Skip to content


Subversion checkout URL

You can clone with
Download ZIP
True infinite scroll pagination using the seek method (keyset paging)
Branch: master
Failed to load latest commit information.
infinite_scroll_pagination bump 0.1.3
tests unicode_literals
.gitattributes Initial commit
.gitignore Initial commit
.travis.yml bump 0.1.3
LICENSE Initial commit Initial commit bump 0.1.3 unicode_literals bump 0.1.3

infinite-scroll-pagination Build Status Coverage Status

infinite-scroll-pagination is a Django app that implements the seek method (AKA Keyset Paging) for scalable pagination.

How it works

Keyset driven paging relies on remembering the top and bottom keys of the last displayed page, and requesting the next or previous set of rows, based on the top/last keyset

This approach has two main advantages over the OFFSET/LIMIT approach:

  • is correct: unlike the offset/limit based approach it correctly handles new entries and deleted entries. Last row of Page 4 does not show up as first row of Page 5 just because row 23 on Page 2 was deleted in the meantime. Nor do rows mysteriously vanish between pages. These anomalies are common with the offset/limit based approach, but the keyset based solution does a much better job at avoiding them.
  • is fast: all operations can be solved with a fast row positioning followed by a range scan in the desired direction

For a full explanation go to the seek method


infinite-scroll-pagination requires the following software to be installed:

  • Python 2.7, 3.3 or 3.4
  • Django 1.5, 1.6 or 1.7


  1. None :smile:


Paging by date or any other field:

Note: has_next() is not reliable, records may have been deleted in between requests. page() will raise an EmptyPage in this case, you should catch it and return a proper response.


from infinite_scroll_pagination.paginator import SeekPaginator, EmptyPage

def pagination_ajax(request, pk=None):
    if not request.is_ajax():
        return Http404()

    if pk is not None:
        # I'm doing an extra query because datetime serialization/deserialization is hard
        date = get_object_or_404(Article, pk=pk).date
        # is requesting the first page
        date = None

    articles = Article.objects.all()
    paginator = SeekPaginator(articles, per_page=20, lookup_field="date")

        page =, pk=pk)
    except EmptyPage:
        data = {'error': "this page is empty", }
        articles_list = [{"title": a.title, } for a in page]
        data = {'articles': articles_list,
                'has_next': page.has_next(),
                'pk': page.next_page_pk()}

    return HttpResponse(json.dumps(data), content_type="application/json")

Paging by pk or id (special case):


def pagination_ajax(request, pk=None):

    page =, pk=pk)


Showing how many objects (or pages) are left:

Note: For true infinite scroll, this is not recommended. Since it does a count() query.

It would be better if you increase an IntegerField every time a record is saved and do some javascript magic to know how many objects are left.


page_first =

data = {'objects_left_count': page_first.objects_left,
        'pages_left_count': page_first.pages_left,


  • get previous page will get implemented in the future, it's not there at this time.
  • Order is DESC (from newest to oldest). You may submit a pull request for ASC order support.
  • Lazy pagination is not supported, yet.


Feel free to check out the source code and submit pull requests.

You may also report any bug or propose new features in the issues tracker

Copyright / License

Copyright 2014 Esteban Castro Borsani.

Licensed under the MIT License.

Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License.

Something went wrong with that request. Please try again.