Wow, I didn't know that... is it a bug?
But, the RETURNING clause will work though :)
stmt = P.__table__.insert(returning=[P.id], values=[{"val": 1}, {"val": 2}])
with engine.connect() as conn:
result = conn.execute(stmt)
# do something with result
hmm, maybe perfomance suffers from huge
On Mar 26, 2014, at 1:56 PM, Staszek wrote:
> On 2014-03-26 18:16, Michael Bayer wrote:
>> It's not totally clear which of the common use cases you have here, pick one:
>>
>> 1. I want to emit a JOIN from User to Car and order by Car.model - don't use
>> eager loading, use query.join():
>
>
On 2014-03-26 18:16, Michael Bayer wrote:
> It's not totally clear which of the common use cases you have here, pick one:
>
> 1. I want to emit a JOIN from User to Car and order by Car.model - don't use
> eager loading, use query.join():
This one.
> http://docs.sqlalchemy.org/en/rel_0_9/faq.ht
On Mar 26, 2014, at 1:04 PM, Staszek wrote:
> Hi
>
> I have a problem with joined relationships, which I will try to describe
> using an example:
>
>
>car = relationship("Car", lazy='joined', innerjoin=True)
>
> Querying for User now seems to generate SQL expressions like this:
>
> SELE
Hi
I have a problem with joined relationships, which I will try to describe
using an example:
class User(Base):
__tablename__ = 'users'
__table_args__ = ...
id = Column(BigInteger, autoincrement=True, primary_key=True)
...
carId = Column(BigInteger, ForeignKey('cars.id'))
reread your example code, that will actually do the trick.
On Wednesday, March 26, 2014 9:07:52 AM UTC-4, Michael Robellard wrote:
>
> I want it mapped to an object because I want to have some
> properties/methods that can be executed on each row of the results that
> will do some additional log
I want it mapped to an object because I want to have some
properties/methods that can be executed on each row of the results that
will do some additional logic.
On Tuesday, March 25, 2014 6:21:31 PM UTC-4, Michael Bayer wrote:
>
>
> On Mar 25, 2014, at 2:43 PM, Michael Robellard
> >
> wrote:
