On Jun 17, 2010, at 3:15 PM, Andi Albrecht wrote:

> Adding these two lines to your test makes it fail:

it turns out there is a test for this, and also that passing "None" does not 
have the same effect as order_by() never being called in the first place - it 
suppresses the order_by setting on the mapper, if one was made.  So not the 
same as "order by was never called", it explicitly can be used to disable an 
order_by setting on a mapper.

I implemented support for passing False to order_by() in r8078b02c5aa6, which 
does in fact make it as though order_by() was never called.



> 
> q.delete()
> eq_(s.query(User).count(), 0)
> 
> Here's the complete test to see it in context again:
> 
> class OrderByTest(QueryTest, AssertsCompiledSQL):
>    def test_cancel_order_by(self):
>        s = create_session()
> 
>        q = s.query(User).order_by(User.id)
>        self.assert_compile(q,
>            "SELECT users.id AS users_id, users.name AS users_name
> FROM users ORDER BY users.id",
>            use_default_dialect=True)
> 
>        q = q.order_by(None)
>        self.assert_compile(q,
>                "SELECT users.id AS users_id, users.name AS users_name
> FROM users",
>                use_default_dialect=True)
> 
>        q.delete()
>        eq_(s.query(User).count(), 0)
> 
> Andi
> On Thu, Jun 17, 2010 at 4:15 PM, Michael Bayer <[email protected]> 
> wrote:
>> OK, it looks like we did add this at some point, though I could not find a 
>> unit test where I would expect.  So I added one in the latest tip.  Here it 
>> is:
>> 
>> class OrderByTest(QueryTest, AssertsCompiledSQL):
>>    def test_cancel_order_by(self):
>>        s = create_session()
>> 
>>        q = s.query(User).order_by(User.id)
>>        self.assert_compile(q,
>>            "SELECT users.id AS users_id, users.name AS users_name FROM users 
>> ORDER BY users.id",
>>            use_default_dialect=True)
>> 
>>        q = q.order_by(None)
>>        self.assert_compile(q,
>>                "SELECT users.id AS users_id, users.name AS users_name FROM 
>> users",
>>                use_default_dialect=True)
>> 
>> 
>> and it passes.   So how to reproduce the failure I assume you are getting ?
>> 
>> 
>> 
>> On Jun 17, 2010, at 3:14 AM, Andi Albrecht wrote:
>> 
>>> It's not different at all. I've just noticed that on the one hand it
>>> *is* possible to use query.order_by(None) to remove any previously set
>>> ordering, but on the other hand that _no_select_modifiers() in
>>> sqlalchemy/orm/query.py checks for False (and not None). It would be
>>> clearer to me if either order_by(False) would work too or - even
>>> better - if _no_select_modifiers() would allow None for _order_by.
>>> 
>>> Andi
>>> 
>>> On Wed, Jun 16, 2010 at 3:30 PM, Michael Bayer <[email protected]> 
>>> wrote:
>>>> its not really provided right now, the same way saying query.filter(None) 
>>>> won't reset any existing WHERE criterion, or join(None) doesn't remove all 
>>>> joins.    This issue has come up before.  How is order_by() different ?
>>>> 
>>>> 
>>>> On Jun 16, 2010, at 8:05 AM, Andi Albrecht wrote:
>>>> 
>>>>> Hi,
>>>>> 
>>>>> I'm a bit curious about how to unset an order_by in 0.6.
>>>>> 
>>>>> query.order_by(None) should do it, but in sqlalchemy/orm/query.py in
>>>>> _no_select_modifiers() the notset value for the _order_by attribute is
>>>>> False. In turn, the order_by() method doesn't seem to accept False.
>>>>> 
>>>>> What would be a proper way to unset an order_by? In my case the
>>>>> order_by is set when classes are mapped.
>>>>> 
>>>>> Best regards,
>>>>> 
>>>>> Andi
>>>>> 
>>>>> --
>>>>> You received this message because you are subscribed to the Google Groups 
>>>>> "sqlalchemy" group.
>>>>> To post to this group, send email to [email protected].
>>>>> To unsubscribe from this group, send email to 
>>>>> [email protected].
>>>>> For more options, visit this group at 
>>>>> http://groups.google.com/group/sqlalchemy?hl=en.
>>>>> 
>>>> 
>>>> --
>>>> You received this message because you are subscribed to the Google Groups 
>>>> "sqlalchemy" group.
>>>> To post to this group, send email to [email protected].
>>>> To unsubscribe from this group, send email to 
>>>> [email protected].
>>>> For more options, visit this group at 
>>>> http://groups.google.com/group/sqlalchemy?hl=en.
>>>> 
>>>> 
>>> 
>>> --
>>> You received this message because you are subscribed to the Google Groups 
>>> "sqlalchemy" group.
>>> To post to this group, send email to [email protected].
>>> To unsubscribe from this group, send email to 
>>> [email protected].
>>> For more options, visit this group at 
>>> http://groups.google.com/group/sqlalchemy?hl=en.
>>> 
>> 
>> --
>> You received this message because you are subscribed to the Google Groups 
>> "sqlalchemy" group.
>> To post to this group, send email to [email protected].
>> To unsubscribe from this group, send email to 
>> [email protected].
>> For more options, visit this group at 
>> http://groups.google.com/group/sqlalchemy?hl=en.
>> 
>> 
> 
> -- 
> You received this message because you are subscribed to the Google Groups 
> "sqlalchemy" group.
> To post to this group, send email to [email protected].
> To unsubscribe from this group, send email to 
> [email protected].
> For more options, visit this group at 
> http://groups.google.com/group/sqlalchemy?hl=en.
> 

-- 
You received this message because you are subscribed to the Google Groups 
"sqlalchemy" group.
To post to this group, send email to [email protected].
To unsubscribe from this group, send email to 
[email protected].
For more options, visit this group at 
http://groups.google.com/group/sqlalchemy?hl=en.

Reply via email to