Mapping capabilities

PCL will match the features of MapServer by using that software as one of its possible mapping engines. It is the only engine available now, but could be joined by any number of others.

Explicit programming

Usage of MapServer's mapscript is very much like map file scripting: set the values of many interdependent parameters and hit the "go" button. Now, I'm not against run-time app configuration in the style of MapServer's map file, but I'd like a more explicit library for programming with an API that is fully separated from the run-time configuration.

Well known objects

The PCL classes will be modeled after the Open Geospatial Consortium's SLD specification and the GeoAPI 1.0

http://geoapi.sourceforge.net/

These have an advantage over the MapServer classes in that they have been through a fair amount of refactoring, whereas the MapServer classes tend to be intertwined and poorly normalized. The disadvantage of SLD and the GeoAPI is that they are very verbose. The Python Cartographic Library will simplify wherever it is possible for maximum usability. We dispense with all the getters and setters of the GeoAPI LineSymbolizer, for example, and use a more natural Python idiom.

The major advantage of using these domain objects is that service applications developed using PCL will find it very easy to describe their capabilities to consumers.

Cartographic Objects for Zope

ZCO is the first application using the cartographic library.

Usage

Presently, ZCO or the PCL unit tests, particularly testrenderer.py, are the best examples of PCL usage.

Here is an example from testrenderer, rendering from a raster layer with no specified style:

from cartography import data, mapping, spatial

# path to MapServer 4.4 testing GeoTIFF file

path = os.path.join(TESTPATH, 'raster.tif')

# create a featurestore, pop into a layer, and make a renderer
# to render the layer into a map image

store = data.DiskFeatureStore(path)
srs = spatial.SpatialReference(epsg=4326)
layer = mapping.Layer(store, 'raster', srs)
mapper = mapping.MapRenderer('MAPSERVER')

# Define the spatial view and dimensions for output image
# The view is a box with upper left corner at -2.0 E 54.0 N
# and lower right corner at 2.0 E 50.0 N, image dimesion
# will be 200 pixels wide by 200 pixels high

view = mapping.View(srs, ul=spatial.Point(-2.0, 54.0),
                         lr=spatial.Point(2.0, 50.0))
size = [200, 200]

# Render the layer to a JPEG map image

image = mapper.render([(layer, None)], view, 'image/jpeg', size)

# Save to a file on disk

f = open('testRasterLayerMap.jpg', 'wb')
f.write(image.getdata())
f.close()

The output: