From: Rueben S. <r_j...@ya...> - 2004-10-01 03:56:13
|
On Wed, 2004-09-29 at 05:48, James Macgill wrote: > I actualy put in a first shot at prj reading for the Shapefile module > last week. At the moment it has a 'fix' method which converts one > name change before having it parsed. (Labert_Conformal_Conic) again > it lacks 1sp/2sp in ESRI's form. > > Feel free to dump this code if the problem can/should be passed on > down the line. > > James > > This should eventually be dumped, but we need to decide which CTS implementation will be used by udig first. The new referencing implementation will allow us to put all this logic about the inconsistent projection names into one place. Right now we are still using find/replace methods (noticably in the EPSGCoordinateSystemFactory) like you did. In the current implementation, I could probably add a new provider for the Lambert_Conformal_Conic (and Mercator). Rueben |