fc5: ir_ingest — Tag rows (skip artist/post) + manifest JSON for sha256-keyed tag_apply

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-05-22 09:00:29 -04:00
parent a13a54206e
commit 77b17f7bb6
2 changed files with 270 additions and 0 deletions
@@ -0,0 +1,129 @@
+"""FC-5: ir_ingest unit tests."""
+import json
+from datetime import UTC, datetime
+
+import pytest
+from sqlalchemy import select
+
+from backend.app.models import Tag
+from backend.app.services.migrators import ir_ingest
+
+pytestmark = pytest.mark.integration
+
+
+def _ir_export(tags=None, artist_assignments=None, tag_associations=None, series_pages=None):
+    return {
+        "source_app": "imagerepo",
+        "schema_version": 1,
+        "exported_at": datetime.now(UTC).isoformat(),
+        "tags": tags or [],
+        "image_artist_assignments": artist_assignments or [],
+        "image_tag_associations": tag_associations or [],
+        "series_pages": series_pages or [],
+    }
+
+
+@pytest.mark.asyncio
+async def test_general_tag_created(db, tmp_path):
+    data = _ir_export(tags=[
+        {"name": "blue_eyes", "kind": "general", "fandom_name": None},
+    ])
+    await ir_ingest.migrate_async(db, data=data, images_root=tmp_path, dry_run=False)
+
+    tag = (await db.execute(
+        select(Tag).where(Tag.name == "blue_eyes", Tag.kind == "general")
+    )).scalar_one()
+    assert tag.fandom_id is None
+
+
+@pytest.mark.asyncio
+async def test_character_tag_resolves_fandom_name(db, tmp_path):
+    data = _ir_export(tags=[
+        {"name": "Wonderland", "kind": "fandom", "fandom_name": None},
+        {"name": "Alice", "kind": "character", "fandom_name": "Wonderland"},
+    ])
+    await ir_ingest.migrate_async(db, data=data, images_root=tmp_path, dry_run=False)
+
+    fandom = (await db.execute(
+        select(Tag).where(Tag.name == "Wonderland", Tag.kind == "fandom")
+    )).scalar_one()
+    alice = (await db.execute(
+        select(Tag).where(Tag.name == "Alice", Tag.kind == "character")
+    )).scalar_one()
+    assert alice.fandom_id == fandom.id
+
+
+@pytest.mark.asyncio
+async def test_artist_and_post_kinds_skipped(db, tmp_path):
+    data = _ir_export(tags=[
+        {"name": "BobArtist", "kind": "artist", "fandom_name": None},
+        {"name": "PostXYZ", "kind": "post", "fandom_name": None},
+        {"name": "general_one", "kind": "general", "fandom_name": None},
+    ])
+    counts = await ir_ingest.migrate_async(db, data=data, images_root=tmp_path, dry_run=False)
+    assert counts["rows_skipped"] >= 2
+
+    bob = (await db.execute(
+        select(Tag).where(Tag.name == "BobArtist")
+    )).scalar_one_or_none()
+    post = (await db.execute(
+        select(Tag).where(Tag.name == "PostXYZ")
+    )).scalar_one_or_none()
+    assert bob is None
+    assert post is None
+
+
+@pytest.mark.asyncio
+async def test_manifest_written_to_disk(db, tmp_path):
+    data = _ir_export(
+        artist_assignments=[{"sha256": "a" * 64, "artist_name": "Alice"}],
+        tag_associations=[
+            {"sha256": "a" * 64, "tag_name": "blue_eyes", "tag_kind": "general", "fandom_name": None},
+        ],
+        series_pages=[
+            {"sha256": "a" * 64, "series_tag_name": "MySeries", "page_number": 1},
+        ],
+    )
+    await ir_ingest.migrate_async(db, data=data, images_root=tmp_path, dry_run=False)
+
+    manifest_file = tmp_path / "_migration_state" / "ir_tag_manifest.json"
+    assert manifest_file.exists()
+    manifest = json.loads(manifest_file.read_text())
+    assert manifest["schema_version"] == 1
+    assert len(manifest["image_artist_assignments"]) == 1
+    assert len(manifest["image_tag_associations"]) == 1
+    assert len(manifest["series_pages"]) == 1
+
+
+@pytest.mark.asyncio
+async def test_dry_run_writes_no_manifest_and_no_tags(db, tmp_path):
+    data = _ir_export(
+        tags=[{"name": "dry", "kind": "general", "fandom_name": None}],
+        artist_assignments=[{"sha256": "a" * 64, "artist_name": "DryArtist"}],
+    )
+    await ir_ingest.migrate_async(db, data=data, images_root=tmp_path, dry_run=True)
+
+    tag = (await db.execute(
+        select(Tag).where(Tag.name == "dry")
+    )).scalar_one_or_none()
+    assert tag is None
+    manifest_file = tmp_path / "_migration_state" / "ir_tag_manifest.json"
+    assert not manifest_file.exists()
+
+
+@pytest.mark.asyncio
+async def test_idempotent_on_rerun(db, tmp_path):
+    data = _ir_export(tags=[
+        {"name": "rerun", "kind": "general", "fandom_name": None},
+    ])
+    await ir_ingest.migrate_async(db, data=data, images_root=tmp_path, dry_run=False)
+    counts2 = await ir_ingest.migrate_async(db, data=data, images_root=tmp_path, dry_run=False)
+    assert counts2["rows_skipped"] >= 1
+
+
+@pytest.mark.asyncio
+async def test_rejects_wrong_source_app(db, tmp_path):
+    bad = {"source_app": "elsewhere", "schema_version": 1, "tags": [],
+           "image_artist_assignments": [], "image_tag_associations": [], "series_pages": []}
+    with pytest.raises(ValueError):
+        await ir_ingest.migrate_async(db, data=bad, images_root=tmp_path, dry_run=False)