Django ORM Query Builder

The Django ORM (Object-Relational Mapper) translates Python method calls into SQL queries. You work with Python classes (models) and chain methods like .filter(), .exclude(), .annotate() — Django generates the SQL automatically. This lets you write database queries without raw SQL while staying portable across PostgreSQL, MySQL, SQLite, and Oracle.

filter() returns a QuerySet (potentially multiple objects) and never raises an exception if no match is found — it returns an empty QuerySet. get() returns a single object and raises DoesNotExist if nothing matches or MultipleObjectsReturned if more than one matches. Use get() only when you expect exactly one result (e.g., lookup by primary key or unique field).

Double underscores are the Django ORM's field lookup separator. They enable two things: (1) field traversal across relationships — e.g., author__name filters via a ForeignKey to the related model's name field; (2) lookup types — e.g., name__icontains for case-insensitive LIKE, age__gte for greater-than-or-equal. You can chain them: order__customer__country__iexact='US'.

select_related performs a SQL JOIN and fetches related objects in a single query — best for ForeignKey and OneToOne relations. prefetch_related performs a separate query for each related model and joins in Python — best for ManyToMany and reverse ForeignKey relations. Both avoid the N+1 query problem where you'd otherwise execute one query per row to fetch related data.

Q() objects let you build complex OR/AND/NOT conditions in Django querysets. Normal .filter() chains are always AND conditions. For OR: Model.objects.filter(Q(status='active') | Q(is_admin=True)). For NOT: filter(~Q(status='deleted')). Q objects can be combined with &, |, and ~ operators and passed as arguments to filter(), exclude(), or get().

values() returns a QuerySet of dictionaries instead of model instances — useful when you only need specific fields and want to avoid instantiating full objects. values_list() returns tuples instead of dicts. Both accept field names to restrict which columns are fetched. values_list('id', flat=True) returns a flat list of IDs — very efficient for bulk operations.