[Django]-How to seed Django project ? – insert a bunch of data into the project for initialization

Similar to Rails, we also have option to seed the database. It is done using management commands. In one of your apps, use the following folder structure

<project>/<app>/management/commands/seed.py

this makes python manage.py seed available as a management command. I personally follow the following structure.

# <project>/<app>/management/commands/seed.py
from django.core.management.base import BaseCommand
import random

# python manage.py seed --mode=refresh

""" Clear all data and creates addresses """
MODE_REFRESH = 'refresh'

""" Clear all data and do not create any object """
MODE_CLEAR = 'clear'

class Command(BaseCommand):
    help = "seed database for testing and development."

    def add_arguments(self, parser):
        parser.add_argument('--mode', type=str, help="Mode")

    def handle(self, *args, **options):
        self.stdout.write('seeding data...')
        run_seed(self, options['mode'])
        self.stdout.write('done.')


def clear_data():
    """Deletes all the table data"""
    logger.info("Delete Address instances")
    Address.objects.all().delete()


def create_address():
    """Creates an address object combining different elements from the list"""
    logger.info("Creating address")
    street_flats = ["#221 B", "#101 A", "#550I", "#420G", "#A13"]
    street_localities = ["Bakers Street", "Rajori Gardens", "Park Street", "MG Road", "Indiranagar"]
    pincodes = ["101234", "101232", "101231", "101236", "101239"]

    address = Address(
        street_flat=random.choice(street_flats),
        street_locality=random.choice(street_localities),
        pincode=random.choice(pincodes),
    )
    address.save()
    logger.info("{} address created.".format(address))
    return address

def run_seed(self, mode):
    """ Seed database based on mode

    :param mode: refresh / clear 
    :return:
    """
    # Clear data from tables
    clear_data()
    if mode == MODE_CLEAR:
        return

    # Creating 15 addresses
    for i in range(15):
        create_address()

In above structure you could, add custom modes, and seed accordingly. Also you could add additional management command arguments (e.g. number_of_addresses and pass it to run seed. the command would be python manage.py seed --mode=refresh --number_of_addresses=15).

Hope this helps. Cheers!

According to django 4.2 docs You can seed your database by creating fixtures.
The most straightforward way of creating a fixture if you’ve already got some data is to use the manage.py dumpdata command. e.g.

python3 manage.py dumpdata <app_name>.<model_name>  > ./fixtures/<model_name>.json

Or, you can write fixtures by hand; fixtures can be written as JSON, XML or YAML.

As an example, though, here’s what a fixture for a simple Person model might look like in JSON:

[
  {
    "model": "myapp.person",
    "pk": 1,
    "fields": {
      "first_name": "John",
      "last_name": "Snow"
    }
  },
  {
    "model": "myapp.person",
    "pk": 2,
    "fields": {
      "first_name": "Paul",
      "last_name": "McCartney"
    }
  }
]

And here’s that same fixture as YAML:

- model: myapp.person
  pk: 1
  fields:
    first_name: John
    last_name: Snow
- model: myapp.person
  pk: 2
  fields:
    first_name: Paul
    last_name: McCartney

You’ll store this data in a fixtures directory inside your app.

Loading data is easy: just call manage.py loaddata <fixturename>, where <fixturename> is the name of the fixture file you’ve created.

I am use this lib

https://pypi.org/project/django-seed/

from django_seed import Seed

seeder = Seed.seeder()

from myapp.models import Game, Player
seeder.add_entity(Game, 5)
seeder.add_entity(Player, 10)

inserted_pks = seeder.execute()

There are multiple ways to provide seed data to Django:

If you are prototyping and don’t care yet about maintaining consistency across environments, using Django’s dumpdata and loaddata is very easy and fast:

$ ./manage.py dumpdata myApp.MyModel [myApp.AnotherModel] --output <myfile>
$ ./manage.py loaddata <path-to-myfile>

The fixture files produced can be different formats (json by default) so they are easy to edit or generate and they include information about which app/model they reference, so they are easy to load.

If you need to maintain consistency across environments, either:

Use data migrations:

./manage.py makemigrations --empty <myapp> --name seed_<my_model>

Or add a custom management command per @anurag’s answer

You can create a data migration…. https://docs.djangoproject.com/en/2.0/topics/migrations/#data-migrations

python manage.py makemigrations --empty appname