A collection of anything from basics to advanced recommended methods and usages with Django REST Framework for creating browsable and awesome web API's. This could also serve as a quick reference guide.
Here is DRF's official documentation in case you need everything in detail.
Summarized from the official docs:
- Web browsable API
- Serialization that supports ORM and non-ORM data sources.
- Authentication is easily implemented.
- Highly customizable in every sense of the word.
Throughout this cheat-sheet, the examples used will be based on this Post and Comment model.
class Post(models.Model):
title = models.CharField(max_length=1024)
text = models.TextField()
created = models.DateField(auto_now_add=True)
class Comment(models.Model):
post = models.ForeignKey(Post, related_name='comments')
user = models.ForeignKey(User)
text = models.TextField()Install the package via pip:
$ pip install djangorestframework
Then, add 'rest_framework' to INSTALLED_APPS in your settings.py file.
INSTALLED_APPS = [
# Rest of your installed apps ...
'rest_framework',
]Serializers allow complex data like querysets and model instances to be converted to native Python datatypes that can then be easily rendered into JSON, XML, and other formats.
Suppose we wanted to create a PostSerializer for our example Post model and CommentSerializer for our Comment model.
class PostSerializer(serializers.ModelSerializer):
class Meta:
model = Post
fields = ('id', 'title', 'text', 'created')
class CommentSerializer(serializers.ModelSerializer):
class Meta:
model = Comment
fields = ('post', 'user', 'text')Or also you could use exclude to exclude certain fields from being seialized. ModelSerializer has default implementations for the create() and update() methods.
By default, instances are serialized with primary keys to represent relationships. To get nested serialization we could use, General or Explicit methods.
Using depth parameter.
class CommentSerializer(serializers.ModelSerializer):
class Meta:
model = Comment
fields = '__all__'
depth = 2You can also define and nest serializers within eachother...
class CommentSerializer(serializers.ModelSerializer):
post = PostSerializer()
class Meta:
model = Comment
fields = '__all__'So here, the comment's post field (how we named it in models.py) will serialize however we defined it in PostSerializer.
This makes your web API a lot more easy to use (in browser) and would be a nice feature to add.
Lets say we wanted to see the comments that every post has in each of the Post instances of our API. With HyperlinkedModelSerializer, instead of having nested primary keys or nested fields, we get a link to each individual Comment (url).
class PostSerializer(serializers.HyperlinkedModelSerializer):
class Meta:
model = Post
fields = ('id', 'title', 'text', 'created', 'comments')
read_only_fields = ('comments',)Note: without the read_only_fields, the create form for Posts would always require a comments input, which doesn't make sense (comments on a post are normally made AFTER the post is created).
Another way of hyperlinking is just adding a HyperlinkedRelatedField definition to a normal serializer.
class PostSerializer(serializers.ModelSerializer):
comments = serializers.HyperlinkedRelatedField(many=True, view_name='comment-detail', read_only=True)
class Meta:
model = Post
fields = ('id', 'title', 'text', 'created', 'comments')This makes your web API a lot more easy for extract limited number of parameter in response. Let's say you want to set which fields should be used by a serializer at the point of initialization.
Just copy below code and past it in your serliazer file
class DynamicFieldsModelSerializer(serializers.ModelSerializer):
def __init__(self, *args, **kwargs):
# Don't pass the 'fields' arg up to the superclass
fields = kwargs.pop('fields', None)
# Instantiate the superclass normally
super(DynamicFieldsModelSerializer, self).__init__(*args, **kwargs)
if fields is not None:
# Drop any fields that are not specified in the `fields` argument.
allowed = set(fields)
existing = set(self.fields.keys())
for field_name in existing - allowed:
self.fields.pop(field_name)Extend DynamicFieldsModelSerializer from your serializer class
class UserSerializer(DynamicFieldsModelSerializer):
class Meta:
model = User
fields = ('id', 'username', 'email')Mention the fields name inside fields
UserSerializer(user, fields=('id', 'email'))Here, you will get only idand email from serializer instead of all.
There are many options for creating views for your web API, it really depends on what you want and personal preference.
