tmstiler

Map tile utilities supporting python3.

This project supports map tile creation using python3. (and the project serves as a personal study of tile maps)

It mimics TileStache in the interface to some degree, but leaves caching to higher levels.

This project contains two classes intended for use in Map Tile generation:

* RasterTileManager

* DjangoRasterTileLayerManager

'RasterTileManager' provides helper methods for raster space to spherical mercator conversion. For transforming spherical mercator (google maps) projected points to tile pixel locations the '.sphericalmercator_to_pixel(zoom, tilex, tiley, xm, ym)' method is available.

'DjangoRasterTileLayerManager' allows you to use data already binned and placed in a django model containing a PointField, to generate .png tiles.

Sample image created with tiles rendered using the tmstiler's 'DjangoRasterTileLayerManager' in the safecasttiles project:

Below is an excerpt of the safecasttiles project, showing how the 'DjangoRasterTileLayerManager' class can be added in django for tile creation.

NOTE: 'DjangoRasterTileLayerManager' requires pillow and django. ('RasterTileManager' alone has no additional dependencies)

from io import BytesIO

from django.http import HttpResponse
from django.views.generic import View

from tmstiler.django import DjangoRasterTileLayerManager

from .models import Measurement


class SafecastMeasurementsTileView(View):

    def __init__(self, **kwargs):
        super().__init__(**kwargs)
        # create per month layers
        months = Measurement.objects.order_by("date").values_list('date', flat=True).distinct()
        layers = OrderedDict()
        legend = Legend()
        for m in months:
            qs = Measurement.objects.filter(date=m)
            month_layername = m.strftime("%Y%m")
            layers[month_layername] = {
                        "pixel_size": 1500,  # size of bin in meters
                        "point_position": "upperleft",
                        "model_queryset": qs,
                        "model_point_fieldname": "location",
                        "model_value_fieldname": "value",
                        "legend_instance": legend,  # returns a hslcolor_str
                        }
        self.tilemgr = DjangoRasterTileLayerManager(layers)

    def get(self, request):
        layername, zoom, x, y, image_format = self.tilemgr.parse_url(request.path)
        mimetype, tile_pil_img_object = self.tilemgr.get_tile(layername, zoom, x, y)
        image_encoding = image_format.replace(".", "")  # change ".png" to just "png"
        # pillow img.tobytes() doesn't seem to work, workaround to serve raw bytes via BytesIO()
        image_fileio = BytesIO()  
        tile_pil_img_object.save(image_fileio, image_encoding)
        image_fileio.seek(0)
        return HttpResponse(image_fileio, content_type=mimetype)

Dependencies

Optional:

The following libraries are needed to make use of the 'DjangoRasterTileLayerManager()' class, for generating custom geo binned data from django Models containing a PointField(). (Currently only square bins are supported, but it should be pretty trival to support Polygon objects)

• django: [geodjango] https://www.djangoproject.com/download/ (optional)
• pillow: https://github.com/python-pillow/Pillow (optional)