Tree query improvements (#3443)

* Allow part category table to be ordered by part count

* Add queryset annotation for part-category part-count

- Uses subquery to annotate the part-count for sub-categories
- Huge reduction in number of queries

* Update 'pathstring' property of PartCategory and StockLocation

- No longer a dynamically calculated value
- Constructed when the model is saved, and then written to the database
- Limited to 250 characters

* Data migration to re-construct pathstring for PartCategory objects

* Fix for tree model save() method

* Add unit tests for pathstring construction

* Data migration for StockLocation pathstring values

* Update part API

- Add new annotation to PartLocationDetail view

* Update API version

* Apply similar annotation to StockLocation API endpoints

* Extra tests for PartCategory API

* Unit test fixes

* Allow PartCategory and StockLocation lists to be sorted by 'pathstring'

* Further unit test fixes

InvenTree/part/api.py

@@ -53,6 +53,13 @@ class CategoryList(ListCreateAPI):
     queryset = PartCategory.objects.all()
     serializer_class = part_serializers.CategorySerializer
 
+    def get_queryset(self, *args, **kwargs):
+        """Return an annotated queryset for the CategoryList endpoint"""
+
+        queryset = super().get_queryset(*args, **kwargs)
+        queryset = part_serializers.CategorySerializer.annotate_queryset(queryset)
+        return queryset
+
     def get_serializer_context(self):
         """Add extra context data to the serializer for the PartCategoryList endpoint"""
         ctx = super().get_serializer_context()
@@ -141,9 +148,11 @@ class CategoryList(ListCreateAPI):
 
     ordering_fields = [
         'name',
+        'pathstring',
         'level',
         'tree_id',
         'lft',
+        'part_count',
     ]
 
     # Use hierarchical ordering by default
@@ -165,6 +174,13 @@ class CategoryDetail(RetrieveUpdateDestroyAPI):
     serializer_class = part_serializers.CategorySerializer
     queryset = PartCategory.objects.all()
 
+    def get_queryset(self, *args, **kwargs):
+        """Return an annotated queryset for the CategoryDetail endpoint"""
+
+        queryset = super().get_queryset(*args, **kwargs)
+        queryset = part_serializers.CategorySerializer.annotate_queryset(queryset)
+        return queryset
+
     def get_serializer_context(self):
         """Add extra context to the serializer for the CategoryDetail endpoint"""
         ctx = super().get_serializer_context()