Remove not always supported auto select_related (#651)

Polymorphic models do not always support `select_related` as of
https://django-polymorphic.readthedocs.io/en/stable/advanced.html?highlight=select_related#about-queryset-methods
and might actually be slower in some cases as explained in https://django-rest-framework-json-api.readthedocs.io/en/stable/usage.html?highlight=usage#performance-improvements

Hence reverting change of AutoPreloadMixin as it has not been released yet.

Author: sliverc

CHANGELOG.md

@@ -17,7 +17,6 @@ any parts of the framework not mentioned in the documentation should generally b
 ### Changed
 
 * Allow to define `select_related` per include using [select_for_includes](https://django-rest-framework-json-api.readthedocs.io/en/stable/usage.html#performance-improvements)
-* Reduce number of queries to calculate includes by using `select_related` when possible
 * Use REST framework serializer functionality to extract includes. This adds support like using
   dotted notations in source attribute in `ResourceRelatedField`.
 
@@ -31,7 +30,6 @@ any parts of the framework not mentioned in the documentation should generally b
 ### Deprecated
 
 * Deprecate `PrefetchForIncludesHelperMixin` use `PreloadIncludesMixin` instead
-* Deprecate `AutoPrefetchMixin` use `AutoPreloadMixin` instead
 
 ## [2.7.0] - 2019-01-14
 

example/tests/test_performance.py

@@ -53,7 +53,7 @@ def test_query_count_include_author(self):
         4. Author types prefetched
         5. Entries prefetched
         """
-        with self.assertNumQueries(4):
+        with self.assertNumQueries(5):
             response = self.client.get('/comments?include=author&page[size]=25')
             self.assertEqual(len(response.data['results']), 25)
 

example/views.py

@@ -12,7 +12,7 @@
 from rest_framework_json_api.filters import OrderingFilter, QueryParameterValidationFilter
 from rest_framework_json_api.pagination import JsonApiPageNumberPagination
 from rest_framework_json_api.utils import format_drf_errors
-from rest_framework_json_api.views import ModelViewSet, PreloadIncludesMixin, RelationshipView
+from rest_framework_json_api.views import ModelViewSet, RelationshipView
 
 from example.models import Author, Blog, Comment, Company, Entry, Project, ProjectType
 from example.serializers import (
@@ -200,12 +200,9 @@ def get_queryset(self, *args, **kwargs):
         return super(CommentViewSet, self).get_queryset()
 
 
-class CompanyViewset(PreloadIncludesMixin, viewsets.ModelViewSet):
+class CompanyViewset(ModelViewSet):
     queryset = Company.objects.all()
     serializer_class = CompanySerializer
-    prefetch_for_includes = {
-        'current_project': ['current_project'],
-    }
 
 
 class ProjectViewset(ModelViewSet):

rest_framework_json_api/views.py

@@ -115,20 +115,18 @@ def get_queryset(self, *args, **kwargs):
         return qs
 
 
-class AutoPreloadMixin(object):
-
+class AutoPrefetchMixin(object):
     def get_queryset(self, *args, **kwargs):
         """ This mixin adds automatic prefetching for OneToOne and ManyToMany fields. """
-        qs = super(AutoPreloadMixin, self).get_queryset(*args, **kwargs)
+        qs = super(AutoPrefetchMixin, self).get_queryset(*args, **kwargs)
+
         included_resources = get_included_resources(self.request)
 
         for included in included_resources + ['__all__']:
             # If include was not defined, trying to resolve it automatically
             included_model = None
             levels = included.split('.')
             level_model = qs.model
-            # Suppose we can do select_related by default
-            can_select_related = True
             for level in levels:
                 if not hasattr(level_model, level):
                     break
@@ -144,12 +142,6 @@ def get_queryset(self, *args, **kwargs):
                 if not (is_forward_relation or is_reverse_relation):
                     break
 
-                # Figuring out if relation should be select related rather than prefetch_related
-                # If at least one relation in the chain is not "selectable" then use "prefetch"
-                can_select_related &= (
-                    issubclass(field_class, (ForwardManyToOneDescriptor, ReverseOneToOneDescriptor))
-                )
-
                 if level == levels[-1]:
                     included_model = field
                 else:
@@ -165,23 +157,11 @@ def get_queryset(self, *args, **kwargs):
                         level_model = model_field.model
 
             if included_model is not None:
-                if can_select_related:
-                    qs = qs.select_related(included.replace('.', '__'))
-                else:
-                    qs = qs.prefetch_related(included.replace('.', '__'))
+                qs = qs.prefetch_related(included.replace('.', '__'))
 
         return qs
 
 
-class AutoPrefetchMixin(AutoPreloadMixin):
-
-    def __init__(self, *args, **kwargs):
-        warnings.warn("AutoPrefetchMixin is deprecated. "
-                      "Use AutoPreloadMixin instead",
-                      DeprecationWarning)
-        super(AutoPrefetchMixin, self).__init__(*args, **kwargs)
-
-
 class RelatedMixin(object):
     """
     This mixin handles all related entities, whose Serializers are declared in "related_serializers"
@@ -259,14 +239,14 @@ def get_related_instance(self):
                 raise NotFound
 
 
-class ModelViewSet(AutoPreloadMixin,
+class ModelViewSet(AutoPrefetchMixin,
                    PreloadIncludesMixin,
                    RelatedMixin,
                    viewsets.ModelViewSet):
     pass
 
 
-class ReadOnlyModelViewSet(AutoPreloadMixin,
+class ReadOnlyModelViewSet(AutoPrefetchMixin,
                            RelatedMixin,
                            viewsets.ReadOnlyModelViewSet):
     pass