We'll write a post_list view with GET and POST methods for reading and creating new Post instances.
from rest_framework.decorators import api_view
from rest_framework.response import Response
from rest_framework.parsers import JSONParser
from rest_framework import status
from posts.models import Post
from posts.serializers import PostSerializer
@api_view(['GET', 'POST'])
def post_list(request, format=None):
if request.method == 'GET':
posts = Post.objects.all()
serializer = PostSerializer(posts, many=True)
return Response(serializer.data)
elif request.method == 'POST':
data = JSONParser().parse(request)
serializer = PostSerializer(data=data)
if serializer.is_valid():
serializer.save()
return Response(serializer.data, status=status.HTTP_201_CREATED)
return Response(serializer.errors, status=status.HTTP_400_BAD_REQUEST)from rest_framework.response import Response
from rest_framework import status
from rest_framework.views import APIView
from posts.models import Post
from posts.serializers import PostSerializer
class PostList(APIView):
def get(self, request, format=None):
snippets = Post.objects.all()
serializer = PostSerializer(snippets, many=True)
return Response(serializer.data)
def post(self, request, format=None):
serializer = PostSerializer(data=request.data)
if serializer.is_valid():
serializer.save()
return Response(serializer.data, status=status.HTTP_201_CREATED)
return Response(serializer.errors, status=status.HTTP_400_BAD_REQUEST)from rest_framework import generics
from posts.models import Post
from posts.serializers import PostSerializer
class PostList(generics.ListCreateAPIView):
queryset = Post.objects.all()
serializer_class = PostSerializerfrom rest_framework import generics, mixins
from posts.models import Post
from posts.serializers import PostSerializer
class PostList(generics.GenericAPIView,
mixins.ListModelMixin,
mixins.CreateModelMixin
):
queryset = Post.objects.all()
serializer_class = PostSerializer
def get(self, request, *args, **kwargs):
return self.list(request, *args, **kwargs)
def post(self, request, *args, **kwargs):
return self.create(request, *args, **kwargs)With ModelViewSet (in this case), you don't have to create separate views for getting list of objects and detail of one object. ViewSet will handle it for you in a consistent way for both methods.
from rest_framework import viewsets
from posts.models import Post
from posts.serializers import PostSerializer
class PostViewSet(viewsets.ModelViewSet):
"""
A viewset for viewing and editing post instances.
"""
queryset = Post.objects.all()
serializer_class = PostSerializerSo basically, this would not only generate the list view, but also the detail view for every Post instance.
Routers in ViewSets allow the URL configuration for your API to be automatically generated using naming standards.
from rest_framework.routers import DefaultRouter
from posts.views import PostViewSet
router = DefaultRouter()
router.register(r'users', UserViewSet)
urlpatterns = router.urlsDRF provides helpers to add custom actions for ad-hoc behaviours with the @action decorator. The router will configure its url accordingly.
For example, we can add a comments action in the our PostViewSet to retrieve all the comments of a specific post as follows:
from rest_framework import viewsets
from rest_framework.decorators import action
from posts.models import Post
from posts.serializers import PostSerializer, CommentSerializer
class PostViewSet(viewsets.ModelViewSet):
...
@action(methods=['get'], detail=True)
def comments(self, request, pk=None):
try:
post = Post.objects.get(id=pk)
except Post.DoesNotExist:
return Response({"error": "Post not found."},
status=status.HTTP_400_BAD_REQUEST)
comments = post.comments.all()
return Response(CommentSerializer(comments, many=True))Upon registering the view as router.register(r'posts', PostViewSet), this action will then be available at the url posts/{pk}/comments/.
This can usually be performed automatically with generic class-based views and viewsets. If you're using APIView, it must be explicitly applied.
For these views to use pagination, all we have to do is override DEFAULT_PAGINATION_CLASS and PAGE_SIZE in our DRF settings in settings.py.
REST_FRAMEWORK = {
'DEFAULT_PAGINATION_CLASS': 'rest_framework.pagination.LimitOffsetPagination',
'PAGE_SIZE': 20
}First we must define the default behaviour for our pagination, which is done in settings.py.
REST_FRAMEWORK = {
'DEFAULT_PAGINATION_CLASS': 'rest_framework.pagination.LimitOffsetPagination',
'PAGE_SIZE': 20
}Now we need to handle our pagination through our view (APIView in this case).
from rest_framework.settings import api_settings
from rest_framework.views import APIView
class PostView(APIView):
# ... queryset, serializer_class, etc
pagination_class = api_settings.DEFAULT_PAGINATION_CLASS
# We need to override the get method to achieve pagination
def get(self, request):
# ...
page = self.paginate_queryset(self.queryset)
if page is not None:
serializer = self.serializer_class(page, many=True)
return self.get_paginated_response(serializer.data)
# ...
# Now add the pagination handlers taken from
# django-rest-framework/rest_framework/generics.py
@property
def paginator(self):
"""
The paginator instance associated with the view, or `None`.
