Django ORM Relationships Cheat Sheet
A cheat sheet for Django ORM relationships — version 2.1
I want to start this story with saying thanks to Mahmoud Zalt, who published a really useful article “Eloquent Relationships Cheat Sheet” about a year ago and also gave me permission to use the same structure/images/examples in my article.
- One to One Relationship
- One to Many Relationship
- Many to Many Relationship
- Polymorphic One to Many Relationship — Generic Relations
- Polymorphic One to Many Relationship — django-polymorphic package
- Polymorphic Many to Many Relationship — django-polymorphic package
One to One Relationship
Demo details:
In this demo we have 2 models (Owner and Car), and 2 tables (owners and cars).
Business Rules:
The Owner can own one Car.
The Car can be owned by one Owner.
Relations Diagram:
Relationship Details:
The Cars table should store the Owner ID.
Models:
class Owner(models.Model):
#...
name = models.CharField(max_length=255)class Car(models.Model):
#...
name = models.CharField(max_length=255)
owner = models.OneToOneField(
Owner,
on_delete=models.CASCADE,
related_name='car'
)
Store Records:
car = Car.objects.get(id=1)
owner = Owner.objects.get(id=1)# Create relation between Owner and Car.owner.car = car
owner.car.save()# Create relation between Car and Owner.car.owner = owner
car.save()
Retrieve Records:
# Get Owner Carowner.car
# Get Car Ownercar.owner;
One to Many Relationship
Demo details:
In this demo we have 2 models (Thief and Car), and 2 tables (thieves and cars).
Business Rules:
The Thief can steal many Cars.
The Car can be stolen by one Thief.
Relations Diagram:
Relationship Details:
The Cars table should store the Thief ID.
Models:
class Thief(models.Model):
# ...
name = models.CharField(max_length=255)
class Car(models.Model):
# ...
name = models.CharField(max_length=255)
thief = models.ForeignKey(
Thief,
on_delete=models.CASCADE,
related_name='cars'
)
Store Records:
thief = Thief.objects.get(id=1)
car1 = Car.objects.get(id=1)
...# Create relation between Thief and Car.thief.cars.add(car1,car2, car3)# Create relation between Car and Thief.car.thief = thief
car.save()# When we creating new car :
car = Car(name = 'test name', thief=thief)
car.save()
Retrieve Records:
# Get Thief Car thief.cars.all()# Get Car Thiefcar.thief
Many to Many Relationship
Demo details:
In this demo we have 2 models (Driver and Car), and 3 tables (drivers, cars and a pivot table named car_drivers).
Business Rules:
The Driver can drive many Cars.
The Car can be driven by many Drivers.
Relations Diagram:
Relationship Details:
The Pivot table “car_driver” should store the Driver ID and the Car ID.
Models:
class Driver(models.Model):
# ...
name = models.CharField(max_length=255)class Car(models.Model):
# ...
name = models.CharField(max_length=255)
drivers = models.ManyToManyField(
Driver,
related_name='cars'
)
Store Records:
# Create relation between Driver and Car.driver = Driver.objects.get(id=1)
car1 = Car.objects.get(id=1)
car2 = Car.objects.get(id=2)driver.cars.add(car1,car2)# Create relation between Car and Driver.
car = Car.objects.get(id=1)
driver1 = Driver.objects.get(id=2)
driver2 = Driver.objects.get(id=3)
car.drivers.add(driver1, driver2)
Retrieve Records:
# Get Driver Cardriver.cars.all()# Get Car Driverscar.drivers.all()
Polymorphic One to Many Relationship
With Django’s Generic Relations
Demo details:
In this demo we have 3 models (Man, Woman and Car), and 3 tables (men, women and cars).
Business Rules:
The Man (buyer) can buy many Cars.
The Woman (buyer) can buy many Cars.
The Car can be bought by one buyer (Man or Woman).
Relations Diagram:
Relationship Details:
The Car table should store the Buyer ID and the Buyer table should store the relation between ID and Type.
