = importer.session.execute(
|
||||
select(Artist).where(Artist.slug == "yuki")
|
||||
).scalar_one()
|
||||
src = importer.session.execute(select(Source)).scalar_one()
|
||||
assert src.artist_id == a.id
|
||||
Reference in New Issue
Block a user