There is one common problem with PIL: if you tried to install it
*before* you install JPEG and other libraries, it won't recompile C
extensions with newer headers and libraries.
The solution is to rebuild PIL with -f switch:
python setup.py build_ext -f
(sudo) python setup.py install
Hope this h
Glad to hear that queryset-refactor is almost ready.
Currently, I noticed that there are few SQL portability issues and few
old queryset API issues (eg in admin). I already filed #6956 and
#6957. I do some heavy testing of this branch and I will report
anything that goes wrong.
Regards,
--
Ivan
> def create_user(self, username, email, password=None):
> def make_random_password(self, length=10,
> allowed_chars='abcdefghjkmnpqrstuvwxyzABCDEFGHJKLMNPQRSTUVWXYZ23456789'):
>
> These two methods seem to involve situations where they need to act on
> a model, but the object instance hasn't been
http://djangoamf.sourceforge.jp/index.php?DjangoAMF_en
On 1 фев, 18:33, "Ronaldo Z. Afonso" <[EMAIL PROTECTED]> wrote:
> Hi,
>
> I'm new in Django and Web development and I just want to know if it's
> possible to have a flash script showing elements that was retrieve from
> a data base by django?
> I'll try to redefine save methods in all the objects to see how much time it
> saves
You may want to redefine the __init__ methods as well (it's called
when you create your objects before save), it adds overhead of two
more signals that are probably unused and can be removed for speed.
You ca
You have two options:
1. Execute raw SQL 'INSERT' queries
2. Override the Model.save() or create new save_fast() method in your
Model class. The main speed eaters in Model.save() are
dispatcher.send() calls - so if you copy/paste the content of save
method from Django code without dispatcher.send
> def save(self):
> if self.stuff !=
> magic_function_that_tells_what_is_in_database_for('stuff')
> do_this()
> super(A, self).save()
This 'magic function' is
self.__class__.objects.get(id=self.id).stuff
if self.stuff != self.__class__.objects.get(id=self.id).stuff:
do_this()
Example of post_save with 'created' flag is in Django tests:
http://code.djangoproject.com/browser/django/trunk/tests/modeltests/signals/models.py
--~--~-~--~~~---~--~~
You received this message because you are subscribed to the Google Groups
"Django users" group.
There is 'post_save' signal with 'created' parameter. But,
unfortunately, this feature is undocumented...
On Jan 31, 6:44 am, Julien <[EMAIL PROTECTED]> wrote:
> Hello there,
>
> Does the title says it all? When overriding the save method of an
> object, I'd like to know if that object is being
If you don't want to rely on Django's smart_str, you can do following:
if isinstance(value, unicode):
value = value.encode('utf-8', 'ignore')
return textile.textile(value, encoding=settings.DEFAULT_CHARSET,
output=settings.DEFAULT_CHARSET)
textile.textile() certainly breaks with your error if
The problem here is not about wrong encoding of strings - it's all
about unicode/string relationship and from-unicode-to-string/from-
string-to-unicode problem.
I really think that it will be fixed if you replace
return textile.textile(value, encoding=settings.DEFAULT_CHARSET,
output=settings.DEF
> Hm, a code running fast is really good but this solution seems to be really
> experimental and hard-to-use, as well as making code kinda unreadable. It
> would be great if it's implemented in next django releases to make django
> fast :). Is it possible to do it?
It depends on Django developers
Jan 30, 8:25 am, Ivan Illarionov <[EMAIL PROTECTED]> wrote:
> You pass unicode value to textile.textile. I was able to repeat your
> error by doing the same: textile.textile breaks with similar error
> when it recieves a unicode string. It can be fixed by encoding your
> value as ut
You pass unicode value to textile.textile. I was able to repeat your
error by doing the same: textile.textile breaks with similar error
when it recieves a unicode string. It can be fixed by encoding your
value as utf-8 before passing to textile.textile.
Your error log has:
return textile.textile(
You have this in your log:
func
ignore_failures
False
Maybe try to change this to True. And double check that everything is
converted with 'ignore'.
Another option: store the textiled xhtml in database.
--~--~-~--~~~---~--~~
You received thi
> Our database is utf8 we explicitly convert ignoring errors to utf8
> before passing to textile. Also it's not 1 page that poops it's every
> site pooping at once.
That means that the problem is inside something that shows on every
page.
--~--~-~--~~~---~--~~
You r
I had A LOT of similar problems when I need to work with cyrillic. I
can easily replay your problem. Fortunately I'm on localized Windows
machine.
So:
>>> path = r'C:\Documents and Settings\vanilla'
>>> os.path.isdir(path)
True
>>> os.listdir(path)[-1]
'\xd8\xe0\xe1\xeb\xee\xed\xfb'
>>> a = os.lis
It looks that you have encoding problem. You have wrong characters
somewhere. The solution is to find the text that causes problems and
create custom encode and/or decode function that fixes this.
--~--~-~--~~~---~--~~
You received this message because you are subsc
After discussion on Django developers group about
Model.__init__(*args) deprecation I found the way to dramatically
optimize instantiation of Django models. Here is a code:
http://pastebin.com/m8e7e365
You may add this code to your Django model class and then instead of
obj = YourModelClass(titl
> Fortnately, Python makes this very easy with the built-in
> property() call:
>
> class MyModel(Model):
> surname = CharField(...)
> forenames = CharField(...)
> def _get_name(self):
> return self.forenames + ' ' + self.surname
> name = property(fget=_get_name)
Python (2.
Made some corrections:
http://pastebin.com/d6337d211
On Jan 28, 10:45 pm, Ivan Illarionov <[EMAIL PROTECTED]>
wrote:
> IMO code_berzerker's approach is the best for complex models. But in
> many cases that don't require a lot of flexibility it can be better to
> use s
IMO code_berzerker's approach is the best for complex models. But in
many cases that don't require a lot of flexibility it can be better to
use simpler approach with text fields for each language and function/
property to get the right field from templates. I even wrote the
metaclass to make this
Models that need flexibility have `lang` and `is_translation_of`
fields. Views (or custom managers) filter the output based on `lang`
and add the link to other language if translation exists. Some models
just have two separate text fields for each language and views (or
custom managers) display th
Piotr, having django-multilingual features in the core won't make them
any better. Complex models and use-cases will still need custom/manual
solutions. It will add unneeded overhead to single-language sites and
may even break the sites that use explicit custom solutions (like
mine).
I am develop
And it's also important to not really delete records from database.
Use the 'delete' boolean field to mark unneeded records.
On 18 янв, 06:18, otonk <[EMAIL PROTECTED]> wrote:
> > Create the sync script using the Python DB API directly (no Django
> > ORM). And then trigger it either automatically
On 17 янв, 05:58, otonk <[EMAIL PROTECTED]> wrote:
> hi, i am currently designing two or more django application, one
> application act as the master application, contains all the data, the
> other application has partial function of the master application. The
> master database belongs to the mas
26 matches
Mail list logo
