A small scala project with the goal of transforming Mapnik map definitions to something GeoTools and GeoServer can deal with.

1. Install SBT. See the instructions at http://www.scala-sbt.org/release/tutorial/Setup.html
2. sbt update to fetch dependencies
3. sbt compile to build everything
4. sbt run to run the gui, or sbt run-main me.winslow.d.mn2gt.Driver mapnikmap.xml for batch mode conversion (see options below)

sbt "run-main me.winslow.d.mn2gt.Driver [options] {mapnik xml file}"

• -- rest= (url to geoserver rest api, like http://localhost:8080/geoserver/rest). This tells mapnik2geotools to make REST calls to GeoServer to duplicate a given mapnik.xml file. Currently this means:

  • Upload one style for every <Style> element referenced in any <Layer>
  • Configure raster/vector layers for the following DataSource types:
    • geotiff
    • shapefile
    • postgis
  • Configure a GeoServer LayerGroup so that you can conveniently reference them all in a single WMS request

  Additional options which affect the configuration set via REST include:

  • --user=(username) [defaults to admin]. This is the username that the tool will use to authenticate itself with GeoServer.
  • --password=(password) [defaults to geoserver]. This is the password that the tool will use to authenticate itself with GeoServer.
  • --prefix=(some xml prefix) [defaults to mn2gt]. This is the prefix of a workspace that will be associated with all layers uploaded (except the containing layergroup since those do not have workspaces). You must create the workspace yourself before running the tool.
  • --namespace=(some xml namespace URI) [defaults to http://mn2gt.com/]. This is the URI for the workspace mentioned above.

If you don't provide a REST URL, the tool instead dumps some output to the filesystem:

• One SLD file per <Style> element in the Mapnik XML
• One (database).sql file for each database referenced in the Mapnik XML. These contain CREATE VIEW statements used to workaround the fact that mapnik allows prefiltering data by using a SQL subquery in place of a table name. mapnik2geotools assumes you've already run these sql scripts before you run the REST configuration.

There's currently no standalone launcher script, so the recommended usage is to start up the SBT prompt by running sbt with no arguments. You can then use the run task to execute the converter. Currently, you have to run it twice to get GeoServer fully setup:

1. First, generate files on disk by running with the name of a mapnik map definition:

   run /path/to/mapnik/map.xml
   

   This creates one SLD file in /path/to/mapnik/output/ for each Style element in the mapfile. You can inspect these manually (when uploading to GeoServer through REST the styles will be rewritten.) There is also a {dbname}.sql file ({dbname} will be substituted for the name of the database) which creates views for each datastore that uses a subquery. If you use Postgis, you should execute this SQL against your database to set up those aliases for GeoServer (as GeoServer does not currently support subqueries in datastore definitions). This looks like:

   psql -d mydb -f /path/to/mapnik/output/mydb.sql
   

2. With the tables and views in place, run the converter with a GeoServer REST URL and datadir:

   run /path/to/mapnik/map.xml --rest http://localhost:8080/geoserver/rest --datadir /path/to/geoserver_data
   

   Optionally, you can also provide --user <geoserver user name> and --password <geoserver user password>. If you omit these, the default admin:geoserver credentials will be assumed.

many. let me know what you find :)

There is an issue tracker for this project on Github: https://github.com/dwins/mapnik2geotools/issues

If you don't want to use Github feel free to ask on the mailing list (noted below) as well. Or just ask in the #geoserver channel on Freenode IRC: http://webchat.freenode.net/?channels=geoserver

There is a public mailing list at https://groups.google.com/a/opengeo.org/group/mapnik2geotools/topics where you can ask for help and discuss new feature ideas.