/django-indices

Create Postgres JSON Indexes concurrently

Primary LanguagePythonMIT LicenseMIT

Create JSON Indexes concurrently

The functionality provided by this Django App is similar to the indexes, except that it provides an option to create indexes concurrently.

Currently, only Postgres JSON Indexes are supported.

Installation

Install the latest version using pip:

pip install django_indices

Add django_indices to INSTALLED_APPS in the django settings file:

INSTALLED_APPS = (
    ...,
    "django_indices",
)

Run the migrations:

python manage.py migrate django_indices

Add the indices you want to the Meta class of the required Model. Then run the migrate_indices management command to add / remove the indices:

python manage.py migrate_indices

Example

from django.db import models
from django.db.models import Q
from django.contrib.postgres.fields import JSONField

from django_indices import JSONIndex

class Inventory(models.Model):
    data = JSONField()

    class Meta:
        indices = [
            # Creates an index on data->'product_type'
            # Indexes are created concurrently by default, which doesn't lock
            # the table.
            JSONIndex(
                field=["data", "product_type"],
                name="idx_inventory_product_type",
            ),

            # Creates a unique index on data->'product'->>'id'
            #
            # concurrently=False ensures that the index creation will execute
            # in a transaction. So the index is either created successfully,
            # or fails cleanly. This will lock the table though.
            JSONIndex(
                field=["data", "product", "id"],
                name="idx_inventory_product_id",
                unique=True,
                concurrently=False,
            ),

            # Creates a partial compound index on
            # (data->'product'->>'brand', data->'product'->>'size')
            # WHERE data->>'product_type' = 'shoe' OR data->>'product_type' = 'crocs'
            JSONIndex(
                field=(["data", "product", "brand"], ["data", "product", "size"]),
                name="idx_inventory_footwear_brand_size",
                where=(Q(data__product_type="shoe") | Q(data__product_type="crocs"),
            ),

        ]