I have a mapper created with a mapper extension that has an
after_update() override.
For a table in the mapper I do an update and then a commit().
This is the resulting order of execution:
update instance (setting an attribute on the mapped class)
commit
after_update called on instance
I.e.
PLEASE IGNORE PREVIOUS.
It turns out that explicitly flushing does change the order (made a
silly coding error before).
I am all set, but the question remains why autoflush isn't enough.
On May 23, 2:15 pm, Moshe C. [EMAIL PROTECTED] wrote:
I have a mapper created with a mapper extension that
On May 23, 2008, at 2:49 AM, schickb wrote:
On May 22, 8:37 pm, schickb [EMAIL PROTECTED] wrote:
Maybe SA is putting second execute in a transaction that isn't
being commited?
That was the problem. I found this in the SA docs: While many DBAPIs
implement a flag called autocommit, the
On May 23, 2008, at 7:26 AM, Moshe C. wrote:
PLEASE IGNORE PREVIOUS.
It turns out that explicitly flushing does change the order (made a
silly coding error before).
I am all set, but the question remains why autoflush isn't enough.
autoflush will occur before every query execute and
Evidently, when the autoflush occurred within the commit(), the database
transaction COMMIT itself happened before the call to after_update() .
The fact is that explicitly calling session.flush() immediately before
calling session.commit(), changed the final state of the DB.
This means the
On May 23, 2008, at 10:44 AM, Moshe Cohen wrote:
Evidently, when the autoflush occurred within the commit(), the
database transaction COMMIT itself happened before the call to
after_update() .
thats not how it works. the steps are:
session.commit()
session.flush()
Geoff wrote:
Does SQLA have any mechanism to use FORCE INDEX?
Not in generated SQL. There is a ticket to add hinting support, but
currently you'd need to use text() selects or join conditions to get the
hints in.
http://www.sqlalchemy.org/trac/ticket/921
All,
Perhaps someone can help me here since I am in that 'bad place' where
I am retrying things again and getting deeper than I want into it.
I am trying to create a database (using elixir model although the
problem appears to be in sqlalchemy so asking here) in Oracle
Express.
The problem I
All,
So have a 'working' solution but think it is pretty bad... Someone
MUST have a better solution. Here is what I had to do:
1. Have create_all explictly not test for tables existing (or else it
finds it in wrong schema)
create_all(checkfirst = False)
2. Explictly set the schema and owner
Hello,
I am doing a A-B join before I do any filtering on a query object
constructed from A. I then to filtering based on C. I have a
column_property mapped to A that gives me the total number of B. This
is executed as a subquery, which means it gives me the total number of
B in the database,
On May 23, 4:33 pm, Justin Tulloss [EMAIL PROTECTED] wrote:
I am doing a A-B join before I do any filtering on a query object
Sorry, this should be A-B-C join.
--~--~-~--~~~---~--~~
You received this message because you are subscribed to the Google Groups
I traced what is happening in the code.
I don't fully understand it but I hope the following will help.
The crucial point is that in my after_update() method I create a
mapped object and call session.save() using the same session (but
different table).
This is the sequence of events:
- I call
Howdy,
I'm a newbie to sqlalchemy and am having trouble understanding how to
turn selects into objects.
I've got two tables mapped into objects like this:
nis_accounts_table = Table( ... )
nis_users_table = Table( ... )
class NisAccount:
pass
class NisUser:
pass
mapper(NisUser,
On May 23, 2008, at 6:54 PM, Moshe C. wrote:
I traced what is happening in the code.
I don't fully understand it but I hope the following will help.
The crucial point is that in my after_update() method I create a
mapped object and call session.save() using the same session (but
On May 23, 2008, at 3:10 PM, Brandon Goldfedder wrote:
All,
Perhaps someone can help me here since I am in that 'bad place' where
I am retrying things again and getting deeper than I want into it.
I am trying to create a database (using elixir model although the
problem appears to be in
On May 23, 2008, at 5:33 PM, Justin Tulloss wrote:
Hello,
I am doing a A-B join before I do any filtering on a query object
constructed from A. I then to filtering based on C. I have a
column_property mapped to A that gives me the total number of B. This
is executed as a subquery, which
On May 23, 2008, at 7:42 PM, Jeff Putsch wrote:
Howdy,
I'm a newbie to sqlalchemy and am having trouble understanding how to
turn selects into objects.
I've got two tables mapped into objects like this:
nis_accounts_table = Table( ... )
nis_users_table = Table( ... )
class
OK, thanks very much.
My wrong assumption was that saves within the hook functions will make
it into the current flush.
Preceding the commit() with a manual flush, causes the commit to flush
this new saves (that occured in the flush).
Makes sense now :-)
On May 24, 2:52 am, Michael Bayer
Michael,
Yes - that is what I am doing now :
ForeignKey(schemaname.tablename.colname)
and I have things working. (see example 3). My problem is these steps
seem really ugly and a lot more work than it should be to get things
working portably and reliably - thus my question on if this is best
On May 23, 2008, at 5:07 PM, Michael Bayer wrote:
Jeff Putsch wrote:
Then I define some selects and execute them:
s = select([nis_accounts_table, nis_users_table],
from_obj=[nis_accounts_table.join(nis_users_table)]).where(
nis_users_table.c.eid != ''
)
20 matches
Mail list logo
