Yes I know the question was a bit vague, but i couldn't think of a simple
way to show some code. A test case would probably require some major
surgery; I'll consider it though.

It is possible that the second instance is associated with a different
session, i'd have to check. What actually happens if you add the same
instance to a second session?

CL


On Mon, Sep 2, 2013 at 8:23 PM, Simon King <[email protected]> wrote:

> Without knowing your application it's very difficult to know...
>
> Is there any chance that the second object is already associated with a
> different session? Perhaps from a different thread, or retrieved from a
> cache or something similar?
>
> Producing a standalone test script would help in diagnosing the problem.
>
> Simon
>
>
> On 2 Sep 2013, at 18:56, Lars van Gemerden <[email protected]> wrote:
>
> > Well, from the message yes, but i am not setting any primary keys
> manually, so where could the second instance come from?
> >
> > CL
> >
> > ====================================
> > Lars van Gemerden
> > [email protected]
> > +31 6 26 88 55 39
> > ====================================
> >
> > On 2 sep. 2013, at 17:39, Simon King <[email protected]> wrote:
> >
> >> On Mon, Sep 2, 2013 at 4:01 PM, lars van gemerden <[email protected]>
> wrote:
> >>> Hi all,
> >>>
> >>> I have a one-to-one and many-to-one relationship (with an association
> table
> >>> in the middle; Report.author-assoctable-Person.reports; why the table
> in the
> >>> middle, you might ask, well it's a long story) and i am setting the
> author
> >>> on a specific Report twice to the same value. This gives the following
> >>> error:
> >>>
> >>>          InvalidRequestError: Can't attach instance <person at
> 0x36a9350>;
> >>> another instance with key (<class 'models.data.classes.person'>, (2,))
> is
> >>> already present in this session.
> >>>
> >>> The first time i set the author, it works and if i change the author it
> >>> works as well.
> >>>
> >>> This seems i little strange to me, why not be able to set the
> attribute to
> >>> the same value twice?
> >>>
> >>> Otherwise i have to check whether the new and old values are equal
> (which
> >>> through the peculiarties of my code is somewhat wastefull)
> >>>
> >>
> >> I think the message is actually telling you that you have 2 different
> >> instances of your "person" class with the same primary key. SQLAlchemy
> >> uses an identity map to ensure that a particular row in the database
> >> is only identified by a single instance within a particular session.
> >> If you violate that assumption (eg. by loading one instance of
> >> "person" from the database, then creating another instance and setting
> >> its primary key), you will get errors like this.
> >>
> >> Does that sound plausible?
> >>
> >> Simon
> >>
> >> --
> >> You received this message because you are subscribed to a topic in the
> Google Groups "sqlalchemy" group.
> >> To unsubscribe from this topic, visit
> https://groups.google.com/d/topic/sqlalchemy/bkm3-nB6d1o/unsubscribe.
> >> To unsubscribe from this group and all its topics, send an email to
> [email protected].
> >> To post to this group, send email to [email protected].
> >> Visit this group at http://groups.google.com/group/sqlalchemy.
> >> For more options, visit https://groups.google.com/groups/opt_out.
> >
> > --
> > You received this message because you are subscribed to the Google
> Groups "sqlalchemy" group.
> > To unsubscribe from this group and stop receiving emails from it, send
> an email to [email protected].
> > To post to this group, send email to [email protected].
> > Visit this group at http://groups.google.com/group/sqlalchemy.
> > For more options, visit https://groups.google.com/groups/opt_out.
>
> --
> You received this message because you are subscribed to a topic in the
> Google Groups "sqlalchemy" group.
> To unsubscribe from this topic, visit
> https://groups.google.com/d/topic/sqlalchemy/bkm3-nB6d1o/unsubscribe.
> To unsubscribe from this group and all its topics, send an email to
> [email protected].
> To post to this group, send email to [email protected].
> Visit this group at http://groups.google.com/group/sqlalchemy.
> For more options, visit https://groups.google.com/groups/opt_out.
>



-- 
====================================
Lars van Gemerden
[email protected]
+31 6 26 88 55 39
====================================

-- 
You received this message because you are subscribed to the Google Groups 
"sqlalchemy" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
To post to this group, send email to [email protected].
Visit this group at http://groups.google.com/group/sqlalchemy.
For more options, visit https://groups.google.com/groups/opt_out.

Reply via email to