I have a model Property with certain fields and a relevant method:
class Property(models.Model):
table = models.ForeignKey(Table)
field1 = models.CharField()
field2 = models.IntegerField()
field3 = models.BooleanField()
class Meta:
abstract = True
def post():
pass
But then I have a definite number of types of columns, conceptually speaking. There is no difference in the fields, only in how the behavior of a certain method is implemented:
class Property1(Property):
def post():
# execute behavior for Property1
pass
class Property2(Property):
def post():
# execute behavior for Property2
pass
and so on.
If I turned Property into an abstract base model class and have the rest inherit it, I will end up with different tables for each property. I am not sure I want that. All tables will look the same, which is redundant.
But at the same time when running a query to get all properties in a table and calling post() I want the corresponding behavior to be executed:
for prop in table.property_set.all():
prop.post()
What are my options?
For that, you can use proxy model. Try like this:
class Property(models.Model):
table = models.ForeignKey(Table)
field1 = models.CharField()
field2 = models.IntegerField()
field3 = models.BooleanField()
class Property1(Property):
class Meta:
proxy = True
def post():
# execute behavior for Property1
pass
class Property2(Property):
class Meta:
proxy = True
def post():
# execute behavior for Property2
pass
As per documentation:
The MyPerson class operates on the same database table as its parent Person class. In particular, any new instances of Person will also be accessible through MyPerson, and vice-versa:
So you can get the proxy instances like this:
Property1.objects.filter(pk__in=table.property_set.all())
Related
I have a class using django.db as follows:
from django.db import models
class myClass(models.Model):
"My class"
class Meta:
db_table = 'my_table'
field1 = models.CharField()
# more fields here
def some_function(self):
return "hello"
I'd like to turn this into a generic class that does not have "my_table" hardcoded, so that I can use it in some way such as:
class GenericClass(models.Model):
class Meta:
db_table = None
# genericClass properties and methods here
class myClass(GenericClass):
id = models.CharField(max_length=20)
def __init__(self, *args, **kwargs):
super(self.__class__, self).Meta.db_table = 'my_table'
super(self.__class__, self).__init__(*args, **kwargs)
However, this gives me the following error (using Python 2.7 / Django 1.11):
Local field u'id' in class 'myClass' clashes with field of similar name from base class 'GenericClass'
Is there any way to create a generic class that extends django.db.models.Model, that supports a non-default database table name without hardcoding Meta.db_table?
I guess what you are asking is a Django equivalent of AbstractSuperClass / SuperClass.
If what you want is AbstractSuperClass for django model you can define it like
class AbstractModel(models.Model):
field1 = ...
field2 = ...
...
class Meta:
abstract = True
class SubClassModel(models.Model):
field3 = ...
class Meta:
db_table=sub_class_db_table
Your AbstractModel won't have a database table. It simply acts as a superclass to store common fields (and even you can define common functions).
If you DESC your sub_class_db_table, you will find field1, field2, field3.
See this for more details.
If you want to create a db_table hierarchy itself (field1 and field2 will be in the super_class_table and field3 will be in sub_class_table), you can use Multi-table inheritance. This will be useful especially if there is a service-contract scenario - A team or a module owns common functionality including the db_table, its maintenance. But this way is less commonly used.
I have two models which are used with a database I don't control. Both are set with managed = False. The first model has a field which is a foreign key to the second model, but it's implemented as a CharField, not as a ForeignKey.
Is it possible to use select_related on the first model to access properties of the key'd second model?
Here's an example:
class Foo(models.Model):
class Meta:
managed = False
fieldone = models.CharField(max_length=10)
myfk = models.CharField(max_length=20) # In practice, this points to Bar.localkey
class Bar(models.Model):
class Meta:
managed = False
localkey = models.CharField(max_length=20)
someotherattribute = models.CharField(max_length=100)
Foo.objects.all().select_related('Bar') # I know this won't work, but is there something that will?
No, because there's nothing related.
But if you (or someone for some reason) have stored the ID (or some unique value such as localkey) from the 'related' object, you could perform a filter based on it.
foo = Foo.objects.first() # Pick one Foo object
foo_bar = Bar.objects.get(localkey=foo.myfk)
To make this looks like select_related you could try this:
class Foo(models.Model):
class Meta:
managed = False
fieldone = models.CharField(max_length=10)
myfk = models.CharField(max_length=20)
def bar(self):
return Bar.objects.get(localkey=self.myfk)
# probably you will need to manage common error when performing a .get()
# DoesNotExist and MultipleObjectsReturned
Then use like this:
foos = Foo.objects.all()
for foo in foos:
print foo.bar()
I am not sure if this is a good idea but you could decorate .bar() method as a property:
...
