Hi all,
Maybe (hopefully;-) i am missing something simple, but i have the following
test code:
engine = create_engine('sqlite:///:memory:', echo=True)
Session = sessionmaker(bind=engine)
Base = declarative_base()
class User(Base):
On Mar 22, 2014, at 9:16 AM, lars van gemerden l...@rational-it.com wrote:
query = session.query(Email)
query = query.join(user) #or query = query.join(user, aliased =
True)
query = query.add_columns(Email.email, User.name)
the add_columns() method does not have the
In the documentation for sqlalchemy.orm.column_property, there is a section
explaining the *active_history* flag. It is very terse. I think it is very
useful but just cannot think of a scenario where I would ever want to use
*previous* value when I'm setting a new value. Could someone use plain
when you're using an event such as before_flush() to check on
attributes.get_history() so that you can see that a new version needs to be
inserted into a version table, you need the old value of the attribute in order
to compare. see the versioned_history example.
On Mar 22, 2014, at 4:34
Hi Michael,
Just a quick line, version table is not something particularly useful for
beginner users, am I right? Thanks.
On Sat, Mar 22, 2014 at 2:52 PM, Michael Bayer mike...@zzzcomputing.comwrote:
when you're using an event such as before_flush() to check on
attributes.get_history() so
heh. reading the SQLA docs cover to cover is not particularly useful for
beginners either... :)
On Mar 22, 2014, at 6:12 PM, Bao Niu niuba...@gmail.com wrote:
Hi Michael,
Just a quick line, version table is not something particularly useful for
beginner users, am I right? Thanks.
On
