Django-Filter Quickstart Documentation: Subclassing of DatatablesFilterSet

First, thanks to the folks who work on this package I really needed it today.

To those trying to implement the Django-filter subpackage and finding it missing, it is not in the current release, but it is in master. See #82

Once you have that, and are following the Django-Filter Quickstart, you might run into some confusion on what gets subclassed.

From my experience trying to repeat the docs, this line:

class AlbumGlobalFilter(GlobalFilter):

triggers the following error:

AttributeError: type object 'AlbumGlobalFilter' has no attribute '_meta'

I believe instead AlbumGlobalFilter should subclass DatatablesFilterSet, which is imported in the example but never implemented. i.e.:

class AlbumGlobalFilter(DatatablesFilterSet):

Which makes some sense given the idea of using the DataTables version of various imports otherwise handled by django-filter i.e. filters.FilterSet.

This may be correct for the example in the docs, but I was also stumped up by the Meta class in the current docs. The fields = '__all__' resulted in an error, I believe because I did not declare all of the fields from my model:

TypeError: 'Meta.fields' contains fields that are not defined on this FilterSet: __all__

I was able to get django-rest-framework-datatables and the django-filter to ultimately work for my subset of model fields. In case someone else is might appreciate an example, it is simply something like:

Declare the field-level global filter

class GlobalCharFilter(GlobalFilter, filters.CharFilter):
    pass

Then declare the Filterset with only the necessary fields:

class FacilityFilterCrud(DatatablesFilterSet):
    name = GlobalCharFilter(lookup_expr='icontains')

    class Meta:
        model = Facility
        fields = ['__name__']

A partial example like this in the docs, might be helpful to help demonstrate some of the more fine-grained control hinted at in the introduction.

[izimobil]

Fixed by https://github.com/izimobil/django-rest-framework-datatables/pull/117 , thanks to all.

[banagale]

@izimobil Sure. I typically look for a go-ahead from a maintainer before investing the time, which was why I created the issue.

I have another outstanding docs PR request in scrapy. I’ll try and handle them both at the same time.

Feel free to assign this to me.