A flexible implementation of parts of the wms standard using Flask for the web stuff. The Flask Blueprint system is used to disclose a variety of datasources.

The main purpose of the library is visualization of fast and accurate 3di flooding calculations.

Features:

• Celery worker for long lasting tasks such as building bathymetry pyramids
• File and only file based configuration
• Modular setup employing Flask's Blueprint system

For the build to work a number of system libraries is required:

$ libhdf5-serial-dev # Not sure about this one.
$ libnetcdf-dev
$ libfreetype6-dev

GDAL 1.9.1 or higher. The server will NOT work correctly with a lower version (grid is misplaced).

First the basic steps:

$ python bootstrap.py
$ bin/buildout

Put your datasets in var/data/3di/<myfolder>. The name of the folder will be the name of the layer in the request. In the folder should be a .nc file and a .tif or .asc bathymetry file. .tif performs better than .asc in the preparation step, but after that there is no difference since the bathymetry is cached in a pyramid oject on the filesystem. Create the tif from the bathimetry .asc file with gdal:

$ gdal_translate mybathimetry.asc myresult.tif

Start the server and the task processor using:

$ bin/supervisord

Go to localhost:5000/3di/demo to see the server in action. To see the requests for timeseries and profiles along a line, check the developertools of the browser while single clicking a point on the map (timeseries) or a single click followed by a double click on some other point (profile along a line) on the demo page.

Examples of requests for contours and quanties directly from the netcdf:

http://localhost:5000/3di/data?request=getcontours&layers=purmer2
http://localhost:5000/3di/data?request=getquantity&layers=purmer2&
    quantity=dep&time=0&decimals=2

The rasterinfo server serves a HTTP layer over gislib functonality

Go to localhost:5000/rasterinfo/ to confirm the service is working. You will see a link to the profile tool

Profile tool: Send a linestring and projection to the server which returns a (json) list of (x, value) pairs where X = distance in projection units from start of line and value = value of the raster at that point.

Example: http://localhost:5000/rasterinfo/profile?geom=LINESTRING(570060.51753709%206816367.7101946,568589.10474281%206815374.028827)&srs=EPSG:900913