On 21 Nov 2008, at 03:32, Michael Bayer wrote:
>
>
> On Nov 20, 2008, at 8:47 PM, Michael Bayer wrote:
>
>>> The first scenario is a single table with 24,000 rows. The problem
>>> is
>>> that using SQLAlchemy through Elixir to map this table to an object,
>>> and performing a fairly naive Mapp
> -Original Message-
> From: sqlalchemy@googlegroups.com
> [mailto:[EMAIL PROTECTED] On Behalf Of SinJax
> Sent: 21 November 2008 10:43
> To: sqlalchemy
> Subject: [sqlalchemy] Re: ORM mapping with Elixir compared to
> raw cursor query
>
[SNIP]
> sql = &qu
On Friday 21 November 2008 12:42:31 SinJax wrote:
and another way:
allAnn = Annotation.query(
).join( 'user').filter_by( py_user = "msn"
).reset_joinpoint().filter( Annotation.subject == somesubject )
and
allAnn = Annotation.query(
).filter_by( subject = somesubject
).join( 'user'
> The result of this query is 23 rows from a 24,000 row table.
>
> I show a query i constructed by hand (Get every annotation of user
> "msn" on subject "2") as i would run using MysqlDb and cursor. I
> time this by measuring the time taken between executing the query
> and putting each result in
I had a feeling the bulk nature of the query would confuse the matter
so i have performed some more tests with a slightly more restricted
query with similarly poor results.
I have not posted much on mail lists much so if it is inappropriate to
post code i apologise.
Here is my code:
On Nov 20, 2008, at 8:47 PM, Michael Bayer wrote:
>> The first scenario is a single table with 24,000 rows. The problem is
>> that using SQLAlchemy through Elixir to map this table to an object,
>> and performing a fairly naive MappedThing.query().all() the process
>> takes roughly 4.8 seconds t
On Nov 20, 2008, at 7:07 PM, SinJax wrote:
>
> Hi, I've made this post already on my blog but it was suggested i post
> here as it might be an interesting point of discussion.
>
> The first scenario is a single table with 24,000 rows. The problem is
> that using SQLAlchemy through Elixir to map