I wanted to know some views about the soft-delete (anti?) pattern and
whether or not I'm going about it the right way in my code base.
We have a piece of code:
class NoDeletedQuery(Query):
def __new__(cls, *args, **kwargs):
if args and hasattr(args[0][0], "deleted_at"):
return Query(*args, **kwargs).filter_by(deleted_at=None)
else:
return object.__new__(cls)
Session = scoped_session(sessionmaker(query_cls=NoDeletedQuery))
This is pretty obvious, just auto adds a filter that ignores the deleted
at, but, I think this is a very hard thing to get right, especially with
joins and whatnot. This works, but I would like to know what's the
recommended idiom? I looked at this stack overflow question:
http://stackoverflow.com/questions/920724/the-right-way-to-auto-filter-sqlalchemy-queries
The mapper solution seems ideal, which would look like:
active_accts_q = (select([
accounts_t
])
.where(accounts_t.c.deleted_at == None).alias()
)
class Account(Base):
__table__ = active_accts_q
def soft_delete(self):
self.deleted_at = func.clock_timestamp()
The problem I get here is that in my code, I do something like:
account.soft_delete()
Session.commit()
and I get a
ObjectDeletedError: Instance '<Account at 0x102b2d390>' has been deleted,
or its row is otherwise not present.
So, I'm guessing I'm just doing it wrong. Any suggestions?
For a little context, we're using soft deletes and and for external
purposes, we need to simulate a resource deletion but *still* have some
attributes of the "deleted" row come back. There are other ways to solve
this, which can be solved by triggers and the like, but I think updating a
foreign key and adding a new column that represents a deleted view is too
complicated for something that can be simplified by just having this
"deleted_at" property.
Thanks!
--
You received this message because you are subscribed to the Google Groups
"sqlalchemy" group.
To post to this group, send email to sqlalchemy@googlegroups.com.
To unsubscribe from this group, send email to
sqlalchemy+unsubscr...@googlegroups.com.
For more options, visit this group at
http://groups.google.com/group/sqlalchemy?hl=en.