Models:
from django.contrib.contenttypes.fields import GenericForeignKey
from django.contrib.contenttypes.fields import GenericRelation
from django.contrib.contenttypes.models import ContentTypeclass Car(models.Model): # ...
name = models.CharField(max_length=255) content_type = models.ForeignKey(ContentType, on_delete=models.CASCADE)
object_id = models.PositiveIntegerField()
content_object = GenericForeignKey()class Woman(models.Model): # ...
name = models.CharField(max_length=255)
cars = GenericRelation(Car)class Man(models.Model): # ...
name = models.CharField(max_length=255)
cars = GenericRelation(Car)
Store Records:
man = Man.objects.get(id=1)
woman = Woman.objects.get(id=1)# Create relation between buyer (Man/Woman) and Car.car = Car.objects.get(id=1)
woman.cars.add(car)# Create relation between Car and buyer (Men/Women).man = Man.objects.get(id=1)
woman = Woman.objects.get(id=1)c = Car(name = 'test name',content_object=man)
c.save()c = Car(name = 'test name',content_object=woman)
c.save()
Retrieve Records:
# Get buyer (Man/Woman) Carsman.cars.all()
woman.cars.all()# Get Car buyer (Man and Woman)car.content_object
Polymorphic One to Many Relationship
With package — django-polymorphic.
Demo details:
In this demo we have 4 models (Buyer, Man, Woman and Car), and 4 tables (buyer, men, women and cars).
Business Rules:
The Man (buyer) can buy many Cars.
The Woman (buyer) can buy many Cars.
The Car can be bought by one buyer (Man or Woman).
Relations Diagram:
Relationship Details:
The Car table should store the Buyer ID and the Buyer table should store the relation between ID and Type.
Models:
from polymorphic.models import PolymorphicModelclass Buyer(PolymorphicModel):
pass
class Woman(Buyer):
# ...
name = models.CharField(max_length=255)
class Man(Buyer):
# ...
name = models.CharField(max_length=255)
class Car(models.Model):
# Fields
name = models.CharField(max_length=255)
buyer = models.ForeignKey(
Buyer,
on_delete=models.CASCADE,
related_name='cars'
)
Store Records:
man = Buyer.objects.get(id=1) # or Man.objecets.get(id=1)
woman = Buyer.objects.get(id=2) # or Woman.objecets.get(id=2)# Create relation between buyer (Man/Woman) and Car.man.cars.add(car1, car2)woman.cars.add(car1, car2)# Create relation between Car and buyer (Men/Women).c = Car(name = 'test name', buyer = man)
c.save()c = Car(name = 'test name', buyer = woman)
c.save()
Retrieve Records:
# Get buyer (Man/Woman) Carsman.cars.all()
woman.cars.all()# Get Car buyer (Man and Woman)car.buyer
Polymorphic Many to Many Relationship
With package — django-polymorphic.
Demo details:
In this demo we have 3 models (Valet, Owner and Car), and 4 tables (valets, owners, cars and drivers).
Business Rules:
The Valet (driver) can drive many Cars.
The Owner (driver) can drive many Cars.
The Car can be driven by many drivers (Valet or/and Owner).
Relations Diagram:
Relationship Details:
The Pivot table “drivers” should store the Driver ID, Driver Type and the CarID.
“driver” is a name given to a group of models (Valet and Owner). And it’s not limited to two. The driver type is the real name of the model.
Models:
from polymorphic.models import PolymorphicModelclass Driver(PolymorphicModel):
pass
class Owner(Driver):
# ...
name = models.CharField(max_length=255)
class Valet(Driver):
# ...
name = models.CharField(max_length=255)
class Car(models.Model):
# ...
name = models.CharField(max_length=255)
drivers = models.ManyToManyField(
Driver,
related_name='cars'
)
Store Records:
# Create relation between driver (Valet/Owner) and Car.owner.cars.add(car1, car2)# Create relation between Car and driver (Valet/Owner).car.drivers.add(owner, valet)
Retrieve Records:
# Get driver (Valet/Owner) Carsvalet.cars.all()
owner.cars.all()# Get Car drivers (Valet and Owner)car.drivers.all()
car1.drivers.instance_of(Valet)
car1.drivers.instance_of(Owner)
I also highly recommend to read the article, that describes polymorphic many to many relationships with another django package (django-gm2m).
Thanks for reading and stay in touch.