Generate MapBox VectorTiles from GeoDjango models

Directly with PostgreSQL/PostGIS 2.4+ or python native mapbox_vector_tile

Read full documentation

Installation

Basic

pip install django-vectortiles

• Without any other option, use only vectortiles.postgis
• Ensure you have psycopg2 set and installed

If you don't want to use Postgis

pip install django-vectortiles[mapbox]

• This will incude mapbox_vector_tiles package and its dependencies
• Use only vectortiles.mapbox

Examples

• assuming you have django.contrib.gis in your INSTALLED_APPS and a gis compatible database backend

# in your app models.py

from django.contrib.gis.db import models


class Layer(models.Model):
    name = models.CharField(max_length=250)


class Feature(models.Model):
    geom = models.GeometryField(srid=4326)
    name = models.CharField(max_length=250)
    layer = models.ForeignKey(Layer, on_delete=models.CASCADE, related_name='features')

Simple model:

# in your view file

from django.views.generic import ListView
from vectortiles.postgis.views import MVTView
from yourapp.models import Feature


class FeatureTileView(MVTView, ListView):
    model = Feature
    vector_tile_layer_name = "features"
    vector_tile_fields = ('other_field_to_include', )


# in your urls file
from django.urls import path
from yourapp import views


urlpatterns = [
    ...
    path('tiles/<int:z>/<int:x>/<int:y>', views.FeatureTileView.as_view(), name="feature-tile"),
    ...
]

Related model:

# in your view file

from django.views.generic import DetailView
from vectortiles.mixins import BaseVectorTileView
from vectortiles.postgis.views import MVTView
from yourapp.models import Layer


class LayerTileView(MVTView, DetailView):
    model = Layer
    vector_tile_fields = ('other_field_to_include', )

    def get_vector_tile_layer_name(self):
        return self.get_object().name

    def get_vector_tile_queryset(self):
        return self.get_object().features.all()

    def get(self, request, *args, **kwargs):
        self.object = self.get_object()
        return BaseVectorTileView.get(self,request=request, z=kwargs.get('z'), x=kwargs.get('x'), y=kwargs.get('y'))


# in your urls file
from django.urls import path
from yourapp import views


urlpatterns = [
    ...
    path('layer/<int:pk>/tile/<int:z>/<int:x>/<int:y>', views.LayerTileView.as_view(), name="layer-tile"),
    ...
]

Usage without PostgreSQL / PostGIS

Just import and use vectortiles.mapbox.view.MVTView instead of vectortiles.postgis.view.MVTView

Usage with DRF

django-vectortiles can be used with DRF if renderer_classes of the view is overridden (see DRF docs). Simply use the right BaseMixin and action on viewsets, or directly a GET method in an APIView, i.e.:

from rest_framework import renderers, views
from vectortiles.postgis.views import MVTView


class MVTRenderer(renderers.BaseRenderer):
    media_type = "application/vnd.mapbox-vector-tile"
    format = "pbf"

    def render(self, data, accepted_media_type=None, renderer_context=None):
        return data


class TileServerView(MVTView, views.APIView):
    renderer_classes = [MVTRenderer]

    def get(...): ...

Development

With docker and docker-compose

docker pull makinacorpus/geodjango:bionic-3.6
docker-compose build
# docker-compose up
docker-compose run /code/venv/bin/python ./manage.py test

Local

• Install python and django requirements (python 3.6+, django 2.2+)
• Install geodjango requirements
• Have a postgresql / postgis 2.4+ enabled database
• Use a virtualenv

pip install .[dev] -U