Please help me in orm level filtering many-to-many relations.
I have to filter all "authors" who does not belongs to any "articles".
Or the opposite of this, list of orphaned articles...
I dont want to write sql in the model if possible.
Thank you,
Charlie.
--
You received this message because
Tomasz Zieliński wrote:
> How about something like this (should work, although I haven't checked
> it):
>
> Author.objects.annotate(article_num=Count('articles')).filter
> (article_num=0)
>
> - where articles=ManyToManyField('Article')
>
This is so elegant, so i have to upgrade to 1.1 now... :)
Hi there,
we have models like this:
1. A car database with freely assignable attributes.
2. An attribute-type table (ex: color, engine, seats)
3. Car attributes (ex: color:red, seats: 5) variable car by car
We have to display a matrix of cars (rows) with different attributes
(columns), but not
Is there anybody out there using mptt?
'full_tree_model' tags and filters are working correctly, but i was
unable to use a 'drilldown_tree_for_node'.
Does anybody have a working example or detailed instructions?
tia,
Charlie.
--
"s minden mestert kinevettem, ki nem nevetett önmagán."
GPG publ
We are starting a new project soon (within a month, after closing a
research phase).
Django V1.0 means (as far as i can recall from wiki) "do not expect
backward compatibility before 1.0".
Obviously major design elements and data models are not affected later,
but (watching django version number c
Joseph Kocherhans írta:
>This is the most up to date plan for 1.0
>
> http://code.djangoproject.com/wiki/VersionOneFeatures
>
>
Thanks, i did not see.
Charlie.
--
"...s minden mestert kinevettem, ki nem nevetett önmagán."
GPG public key: http://www.rendszergazda.com/gpg/charlie-gpg-publi
I have a tree model data and a foreign key is defined with a null=True.
Generated db table is correct (field is NULL instead of NOT NULL), but
in the admin interface i cannot add an object with empty parent. (With
not null parent it works.)
And obviously i cannot add first one at all.
What was i a
patrickk írta:
>maybe this helps:
>http://www.djangoproject.com/documentation/model_api/#blank
>
>
Helps? Works!
Thank you, Patrick.
Charlie.
--
"...s minden mestert kinevettem, ki nem nevetett önmagán."
GPG public key: http://www.rendszergazda.com/gpg/charlie-gpg-public-key.asc
--~--~---
Is it a desired behaviour of templating system to equalize text input
widgets?
Although in my model two fields are different sizes (maxlength=20 and
maxlength=200), inserting into template it generates with "size=30" tag
equally.
I can manually control it, but why should i do if it knows everythin
patrickk írta:
>maxlength is not a size-attribute.
>you probably don´t want size=200, just because maxlength is 200.
>
>
Probably. But i don't want "size=30" either.
Is "30" hard-wired or reflects to something?
Charlie.
--
"...s minden mestert kinevettem, ki nem nevetett önmagán."
GPG public
Matt the Destroyer írta:
>size=30 simply tells the browser to make the input field a size of 30
>characters wide. it has no bearing on input data or transmission.
>
>in the past, i have simply hard-coded my own input fields into my
>templates so I can directly control the size and other attribut
Malcolm Tredinnick írta:
>Sorry, that came across a bit more negatively than I intended: it almost
>certainly will be accepted (see the threads linked from the bug). So
>it's not a roll of the dice -- the solution isn't ideal, but barring a
>better solution, it is a reasonable idea.
>
>
The goa
Steven Armstrong írta:
>I've extended some of django's built in model and form fields to accept
>an additional kwarg named 'html_attributes'.
>
>It works like this:
>
>class Whatever(models.Model):
> ...
> title = models.CharField(maxlength=128,
> html_attributes={
> 'size': 15,
>
Malcolm Tredinnick írta:
>This is the sort of thing that maybe starts crossing the line between
>representation and presentation a bit too much. You are putting a lot of
>presentation information into the data model description and they
>should, as much as possible be separate. That is why I (and
Steven Armstrong írta:
>Instead of writing a patch, I've extended/wrapped djangos model and form
>fields. This way you can update django without having to reapply patches.
>
>You can find the code and instructions under [1].
>
>Just for the record. I'm totally with Malcolm that this is not a pre
15 matches
Mail list logo