#property
def bar(self):
return Bar.objects.get(localkey=self.myfk)
And then call it like this:
foo # some random Foo object
foo.bar # this should return the 'related' Bar object
Here's my attempt at a generalized natural key model manager. It's like the docs except it tries (unsuccessfully) to determine the natural key field names from the Meta.unique_together attribute.
class NaturalKeyModelManager(Manager):
def get_by_natural_key(self, *args):
field_dict = {}
for i, k in enumerate(self.model.Meta.unique_together[0]):
field_dict[k] = args[i]
return self.get(**field_dict)
If I insert a debug print just before the for loop like this:
print dir(self.model.Meta)
it doesn't list the unqiue_together attribute at all:
['__doc__', '__module__', 'abstract']
The 'abstract' bit worried me, but another debug print shows that the model I'm trying manage with natural keys is not abstract:
>>> print self.model.Meta.abstract
False
I am mixing in a lot of abstract base classes. Could that be the problem?
class MixedModel(NamedModel, TimeStampedModel, VersionedModel, Model):
objects = NaturalKeyModelManager()
class Meta:
unique_together = (('name', 'version',),)
For completeness here's one of the mixins:
class TimeStampedModel(Model):
created = DateTimeField(_("Created"), auto_now_add=True, null=True, editable=False)
updated = DateTimeField(_("Updated"), auto_now=True, null=True, editable=True)
class Meta:
abstract = True
The hard-coded model manager works just fine:
class MixedModelManager(Manager):
def get_by_natural_key(self, name, version):
return self.get(name=name, version=version)
In order to get the actual options passed to meta, you should use self.model._meta rather than self.model.Meta
I have a Django Model that looks like this:
class MyModel(models.Model):
field1 = models.IntegerField()
field2 = models.IntegerField()
nonDbField = SomeObject()
objects = MyCustomManager()
field1 is actually a PK to an abstract class of SomeObject.
I want a custom manager that for every value returned by any of the functions (all, filter, get, etc) does the following:
value.nonDbField = SomeObject.objects.get(pk=value.field1)
I've tested that I can manually override get like so:
class MyCustomManager(models.Manager):
def get(self, *args, **kwargs):
value = super(MyCustomManager, self).get(*args, **kwargs)
value.nonDbField = SomeObject.objects.get(listid=value.itemListID)
return value
but wondered if there was an easier way to do it across all functions.
There's going to be plenty of you that will say, "Why are you doing this?". It has to do with a model inheritance of a legacy, but still active database.
If you need nonDbField's value to be related to the field1 (or any other field in the model) you can try something like this:
Class MyModel(models.Model):
# your fields here...
def _nonDbField(self):
return SomeObject.objects.get(pk=self.field1)
nonDbField = property(_nonDbField)
This allows you to do something like this:
MyModel.objects.get(pk=1).nonDbField
Keep in mind that you are making a database query each time you access nonDbField (which may or may not be detrimental to your DB performance).
you can use property for your calculated fields
Class MyModel(models.Model):
# your fields here...
first_name = models.CharField()
last_name = models.CharField()
#property
def fullname(self):
return f"{self.first_name} {self.last_name}"
This allows you to do something like this:
obj = MyModel.objects.get(pk=1)
print(obj.fullname)
I have a normal model and an abstract model like so:
class TaggedSubject(models.Model):
user = models.ForeignKey(User, null=True, blank=True)
category = models.CharField(max_length=200)
foo = models.CharField(max_length=50)
bar = models.CharField(max_length=50)
# etc
content_type = models.ForeignKey(ContentType)
content_object_pk = models.CharField(max_length=255)
content_object = generic.GenericForeignKey("content_type", "content_object_pk")
def __unicode__(self):
if self.user:
return "%s" % (self.user.get_full_name() or self.user.username)
else:
return self.label
class Taggable(models.Model):
tagged_subjects = generic.GenericRelation(TaggedSubject, content_type_field='content_type', object_id_field='content_object_pk')
#property
def tagged_users(self):
return User.objects.filter(pk__in=self.tagged_subjects.filter(user__isnull=False).values("user"))
class Meta:
abstract = True
The Taggable abstract model class then gets used like so:
class Photo(Taggable):
image = models.ImageField(upload_to="foo")
# ... etc
So if we have a photo object:
photo = Photo.objects.all()[0]
I can all the users tagged in the photo with photo.tagged_users.all()
I want to add the inverse relation to the user object, so that if I have a user:
user = User.objects.filter(pk__in=TaggedSubject.objects.exclude(user__isnull=True).values("user"))[0]
I can call something like user.tagged_photo_set.all() and have it return all the photo objects.
I suspect that since TaggedSubject connects to the Taggable model on a generic relation that it won't be possible to use it as a through model with a ManyToMany field.
Assuming this is true, this is the function I believe I'd need to add (somehow) to the User model:
def tagged_photo_set(self):
Photo.objects.filter(pk__in=TaggedSubject.objects.filter(user=self, content_type=ContentType.objects.get_for_model(Photo))
I'm wondering if it's possible to set it up so that each time a new model class is created based on Taggable, it creates a version of the function above and adds it (ideally as a function that behaves like a property!) to User.
Alternatively, if it is somehow possible to do ManyToMany field connections on a generic relation (which I highly doubt), that would work too.
Finally, if there is a third even cooler option that I am not seeing, I'm certainly open to it.
You could use add_to_class and the class_prepared signal to do some post processing when models subclassing your base class are set up:
def add_to_user(sender, **kwargs):
def tagged_FOO_set(self):
return sender.objects.filter(pk__in=TaggedSubject.objects.filter(
user=self,
content_type=ContentType.objects.get_for_model(sender)))
if issubclass(sender, MyAbstractClass):
method_name = 'tagged_{model}_set'.format(model=sender.__name__.lower())
User.add_to_class(method_name, property(tagged_FOO_set))
class_prepared.connect(add_to_user)