In SQLAlchemy (ORM) land, the standard is that the cursor object provides inserted_primary_key and lastrowid attributes. However, currently, both of them do not return anything sensible.
dataset uses SQLAlchemy's inserted_primary_key attribute for returning the pk value from its insert method, see https://github.com/pudo/dataset/blob/1.5.2/dataset/table.py#L120-L121. It is important to make this work in order to be able to reference records just inserted, for manipulating them later, or for referencing them elsewhere.
Other statements
.inserted_primary_key returns a tuple of the primary key of the single inserted row using the connection.execute() command. A good reason to use .inserted_primary_key instead of .lastrowid: .lastrowid is database specific, it may work for some databases while not for others. But .inserted_primary_key will always work regardless of the underlying database.
Problem
In SQLAlchemy (ORM) land, the standard is that the cursor object provides
inserted_primary_key
andlastrowid
attributes. However, currently, both of them do not return anything sensible.Motivation
dataset
uses SQLAlchemy'sinserted_primary_key
attribute for returning the pk value from itsinsert
method, see https://github.com/pudo/dataset/blob/1.5.2/dataset/table.py#L120-L121. It is important to make this work in order to be able to reference records just inserted, for manipulating them later, or for referencing them elsewhere.Other statements