knowledge-platform/tests/integration/test_document_migration.py
2026-09-02 20:42:52 +08:00

123 lines
3.5 KiB
Python

"""Migration tests for canonical Document metadata."""
from pathlib import Path
from alembic import command
from alembic.config import Config
from sqlalchemy import create_engine, inspect, text
def test_document_migration_matches_metadata(tmp_path: Path) -> None:
project_root = Path(__file__).resolve().parents[2]
database_path = tmp_path / "migration.sqlite3"
database_url = f"sqlite+pysqlite:///{database_path}"
config = Config(project_root / "alembic.ini")
config.attributes["database_url"] = database_url
# The production OceanBase database is shared. A foreign table must survive
# migration and remain outside Alembic's autogenerate/drift ownership.
with create_engine(database_url).begin() as connection:
connection.execute(text("CREATE TABLE unrelated_provider_table (id INTEGER PRIMARY KEY)"))
command.upgrade(config, "head")
command.check(config)
inspector = inspect(create_engine(database_url))
assert {
"documents",
"chunks",
"chunk_index_states",
"index_profiles",
"ingestion_jobs",
"ingestion_attempts",
"ingestion_batches",
"ingestion_batch_items",
}.issubset(inspector.get_table_names())
assert "unrelated_provider_table" in inspector.get_table_names()
assert {column["name"] for column in inspector.get_columns("documents")} == {
"id",
"filename",
"media_type",
"size_bytes",
"storage_key",
"parsed_storage_key",
"status",
"error_message",
"created_at",
"updated_at",
}
assert {column["name"] for column in inspector.get_columns("chunks")} == {
"chunk_id",
"document_id",
"filename",
"page",
"page_end",
"section",
"section_path",
"chunk_index",
"content",
"token_count",
"strategy_version",
"created_at",
}
assert {column["name"] for column in inspector.get_columns("chunk_index_states")} == {
"chunk_id",
"index_profile_id",
"embedding_model",
"vector_dimension",
"indexed_at",
}
assert {column["name"] for column in inspector.get_columns("index_profiles")} == {
"profile_id",
"chunk_strategy_version",
"embedding_model",
"vector_dimension",
"status",
"created_at",
"activated_at",
"active_marker",
}
assert {column["name"] for column in inspector.get_columns("ingestion_jobs")} == {
"id",
"document_id",
"status",
"stage",
"attempt_count",
"max_attempts",
"error_message",
"next_run_at",
"lease_expires_at",
"created_at",
"updated_at",
"completed_at",
}
assert {column["name"] for column in inspector.get_columns("ingestion_attempts")} == {
"id",
"job_id",
"attempt_number",
"status",
"stage",
"retryable",
"error_message",
"active_marker",
"started_at",
"finished_at",
}
assert {column["name"] for column in inspector.get_columns("ingestion_batches")} == {
"id",
"total_files",
"created_at",
}
assert {column["name"] for column in inspector.get_columns("ingestion_batch_items")} == {
"id",
"batch_id",
"item_index",
"filename",
"acceptance_status",
"document_id",
"job_id",
"error_message",
"created_at",
"updated_at",
}