V5.1 Modify Person table: split full name into first & last, added tags support

tests/test_db.py

@@ -20,6 +20,7 @@ from transcription.db import dispose_database_runtime
 from transcription.db import initialize_database_runtime
 from transcription.db import reconcile_canonical_media_paths
 from transcription.db import reconcile_legacy_job_source_columns
+from transcription.db import reconcile_person_name_columns
 from transcription.db import session_scope
 from transcription.db.models import DocumentType
 from transcription.db.models import PersonRole
@@ -55,6 +56,7 @@ async def test_create_all_creates_expected_tables(tmp_path):
         assert "person_role" in table_names
         assert "document_person" in table_names
         assert "document_tag" in table_names
+        assert "person_tag" in table_names
         assert "job" in table_names
         assert "source" in table_names
         assert "job_source" in table_names
@@ -141,7 +143,7 @@ async def test_create_all_declares_hot_path_indexes(tmp_path):
                 database = inspect(sync_connection)
                 indexes = {
                     table: [index["column_names"] for index in database.get_indexes(table)]
-                    for table in ("job", "source", "job_source", "document", "document_person", "document_tag")
+                    for table in ("job", "source", "job_source", "document", "document_person", "document_tag", "person_tag")
                 }
                 job_source_unique = [
                     constraint["column_names"]
@@ -151,9 +153,13 @@ async def test_create_all_declares_hot_path_indexes(tmp_path):
                     constraint["column_names"]
                     for constraint in database.get_unique_constraints("document_tag")
                 ]
-                return indexes, job_source_unique, document_tag_unique
+                person_tag_unique = [
+                    constraint["column_names"]
+                    for constraint in database.get_unique_constraints("person_tag")
+                ]
+                return indexes, job_source_unique, document_tag_unique, person_tag_unique
 
-            indexes, job_source_unique, document_tag_unique = await connection.run_sync(collect)
+            indexes, job_source_unique, document_tag_unique, person_tag_unique = await connection.run_sync(collect)
 
         assert ["status", "date_created"] in indexes["job"]
         assert ["document_id"] in indexes["job"]
@@ -168,6 +174,46 @@ async def test_create_all_declares_hot_path_indexes(tmp_path):
         assert ["document_id"] in indexes["document_tag"]
         assert ["tag_id"] in indexes["document_tag"]
         assert ["document_id", "tag_id"] in document_tag_unique
+        assert ["person_id"] in indexes["person_tag"]
+        assert ["tag_id"] in indexes["person_tag"]
+        assert ["person_id", "tag_id"] in person_tag_unique
+    finally:
+        await dispose_database_runtime()
+
+
+@pytest.mark.asyncio
+async def test_reconcile_person_name_columns_backfills_split_names(tmp_path):
+    settings = Settings(
+        openrouter_api_key="test-key",
+        database=SqliteSettings(path=str(tmp_path / "legacy-person-name.db")),
+        environment="test",
+    )
+    runtime = initialize_database_runtime(settings=settings)
+
+    try:
+        await create_all(engine=runtime.engine)
+        async with runtime.engine.begin() as connection:
+            await connection.execute(text('alter table "person" add column "full_name" varchar'))
+            await connection.execute(
+                text(
+                    'insert into "person" (id, full_name, given_names, last_name, created_at, updated_at) '
+                    "values (:id, :full_name, '', '', CURRENT_TIMESTAMP, CURRENT_TIMESTAMP)"
+                ),
+                {"id": "55" * 16, "full_name": "Ada Lovelace"},
+            )
+
+        changed = await reconcile_person_name_columns(engine=runtime.engine)
+        assert changed >= 1
+
+        async with runtime.engine.connect() as connection:
+            row = (
+                await connection.execute(
+                    text('select given_names, last_name from "person" where id = :id'),
+                    {"id": "55" * 16},
+                )
+            ).one()
+        assert row[0] == "Ada"
+        assert row[1] == "Lovelace"
     finally:
         await dispose_database_runtime()
 
@@ -227,10 +273,10 @@ async def test_reconcile_canonical_media_paths_normalizes_source_and_photo_paths
         async with runtime.engine.begin() as connection:
             await connection.execute(
                 text(
-                    'insert into "person" (id, full_name, created_at, updated_at) '
-                    'values (:id, :full_name, CURRENT_TIMESTAMP, CURRENT_TIMESTAMP)'
+                    'insert into "person" (id, given_names, last_name, created_at, updated_at) '
+                    'values (:id, :given_names, :last_name, CURRENT_TIMESTAMP, CURRENT_TIMESTAMP)'
                 ),
-                {"id": "11" * 16, "full_name": "Portrait"},
+                {"id": "11" * 16, "given_names": "Portrait", "last_name": "Person"},
             )
             await connection.execute(
                 text(