WiscLinc subtitle
Link to FGDC Schema FGDC Schema Link to Summary and ResultsSummary and Results

WFS Service

This project particularly investigates web feature services using a MapServer software base with three types of schema transformation methods: XSLT, STX, and Mapserver's built in transformation capabilities.

The MapServer WFS server "map file" (MapServer's configuration file) for our PostGIS database can be downloaded here. This geodetic control WFS service can be accessed and queried using OGC-Compliant WFS viewers such as Gaia (http://www.thecarbonproject.com/products/gaia.html ). This following figure show the NGS WFS Server viewed by Gaia:




Another WFS viewer is the interoperability extension for ESRI’s ArcGIS, CarbonArc Lite. Below is a figure showing our service viewed using this extension:



Other Possibilities

Another potential wfs server is Geoserver. GeoServer requires Java 1.4 using Java socket technique. GeoServer can store server shape data in data/featureTypes folder, under which a file info.xml maintains metadata (source data, bbox, etc.) and sources data names and locations. The featureTypes is the data source to be provided as a feature. GeoServer maintain catalog.xml and services.xml to track featureTypes and services. GeoServer is also able to access remote shape file but not recommended, can access PostGIS tables. GeoServer is developed based on GeoTools (http://www.geotools.org/), which is an open source Java GIS toolkit for developing standards compliant solutions.

GeoServer support simple manipulation of schema, including changing column data type and omitting some columns. See here for a screenshot of GeoServer schema when defining feature type.

An interesting proposal of schema manipulation by GeoServer is based on virtual data store. A virtual data store can be dynamically obtained from database query, i.e., selectDerivedFeatureType (http://docs.codehaus.org/display/GEOS/selectDerivedFeatureType ). A virtual data store wraps another datastore and produces a new FeatureType computed from the FeatureType produced by the wrapped datastore. Another implementation is using aggregation fuction of database query, i.e., aggregatederivedfeaturetype (http://docs.codehaus.org/display/GEOS/aggregatederivedfeaturetype).