On Wednesday, April 4, 2018 at 4:49:16 PM UTC-4, Mike Bayer wrote:
>
> I've not seen any context for what you're doing so is using
> "query.select_from()" part of how you want to handle this?
>
My personal fix was just to flip the order of entities submitted to
`.query()` - but I could have
I've not seen any context for what you're doing so is using
"query.select_from()" part of how you want to handle this?
On Wed, Apr 4, 2018 at 1:56 PM, Jonathan Vanasco wrote:
> I built a simplified test-case, then tried to simplify it more into a second
> form -- and got a
I built a simplified test-case, then tried to simplify it more into a
second form -- and got a different error.
The second error message, and sqlalchemy source code lines, confirmed my
initial belief this is an intentional behavior, which is not reflected in
the `query` or `join` API docs.
On Wednesday, April 4, 2018 at 4:31:36 AM UTC-4, Simon King wrote:
> Where is the traceback raised? On the evaluation of the
> "subquery.c.event_timestamp" expression? That wouldn't really make any
> sense. Or is it when the query itself is evaluated? We need to see at
> least the
On Wed, Apr 4, 2018 at 4:08 AM, Jonathan Vanasco wrote:
>
>
> On Tuesday, April 3, 2018 at 10:34:03 PM UTC-4, Mike Bayer wrote:
>>
>> "does not work" ?
>
>
> wow, i am an awful person. sorry. that is the least helpful description of
> what happens i can think of.
>
> second
On Tuesday, April 3, 2018 at 10:34:03 PM UTC-4, Mike Bayer wrote:
>
> "does not work" ?
wow, i am an awful person. sorry. that is the least helpful description of
what happens i can think of.
second try:
if the subquery is the 1st element, the select compiles correctly and gives
me the
"does not work" ?
On Tue, Apr 3, 2018 at 6:43 PM, Jonathan Vanasco wrote:
> I couldn't find docs on this behavior, so wanted to ask before filing a
> ticket with a test-case. If this isn't the expected behavior, I'll generate
> a SSCCE. I think it might be expected though.
I couldn't find docs on this behavior, so wanted to ask before filing a
ticket with a test-case. If this isn't the expected behavior, I'll
generate a SSCCE. I think it might be expected though.
I have a form of a query that uses an ORM object joined against a subquery
It works when the