Django Remote Queryset

Description

Django Remote Queryset (a.k.a. DRQ) is a library to execute rich ORM queries from client-side using the same interface of Django ORM.

How it works

DRQ defines a JSON syntax to call all Django's Queryset methods.

Consider this example:

    
    from myapp.models import MyModel
    
    not_null_title = MyModel.objects.filter(title__isnull=False)
   

DRQ query syntax for the previous filter is the following

   {
       "_query_class" : "filter",
       "_condition" : "title__isnull",
       "_value" : false
   }

DRQ supports Django Rest Framework and automatically decodes and applies this filter to any GET request.

Installation

DRQ is distrubuted through pypi

   
   $ pip install django-remote-queryset
   

Usage

1. Add django_remote_queryset to your INSTALLED_APPS:

    INSTALLED_APPS = [
        
        ...
        
        'rest_framework',
        'django_remote_queryset',
        
        ...
        
    ]

Django rest_framework is strongly recommended but if you prefer to use DRQ without it you can see Advanced Usage section.

2. Extend DRQModelViewSet for your model:

    from django_remote_queryset.viewset import DRQModelViewSet

    class MyModelModelViewSet(DRQModelViewSet):
        queryset = MyModel.objects.all()
        serializer_class = MyModelModelSerializer

And use it like any other Django REST Framework ViewSet!

3. Enjoy filtering from your browser!

   GET: http://yourdomain/my_model_rest_path/?query=
   
   query = {
       "_query_class" : "filter",
       "_condition" : "title__isnull",
       "_value" : false
   }

For further informations on DRF ViewSets see Django Rest Framework documentation.

Example project

This repository contains a example_django_project folder. Try it doing the following:

    $ cd example_django_app
    # Install site dependencies
    $ pip install -r requirements.txt
    # Create sqlite database
    $ python manage.py migrate
    # Create demo data
    $ python manage.py create_demo_data
    # Run the server
    $ python manage.py runserver

Then:

1. Go to http://localhost:8000 and you should see the usual Django Rest Framework page.
2. Go to http://localhost:8000/polls and you should see the list of all polls available
3. Go to http://localhost:8000/polls/?query={"_query_class":"filter","_condition":"title__icontains","_value":"DRQ"}

Advanced Usage

It's possible to extend/customize DRQ using it in your extisting ViewSets or integrating it in your own custom Views/Django code.

Use the DRQFilterBackend

DRQ is implemented using DRF's filter_backends, you can add it to your ModelViewSet easily

    from django_remote_queryset.backend import DRQFilterBackend

    class AnotherModelViewSet(ModelViewSet):
        filter_backends = ( ..., DRQFilterBackend, ...)

Use the QueryDecoder

You can use directly the QueryDecoder to decode and apply the json_query to any queryset

    from django_remote_queryset.queries import QueryDecoder
    
    ...

    original_queryset = MyModel.objects.all()
    
    
    json_query = {
       "_query_class" : "filter",
       "_condition" : "title__isnull",
       "_value" : false
   }
    
    query = QueryDecoder \
            .decodeJSONQuery(json_query)
    
    if query is not None:
        filtered_queryset = query.applyOnQuerySet(original_queryset)

DRQ JSON Syntax Specification

DRQ Queries are built using the Composite Pattern. Several queries are already available and the Framework could be extended with custom ones. All the queries are contained in django_remote_queryset.queries module.

Query

This is the abstract one, defines the interface for every child.

    class Query():
        def __init__(self, json_query):
            """
            Initializes the sub_tree of this query starting from the json_query parameter
            :param json_query: dict
            """
            pass
        
        def applyOnQuerySet(self, queryset):
            """
            Applies the sub_tree of this query to the queryset passed as parameter
            :param queryset: Queryset
            :return: Queryset
            """
            return queryset 

It's JSON syntax is the following

    {
      "_query_class":"query"
    }

Filter

This query applies the .filter( ... ) method on the queryset

It's JSON syntax is the following

    {
      "_query_class": "filter",
      "_condition": [string] | string,
      "_value": [*] | * 
    }

Examples:

1. Single lookup

    
    Person.filter(age__gte=4)

    {
      "_query_class": "filter",
      "_condition": "age__gte",
      "_value": 4 
    }

2. Multiple lookups

    
    Person.filter(age__gte=4, name__icontains='Nic')

    {
      "_query_class": "filter",
      "_condition": ["age__gte","name__icontains"],
      "_value": [4,"Nic"] 
    }

Exclude

This query applies the .exclude( ... ) method on the queryset

It's JSON syntax is the following

    {
      "_query_class": "exclude",
      "_condition": [string] | string,
      "_value": [*] | * 
    }

Examples:

1. Single lookup

    
    Person.exclude(age__gte=4)

    {
      "_query_class": "exclude",
      "_condition": "age__gte",
      "_value": 4 
    }

2. Multiple lookups

    
    Person.exclude(age__gte=4, name__icontains='Nic')

    {
      "_query_class": "exclude",
      "_condition": ["age__gte","name__icontains"],
      "_value": [4,"Nic"] 
    }

CompositeQuery

This query creates a chain of subqueries applying all it's sub_tree on the queryset.

It's JSON syntax is the following

    {
      "_query_class": "compositequery",
      "_sub_queries":[
        ... other queries ...
      ]
    }

The subqueries are applied as a chain.

    for query in self._sub_queries:
        queryset = query.applyOnQuerySet(queryset)

Examples:

1. Chained filters

    Person.objects.filter(age__gte=4).filter(name__icontains='Nic')

    {
      "_query_class": "compositequery",
      "_sub_queries":[
        {
            "_query_class": "filter",
            "_condition": "age__gte",
            "_value": 4 
        },
        {
            "_query_class": "filter",
            "_condition": "name__icontains",
            "_value": "Nic"
        }
      ]
    }

2. Chained mixed queries

    Person.objects.filter(age__gte=4).exclude(name__icontains='Nic')

    {
      "_query_class": "compositequery",
      "_sub_queries":[
        {
            "_query_class": "filter",
            "_condition": "age__gte",
            "_value": 4 
        },
        {
            "_query_class": "exclude",
            "_condition": "name__icontains",
            "_value": "Nic"
        }
      ]
    }