Querying and Filtering

DRF provides tools to efficiently query, filter, and optimize data retrieval from the database. These tools ensure that APIs deliver only the required data while maintaining performance.

Filtering and Ordering

Overview

Filtering and ordering allow you to return subsets of data or sort results based on specific fields.

Using filter_backends

• DRF provides built-in filter backends for basic filtering, search, and ordering.
• Specify filter_backends in the view or globally in the settings.

Example Setup

1. Enable filter backends:

   from rest_framework.filters import OrderingFilter
   
   class MyViewSet(viewsets.ModelViewSet):
       queryset = MyModel.objects.all()
       serializer_class = MySerializer
       filter_backends = [OrderingFilter]
       ordering_fields = ['field1', 'field2']  # Specify fields allowed for ordering

2. Global configuration:

   REST_FRAMEWORK = {
       'DEFAULT_FILTER_BACKENDS': [
           'rest_framework.filters.OrderingFilter',
       ],
   }

Filtering

1. Simple Filtering: Use django_filters to apply simple or complex filtering based on query parameters.

   from django_filters.rest_framework import DjangoFilterBackend
   
   class MyViewSet(viewsets.ModelViewSet):
       queryset = MyModel.objects.all()
       serializer_class = MySerializer
       filter_backends = [DjangoFilterBackend]
       filterset_fields = ['field1', 'field2']  # Specify fields to filter
   • Example URL: /api/items/?field1=value

2. Custom Filtering: Define a custom filter backend for advanced filtering logic.

   from rest_framework.filters import BaseFilterBackend
   
   class CustomFilterBackend(BaseFilterBackend):
       def filter_queryset(self, request, queryset, view):
           if 'custom_param' in request.query_params:
               return queryset.filter(custom_field=request.query_params['custom_param'])
           return queryset
   
   class MyViewSet(viewsets.ModelViewSet):
       filter_backends = [CustomFilterBackend]

Ordering

• Add an OrderingFilter backend and specify ordering_fields or allow all fields.

  from rest_framework.filters import OrderingFilter
  
  class MyViewSet(viewsets.ModelViewSet):
      queryset = MyModel.objects.all()
      serializer_class = MySerializer
      filter_backends = [OrderingFilter]
      ordering_fields = ['field1', 'field2']  # Fields for ordering
      ordering = ['field1']  # Default ordering
  • Example URL: /api/items/?ordering=field1

Pagination

Overview

Pagination splits large datasets into smaller chunks, making responses manageable and efficient.

DRF Pagination Classes

1. PageNumberPagination:

   • Default pagination using page numbers.

   from rest_framework.pagination import PageNumberPagination
   
   class MyPagination(PageNumberPagination):
       page_size = 10  # Number of items per page
       page_size_query_param = 'page_size'  # Allow clients to set page size
       max_page_size = 100  # Maximum allowed page size

2. LimitOffsetPagination:

   • Allows clients to specify an offset and limit.

   from rest_framework.pagination import LimitOffsetPagination
   
   class MyPagination(LimitOffsetPagination):
       default_limit = 10
       max_limit = 100

3. CursorPagination:

   • Uses an opaque cursor for pagination, ensuring stable ordering.

   from rest_framework.pagination import CursorPagination
   
   class MyPagination(CursorPagination):
       page_size = 10
       ordering = 'created_at'  # Field used for cursor

Using Custom Pagination

Assign the pagination class to a view:

class MyViewSet(viewsets.ModelViewSet):
    queryset = MyModel.objects.all()
    serializer_class = MySerializer
    pagination_class = MyPagination

Global Pagination

Set pagination globally in settings.py:

REST_FRAMEWORK = {
    'DEFAULT_PAGINATION_CLASS': 'rest_framework.pagination.PageNumberPagination',
    'PAGE_SIZE': 10,
}

Searching

Overview

DRF provides a SearchFilter backend for simple search functionality using query parameters.

Adding Search to a View

1. Include the SearchFilter backend:

   from rest_framework.filters import SearchFilter
   
   class MyViewSet(viewsets.ModelViewSet):
       queryset = MyModel.objects.all()
       serializer_class = MySerializer
       filter_backends = [SearchFilter]
       search_fields = ['name', 'description']  # Fields to search

2. Search query example:

   • URL: /api/items/?search=keyword

Customizing Search

1. Partial Matching:

   • Use ^ (starts with), @ (search), or $ (regex) to control field behavior:

     search_fields = ['^name', '=description', '@tags']

2. Case-Insensitive Search:

   • Apply filters like icontains in a custom backend.

Full-Text Search

For advanced searching, integrate with third-party libraries like Elasticsearch, Solr, or Django’s PostgreSQL full-text search.

Advanced Query Optimizations

Overview

Query optimizations reduce database queries and improve performance when fetching related objects.

Techniques

1. select_related:

   • Fetches related objects in a single query for ForeignKey relationships.

   queryset = MyModel.objects.select_related('related_field')

2. prefetch_related:

   • Fetches related objects for ManyToMany or reverse ForeignKey relationships.

   queryset = MyModel.objects.prefetch_related('related_field')

3. Using Annotate and Aggregate:

   • Add computed values to a query to reduce separate calculations.

   from django.db.models import Count
   
   queryset = MyModel.objects.annotate(related_count=Count('related_field'))

4. Avoid N+1 Query Problem:

   • Avoid queries within loops by preloading related data.

   queryset = MyModel.objects.select_related('related_field')
   for obj in queryset:
       print(obj.related_field.name)  # Avoids additional queries

5. Use QuerySet Caching:

   • Cache frequently accessed data using Django’s caching framework.

6. Index Database Fields:

   • Add database indexes for fields frequently used in filtering or ordering.

Example of Optimized QuerySet

queryset = MyModel.objects.select_related('foreign_key_field').prefetch_related('many_to_many_field').annotate(related_count=Count('related_field'))