feat: add asset manifest knowledge ingest

tests/test_alembic.py

@@ -13,11 +13,13 @@ full backing-service compose stack to be online.
 
 from __future__ import annotations
 
-import os
 import sys
 from pathlib import Path
 
 import pytest
+from alembic import command
+from alembic.config import Config
+from alembic.script import ScriptDirectory
 
 ROOT = Path(__file__).resolve().parents[1]
 sys.path.insert(0, str(ROOT))
@@ -29,9 +31,6 @@ def alembic_cfg(tmp_path, monkeypatch):
     db_file = tmp_path / "legacyhub.db"
     monkeypatch.setenv("POSTGRES_HOST", "127.0.0.1")
     monkeypatch.setenv("POSTGRES_PORT", "5432")
-    # Force a fresh Settings + Alembic env that ignores the configured PG.
-    from alembic.config import Config
-
     cfg = Config(str(ROOT / "alembic.ini"))
     cfg.set_main_option("script_location", str(ROOT / "app" / "db" / "migrations"))
     cfg.set_main_option("sqlalchemy.url", f"sqlite:///{db_file}")
@@ -40,11 +39,10 @@ def alembic_cfg(tmp_path, monkeypatch):
 
 def test_migration_offline_emits_sql(alembic_cfg, tmp_path):
     """Offline mode generates SQL for every table; verify ``documents`` appears
-    and at least one JSONB-equivalent column is rendered. SQLite has no JSONB
-    but Alembic's offline mode happily emits the raw DDL for inspection.
+    and the TeamHUB AssetManifest ingest job table is part of the linear head.
+    SQLite has no JSONB but Alembic's offline mode happily emits the raw DDL for
+    inspection.
     """
-    from alembic import command
-
     out_file = tmp_path / "upgrade.sql"
     # ``--sql`` mode bypasses dialect-specific runtime, perfect for a fast check.
     with out_file.open("w", encoding="utf-8") as f:
@@ -59,17 +57,18 @@ def test_migration_offline_emits_sql(alembic_cfg, tmp_path):
     assert "CREATE TABLE documents" in sql
     assert "CREATE TABLE chunks" in sql
     assert "CREATE TABLE processing_events" in sql
+    assert "CREATE TABLE asset_ingest_jobs" in sql
     # Constraint sanity
     assert "uq_chunks_doc_idx" in sql
     assert "uq_pages_doc_page" in sql
+    assert "uq_asset_ingest_identity" in sql
 
 
 def test_revision_history_is_linear(alembic_cfg):
-    """The current project has a single linear history at 0001_initial."""
-    from alembic.script import ScriptDirectory
-
+    """The project keeps a single linear migration head."""
     script = ScriptDirectory.from_config(alembic_cfg)
     heads = script.get_heads()
     assert len(heads) == 1, f"expected one head, got: {heads}"
-    initial = next(iter(script.walk_revisions()))
-    assert initial.revision == "0001_initial"
+    assert heads == ["0002_asset_ingest_jobs"]
+    revisions = [rev.revision for rev in script.walk_revisions()]
+    assert revisions == ["0002_asset_ingest_jobs", "0001_initial"]