QuerySets and the ORM

Published by Programming Line in Django. Last updated .

Django's ORM builds SQL through composable QuerySets. QuerySets are lazy: filtering constructs a query description, and database work begins only when the results are evaluated.

This lesson targets Django 6.1 and modern Python. It focuses on the framework contract behind the feature so that you can adapt the example rather than copying it without understanding the request, data, and security boundaries involved.

How QuerySets and the ORM Works

filter and exclude return new QuerySets, get returns exactly one object, and Q objects express grouped Boolean logic. select_related joins single-valued relationships, while prefetch_related performs a separate query for collections and combines results in Python.

Django favors explicit configuration and reusable components. Follow the data from the incoming request through routing and view logic to the response, and keep database access, validation, and presentation in their appropriate layers. This separation improves both testability and security review.

Core Syntax

The first example shows the smallest useful shape of the feature:

from django.db.models import Q
from .models import Course

results = (
    Course.objects
    .filter(Q(title__icontains="python") | Q(title__icontains="django"))
    .filter(published=True)
    .order_by("title")
)

Place the code in the file indicated by its comments or surrounding lesson text. Django imports project modules during startup, so syntax and import paths must be correct before the development server can route a request. Prefer named objects and readable configuration over clever indirection.

Practical Example

from django.db.models import Count
from .models import Course

courses = (
    Course.objects
    .filter(published=True)
    .annotate(lesson_count=Count("lessons"))
    .prefetch_related("lessons")
)

This example adds the structure commonly required in a real project. Read it from top to bottom and identify the public input, the validation or query boundary, and the returned result. In a running project, verify it with a focused test before relying on browser-only manual testing.

Execution Step by Step

  1. Django loads settings and builds the application registry.
  2. The server converts an HTTP request into an HttpRequest object.
  3. URL resolution selects a view and supplies captured parameters.
  4. The view validates input and collaborates with forms, models, or services as needed.
  5. The view returns an HttpResponse, often produced by rendering a template.
  6. Middleware processes the response before the server sends it to the client.

Not every feature participates visibly in every step, but the same lifecycle helps locate bugs. A 404 usually points to routing or lookup behavior, a form error belongs to validation, and repeated database work can often be found where a QuerySet is evaluated.

Common Mistakes

  • Putting secrets in settings.py or committing a production environment file.
  • Trusting a hidden form field, URL parameter, or template condition as an authorization check.
  • Calling the database repeatedly from a loop instead of shaping the QuerySet for the required relationships.
  • Editing a generated migration after it has already been applied in a shared environment.

Django’s defaults help only when framework APIs are used correctly. Raw SQL, mark_safe, csrf_exempt, DEBUG in production, or an unrestricted ALLOWED_HOSTS setting deliberately remove protections and require careful justification.

Best Practices

  • Keep views small and move reusable domain behavior into well-named functions or model methods.
  • Use named URLs and reverse them instead of hard-coding internal links.
  • Validate all external input and enforce authorization on the server.
  • Write tests for success, invalid input, anonymous access, and forbidden access.
  • Run system checks, migrations, and dependency updates as controlled release steps.

Practice Exercises

  • Rebuild the first QuerySets and the ORM example in a fresh app and explain the role of every imported object.
  • Add one invalid-input or unauthorized-user test and assert the exact status code.
  • Extend the practical example without hard-coding a URL or exposing an unrestricted database query.

Summary

  • QuerySets and the ORM fits into Django’s request, routing, data, and response lifecycle.
  • Explicit validation and server-side authorization are essential even when the interface hides an action.
  • Small components, named routes, deliberate QuerySets, and automated tests make Django projects easier to maintain.