It should be enough for a subclass to implement {@linkPlain org.vfny.geoserver.responses.wms.GetLegendGraphicProducer#writeTo(OutputStream)}and getContentType() in order to encode the BufferedImage produced by this class to the appropiate output format.
This class takes literally the fact that the arguments WIDTH and HEIGHT are just hints about the desired dimensions of the produced graphic, and the need to produce a legend graphic representative enough of the SLD style for which it is being generated. Thus, if no RULE parameter was passed and the style has more than one applicable Rule for the actual scale factor, there will be generated a legend graphic of the specified width, but with as many stacked graphics as applicable rules were found, providing by this way a representative enough legend.
| |