Sideloading Resources

[zacid]

Hi there,

I am using rest_framework_ember and am at the point where I'd like to start side loading resources.

I am using viewsets to setup my resource endpoints:

class DocumentViewSet(viewsets.ModelViewSet):
    """
    API endpoint that allows users to be viewed or edited.
    """
    queryset = Document.objects.all()
    serializer_class = DocumentSerializer

class ContactViewSet(viewsets.ModelViewSet):
    """
    API endpoint that allows users to be viewed or edited.
    """
    queryset = Contact.objects.all()
    serializer_class = ContactSerializer

A document model has a relationship with a contact model so in order to sideload this data I'd have to somehow get something like this:

{
  "document": {
    "id": 1,
    "title": "This is a document",
    "contacts": [1, 2]
  },

  "contacts": [{
    "id": 1,
    "first_name": "my name"
  },
  {
    "id": 2,
    "first_name": "your name"
  }]
}

Can this be accomplished using rest_framework_ember? I saw that there is a section in the readme about side loading relationships but it didn't exactly makes sense when it comes to my problem.

Any help in this regard would be greatly appreciated.

Thanks!

[jerel]

@Zac156 you can accomplish this fairly easily. What you need to do is create a FooViewSet endpoint with resource_name = False to turn off the automatic root key generation. Then output the combined data something like this https://github.com/ngenworks/rest_framework_ember/blob/master/example/api/api.py#L82-L101

Re-open if you still have troubles or questions.

[zacid]

Thank you, but what about sideloading related associates on objects? I.E. in your example if cars was a related field to the user table? How would one get that to work then?

cars = models.ManyToManyField('User')

And when I set resource_name = False :

class DocumentViewSet(viewsets.ModelViewSet):
    """
    API endpoint that allows users to be viewed or edited.
    """
    resource_name = False

    queryset = Document.objects.all()
    serializer_class = DocumentSerializer

    def list(self, request, *args, **kwargs):
            data = {
                'document': self.get_queryset(),
                'contacts': Contact.objects.filter(id=1)
            }
            serializer = DocumentContactSerializer(data)
            return Response(serializer.data)

These are the results, is this correct? Or a bug?:

{
    "false": {
        "document": [
            {
                "contacts": [
                    12
                ], 
                "files": [], 
                "last_update_date": 1414130476, 
                "id": 3, 
                "name": "Z Doc", 
                "title": null, 
                "type": "Will", 
                "notes": "Note", 
                "vault": null
            }
        ], 
        "contacts": [
            {
                "emails": [], 
                "addresses": [], 
                "telephones": [], 
                "id": 1, 
                "title": "Mr", 
                "role": "R", 
                "first_name": "Zac", 
                "last_name": "Last Name", 
                "notes": "N", 
                "company": "C", 
                "organisation": null, 
                "beneficiary": null, 
                "financials": [], 
                "messages": [], 
                "documents": [], 
                "policies": []
            }
        ]
    }
}

[jerel]

If I'm understanding you correctly you are asking about what Ember Data calls embedded records: returning nested data inside the parent object. Here's the mixin required on the Ember Data side: http://emberjs.com/api/data/classes/DS.EmbeddedRecordsMixin.html I've not had a need to use embedded records yet but it should work fine if you make a SerializerMethodField on the parent that serializes the child.

The JSON you pasted with false as the root key is incorrect. documents and contacts should be top level. It may be a bug; we do have a test that covers that scenario but uses generics.GenericAPIView rather than viewsets.ModelViewSet. Here is the line that should check for that if you want to drop a debug statement in there: https://github.com/ngenworks/rest_framework_ember/blob/master/rest_framework_ember/renderers.py#L24

What version are you running? The resource_name = False option wasn't added until one of the later releases.

[zacid]

Hi Jerel,

I am using rest-framework-ember==1.0.3 and have since gotten sideloading to work correctly with the resource_name = False flag. Here is my code below:

class DocumentViewSet(viewsets.ModelViewSet):
    """
    API endpoint that allows documents to be viewed or edited.
    """

    queryset = Document.objects.all()
    serializer_class = DocumentSerializer

    # Side loading code for documents
    resource_name = False
    # renderer_classes = (JSONRenderer, BrowsableAPIRenderer)

    def list(self, request, *args, **kwargs):
        # import IPython
        # IPython.embed()
        data = {'document': []}
        for doc in self.get_queryset():
            data['document'].append(doc)
            data['contacts'] = doc.contacts.all()

        serializer = DocumentContactSerializer(data)

        return Response(serializer.data)

This works as I'd like it to work. For now I'd really like to get sideloading to work rather than embedded records but eventually I'd do either or, depending on what actually works.

The problem now is that since I've implemented this and overwritten the list() method on the ModelViewSet whenever a new object is created on a POST I am getting this error:

'NoneType' object has no attribute '__getitem__'

If I comment out the resource_name = False then POST works as expected again.

Would you know what could be causing this? Thanks for all your help so far.

[zacid]

Hi,

Just wondering if you possibly have any feedback yet regarding this issue as I'm really struggling to get post/create to work once I've enabled sideloading with 'resource_name=False'..

Is this something I am misunderstanding or maybe I have implemented this wrong?

Thanks again for all your help

[jerel]

@Zac156 I'm guessing that is is a bug. Probably from this line here: https://github.com/ngenworks/rest_framework_ember/blob/master/rest_framework_ember/parsers.py#L32 You could try changing https://github.com/ngenworks/rest_framework_ember/blob/master/rest_framework_ember/parsers.py#L25 to parser_context={} as a None type will cause line 32 to fail.

[zacid]

@jerel,

I've tried as you've suggested but am still getting the same error. Anything else that you could think of? Are you making use of POST/create when it comes to sideloading? I don't know what I am missing.

[zacid]

@jerel maybe this will help? http://dpaste.com/0AK4GF1

[zacid]

Looks like this is causing the error:
https://github.com/ngenworks/rest_framework_ember/blob/master/rest_framework_ember/parsers.py#L33

Line 33 shows that it's trying to get the response based on the key resource_name, the problem is that now we have assigned the resource_name to be False which is not a valid key for the result dictionary.

How could we work around this?

[jerel]

@Zac156 after reading this issue again I'm not able to replicate the issue. A couple thoughts:

• you could use self.resource_name inside your function rather than setting it on the class if you only need it on POST and not GET
• are you trying to POST sideloaded data or only GETing it and POSTing one item thing back?

Would it be possible for you to write a test illustrating the fault? I'd be happy to take a look at it if I had something more to work with.

[jerel]

Closing due to inactivity/incomplete data