def create_sql_engine(cls, url_or_con, table):
con = ConnectionHandler.get_connection(url_or_con)
if isinstance(con, DuckDBPyConnection):
return DuckDBEngine(con, table)
elif isinstance(con, Connection):
return SQLAlchemyEngine(con, table)
else:
raise UnsupportedConnectionObjectException(f"Connection object '{con.__class__.__name__}' not supported.")
def create_sql_metadata_repository(cls,
url_or_con) -> SQLMetadataRepository:
con = ConnectionHandler.get_connection(url_or_con)
if isinstance(con, DuckDBPyConnection):
return DuckDBMetadataRepository(con)
elif isinstance(con, Connection):
return SQLAlchemyMetadataRepository(con)
else:
raise UnsupportedConnectionObjectException(
f"Connection object {con} not supported.")
def __init__(self, data: DataFrame, no_sharing=False):
super().__init__()
con = ConnectionHandler.get_connection("duckdb://:memory:")
# activate multiprocessing with available cpu's but one
con.execute(f"PRAGMA threads={multiprocessing.cpu_count()}")
# register df
con.register("data", data)
self.engine = DuckDBEngine(con, "data", no_sharing=no_sharing)
self.data = data
self.table = "data"