Justin Deoliveira wrote:

It appears that when a feature is selected with the selection tool, a filter gets created which contains functions. I beleive this is happening when the render actually tries to render the selection.

For instance:

<BBOX>
  <PropertyName>the_geom</PropertyName>...
</BBOX>
<Or>
   <PropertyIsEqualTo>
     <Function name="geometryType">
    <PropertyName>the_geom</PropertyName></Function>
          <Literal>Point</Literal>
        </PropertyIsEqualTo>
   <PropertyIsEqualTo>
   <Function name="geometryType">
        <PropertyName>the_geom</PropertyName></Function>
         <Literal>MultiPoint</Literal>
        </PropertyIsEqualTo>
   </Function>
</Or>

I beleive this is killing the underlying datastores as they dont understand functions. WFS is the only one that i know of that can and in this case of the "geometryType" function, it only has a hope of working with GeoServer since they uses the same function implementation.

I've commented out some lines of SelectionStyleContent which were adding these filters. Now selection works nicely for a geoserver wfs, but it remains to be seen why those were there in the first place (PointSymbolizers had a filter restricting their use to points, but I would expect a renderer would be smart to realize that PointSymbolizers are for points)... Unless there are crazy situations where one could have multiple geometry types within the same feature type. Hopefully I haven't fixed one thing at the cost of breaking another. This was added at revision 18962 by jesse -- any thoughts on what the intent was here? (net.refractions.udig.project/src/net/refractions/udig/project/internal/render/SelectionStyleContent.java)

There is identical code in SLDContent -- and there we have a problem where when we hit the WFS, visiting all the features a calculating the min, max, quantile, etc, the datastore just pukes on us... but it does work on occasion. I had hoped commenting out the same lines would magically fix things, but we'll have to look at this again tomorrow and figure out what is really going on...

Cheers,
Cory.
_______________________________________________
User-friendly Desktop Internet GIS (uDig)
http://udig.refractions.net
http://lists.refractions.net/mailman/listinfo/udig-devel

Reply via email to