fc5: tag_apply — joins manifest entries to ImageRecord by sha256 after filesystem scan

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

tests/test_tag_apply.py

@@ -0,0 +1,156 @@
+"""FC-5: tag_apply unit tests."""
+import json
+
+import pytest
+from sqlalchemy import select
+
+from backend.app.models import Artist, ImageRecord, SeriesPage, Tag, TagKind, image_tag
+from backend.app.services.migrators import tag_apply
+
+pytestmark = pytest.mark.integration
+
+
+async def _seed_image(db, sha: str, *, suffix="x"):
+    img = ImageRecord(
+        path=f"/images/imported/aa/{suffix}.jpg",
+        sha256=sha, size_bytes=10, mime="image/jpeg",
+        width=10, height=10, origin="imported_filesystem",
+    )
+    db.add(img)
+    await db.flush()
+    return img
+
+
+def _write_manifest(tmp_path, *, artist_assignments=None, tag_associations=None, series_pages=None):
+    d = tmp_path / "_migration_state"
+    d.mkdir(parents=True, exist_ok=True)
+    (d / "ir_tag_manifest.json").write_text(json.dumps({
+        "schema_version": 1,
+        "image_artist_assignments": artist_assignments or [],
+        "image_tag_associations": tag_associations or [],
+        "series_pages": series_pages or [],
+    }))
+
+
+@pytest.mark.asyncio
+async def test_artist_assignment_sets_image_artist_id(db, tmp_path):
+    sha = "a" * 64
+    await _seed_image(db, sha)
+    await db.commit()
+
+    _write_manifest(tmp_path, artist_assignments=[
+        {"sha256": sha, "artist_name": "BobAuthor"},
+    ])
+
+    result = await tag_apply.apply_async(db, images_root=tmp_path, dry_run=False)
+    assert result["counts"]["rows_inserted"] >= 1
+
+    img = (await db.execute(
+        select(ImageRecord).where(ImageRecord.sha256 == sha)
+    )).scalar_one()
+    artist = (await db.execute(
+        select(Artist).where(Artist.id == img.artist_id)
+    )).scalar_one()
+    assert artist.name == "BobAuthor"
+
+
+@pytest.mark.asyncio
+async def test_tag_association_creates_image_tag(db, tmp_path):
+    sha = "b" * 64
+    await _seed_image(db, sha, suffix="b")
+    # Seed the tag itself (ir_ingest would have created it).
+    tag = Tag(name="blue_eyes", kind=TagKind.general)
+    db.add(tag)
+    await db.commit()
+
+    _write_manifest(tmp_path, tag_associations=[
+        {"sha256": sha, "tag_name": "blue_eyes", "tag_kind": "general", "fandom_name": None},
+    ])
+
+    await tag_apply.apply_async(db, images_root=tmp_path, dry_run=False)
+
+    img_id = (await db.execute(
+        select(ImageRecord.id).where(ImageRecord.sha256 == sha)
+    )).scalar_one()
+    row = (await db.execute(
+        select(image_tag.c.tag_id).where(image_tag.c.image_record_id == img_id)
+    )).all()
+    assert len(row) == 1
+
+
+@pytest.mark.asyncio
+async def test_unmatched_sha_appears_in_unmatched_list(db, tmp_path):
+    _write_manifest(tmp_path, tag_associations=[
+        {"sha256": "z" * 64, "tag_name": "missing_image", "tag_kind": "general", "fandom_name": None},
+    ])
+    result = await tag_apply.apply_async(db, images_root=tmp_path, dry_run=False)
+    assert any(e["sha256"] == "z" * 64 for e in result["unmatched"])
+
+
+@pytest.mark.asyncio
+async def test_series_page_created(db, tmp_path):
+    sha = "c" * 64
+    await _seed_image(db, sha, suffix="c")
+    series_tag = Tag(name="MySeries", kind=TagKind.series)
+    db.add(series_tag)
+    await db.commit()
+
+    _write_manifest(tmp_path, series_pages=[
+        {"sha256": sha, "series_tag_name": "MySeries", "page_number": 1},
+    ])
+
+    await tag_apply.apply_async(db, images_root=tmp_path, dry_run=False)
+
+    img_id = (await db.execute(
+        select(ImageRecord.id).where(ImageRecord.sha256 == sha)
+    )).scalar_one()
+    sp = (await db.execute(
+        select(SeriesPage).where(SeriesPage.image_id == img_id)
+    )).scalar_one()
+    assert sp.page_number == 1
+
+
+@pytest.mark.asyncio
+async def test_idempotent_on_rerun(db, tmp_path):
+    sha = "d" * 64
+    await _seed_image(db, sha, suffix="d")
+    tag = Tag(name="rerun_tag", kind=TagKind.general)
+    db.add(tag)
+    await db.commit()
+
+    _write_manifest(tmp_path, tag_associations=[
+        {"sha256": sha, "tag_name": "rerun_tag", "tag_kind": "general", "fandom_name": None},
+    ])
+
+    await tag_apply.apply_async(db, images_root=tmp_path, dry_run=False)
+    result2 = await tag_apply.apply_async(db, images_root=tmp_path, dry_run=False)
+    assert result2["counts"]["rows_skipped"] >= 1
+
+
+@pytest.mark.asyncio
+async def test_missing_manifest_raises(db, tmp_path):
+    with pytest.raises(FileNotFoundError):
+        await tag_apply.apply_async(db, images_root=tmp_path, dry_run=False)
+
+
+@pytest.mark.asyncio
+async def test_dry_run_makes_no_changes(db, tmp_path):
+    sha = "e" * 64
+    await _seed_image(db, sha, suffix="e")
+    tag = Tag(name="dry_tag", kind=TagKind.general)
+    db.add(tag)
+    await db.commit()
+
+    _write_manifest(tmp_path, tag_associations=[
+        {"sha256": sha, "tag_name": "dry_tag", "tag_kind": "general", "fandom_name": None},
+    ])
+
+    await tag_apply.apply_async(db, images_root=tmp_path, dry_run=True)
+
+    img_id = (await db.execute(
+        select(ImageRecord.id).where(ImageRecord.sha256 == sha)
+    )).scalar_one()
+    rows = (await db.execute(
+        select(image_tag.c.tag_id).where(image_tag.c.image_record_id == img_id)
+    )).all()
+    assert len(rows) == 0