On 20/05/2010, at 4:53 PM, Michael Bedward wrote: > Hi Jody, > >> There is something you may of missed in the spec... >> >> See where it returns se:Function? That mandates that a "fallback" value is >> provided; I >> think that the fallback value may be used in cases where lookup value is not >> found. > > Unfortunately I don't think that's right. For functions generally, the > doc states... > > "The value of the fallbackValue attribute is used as a default value, > if the SE implementation does not support the function. If the > implementation supports the function, then the result value is > determined by executing the function." > > So "fallback" does not equal "default" value. Andrea pointed this > clause out when we were adding the "env" function a little while ago.
So we are back to interpretative dance. Weeeee.... What would you like to do? Returning null is nice in such cases in that it makes it someone else's problem. There is no sense of "order" here. You could make a convention for <Literal/>*</Literal> which matches any value; or an empty key (ie <Literal/>) which answers any value. >> You may also enjoy discussing these things on the [email protected] >> email list >> for a wider perspective. > > Yikes... This must be a definition of "enjoy" that I wasn't previously aware > of... Hey they are friendly. It could be worse you could get off the open source lists and on to the ogc lists (and then we would be forced to start an intervention). The [email protected] are still bound by actually wanting to implement. Jody ------------------------------------------------------------------------------ _______________________________________________ Geotools-devel mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/geotools-devel