"""
if not hasattr(self, '_paginator'):
if self.pagination_class is None:
self._paginator = None
else:
self._paginator = self.pagination_class()
return self._paginator
def paginate_queryset(self, queryset):
"""
Return a single page of results, or `None` if pagination is disabled.
"""
if self.paginator is None:
return None
return self.paginator.paginate_queryset(queryset, self.request, view=self)
def get_paginated_response(self, data):
"""
Return a paginated style `Response` object for the given output data.
"""
return self.paginator.get_paginated_response(data)We can define our own pagination class and override default attributes. Using this new class, you can make one, multiple or all views use this instead of LimitOffsetPagination or other ones provided by DRF.
class CustomPagination(PageNumberPagination):
"""
Client controls the response page size (with query param), limited to a maximum of `max_page_size` and default of `page_size`.
"""
page_size = 100
page_size_query_param = 'page_size'
max_page_size = 10000DRF has ready-to-use and integrated authentication schemes, but if you need something more specific, you can customize your own scheme.
SessionAuthentication uses the default session backend for authentication provided by Django, which is more practical for us devs. Once a user has been successfully authenticated, a User instance is stored in request.user.
The recommended use for theTokenAuthentication class is when client-server setups like native apps.
First, add 'rest_framework.authtoken' to your INSTALLED_APPS:
INSTALLED_APPS = [
# Rest of your installed apps ...
'rest_framework',
'rest_framework.authtoken'
]Using signals (in models.py).
from django.conf import settings
from django.db.models.signals import post_save
from django.dispatch import receiver
from rest_framework.authtoken.models import Token
# For existing users
for user in User.objects.all():
Token.objects.get_or_create(user=user)
# For newly created users
@receiver(post_save, sender=settings.AUTH_USER_MODEL)
def create_auth_token(sender, instance=None, created=False, **kwargs):
if created:
Token.objects.create(user=instance)DRF provides a built in view to obtain tokens given username and password.
from rest_framework.authtoken import views
urlpatterns += [
url(r'^api-token-auth/', views.obtain_auth_token)
]OAuth and OAuth2 were previously integrated in DRF, but the corresponding modules were moved and is now supported as a third-party package. There are also other very cool and handy packages that can be easily implemented.
- Django Rest Framework OAuth
- Django OAuth Toolkit (recommended for OAuth2)
If that isn't enough, there's a few more here.
The default look DRF gives you for the browsable API is pretty cool on its own, but in case you don't like it, there are provided ways of customization.
First thing you must do is create a template in templates/rest_framework/api.html that extends rest_framework/base.html.
{% extends "rest_framework/base.html" %}
Now you can modify the many block components that are included in the base.html to your styling. Just as you would do on normal Django templates when you have a base.html.
body- Whole HTML body.bootstrap_theme- CSS for the Bootstrap theme.bootstrap_navbar_variant- CSS for only the Navbar.branding- Brand component in Navbar (top left).script- Custom Javascript.style- Custom CSS.
Those are the common ones, here's all of 'em.
{% block branding %}
<a class="navbar-brand" rel="nofollow" href="erickdelfin.me">
Nifled's Blog
</a>
{% endblock %}This is how you would modify any other block.
If you don't dig the Bootstrap look, you can just drop the whole default look and fully customize it on your own. There is a context provided that you could work with.
api_settings- API settingscontent- The content of the API responserequest- The request objectresponse- The response objectview- The view handling the request
Full list of context variables here.
Take a look at the actual base HTML source code for the API in DRF to get an idea of how it's actually made.
It's important to test your API to make sure it works, of course.
It has a similar name to Django's RequestFactory class, because it extends it. You can use the different class methods to send test requests to your API.
Lets say we want to write test requests for our example Post model.
from django.test import TestCase
from rest_framework.test import APIRequestFactory
from posts.models import Post
from posts.views import PostList
class PostTest(TestCase): # Post object, not HTTP method POST.
"""We'll be testing with the PostList view (class-based view)."""
def setUp(self):
self.factory = APIRequestFactory()
self.post = Post.objects.create(title='Post example', text='Lorem Ipsum')
# For HTTP method GET
def get(self):
view = PostList.as_view()
request = self.factory.get('/posts/')
response = view(request)
self.assertEqual(response.status_code, 200) # 200 = OK
# For HTTP method POST
def post(self):
view = PostList.as_view()
# Generating the request
request = self.factory.post('/posts/', {'title': 'Post example', 'text': 'Lorem Ipsum'})
response = view(request)
expected = {'title': self.post.title, 'text': self.post.text}
self.assertEqual(response.status_code, 201) # 201 = created
self.assertEqual(response.data, expected)