[sqlalchemy] Re: Generated slow JOIN sql statement/specifying a left JOIN

2007-04-26 Thread Andreas Jung
--On 25. April 2007 07:13:19 -0400 Michael Bayer [EMAIL PROTECTED] wrote: well i just noticed youre sticking an external table into your primaryjoinbut i see nothing there that implies the usage of a LEFT OUTER JOIN ? are you saying that you dont need the rows from AMH_View to be

[sqlalchemy] Re: Generated slow JOIN sql statement/specifying a left JOIN

2007-04-26 Thread Michael Bayer
On Apr 26, 2007, at 2:44 AM, Andreas Jung wrote: --On 25. April 2007 07:13:19 -0400 Michael Bayer [EMAIL PROTECTED] wrote: well i just noticed youre sticking an external table into your primaryjoinbut i see nothing there that implies the usage of a LEFT OUTER JOIN ? are you

[sqlalchemy] Re: Generated slow JOIN sql statement/specifying a left JOIN

2007-04-25 Thread Michael Bayer
On Apr 25, 2007, at 12:25 AM, Andreas Jung wrote: hmm..Why has this to do with self-referential mappers? Wouldn't the generated SQL be same if it wasn't a self-referential mapper but just mapper with a property for a one-to-many relationship? well no, the eager load paradigm is:

[sqlalchemy] Re: Generated slow JOIN sql statement/specifying a left JOIN

2007-04-24 Thread Michael Bayer
On Apr 24, 2007, at 4:27 AM, Andreas Jung wrote: Because both tables are big the query takes forever. Using a LEFT JOIN would definitely be faster. Is there a way to configure the 'tools' property in a smarter way? not within the relation(), we dont support self-referential eager loads

[sqlalchemy] Re: Generated slow JOIN sql statement/specifying a left JOIN

2007-04-24 Thread Andreas Jung
--On 24. April 2007 08:54:55 -0400 Michael Bayer [EMAIL PROTECTED] wrote: On Apr 24, 2007, at 4:27 AM, Andreas Jung wrote: Because both tables are big the query takes forever. Using a LEFT JOIN would definitely be faster. Is there a way to configure the 'tools' property in a smarter