Polygon and MultiPolygon Methods !

Topics: Algorithms
Jun 28, 2007 at 12:45 PM
Hi evereybody ;)
(sorry for my poor english, I'm french...)
I implement some methods to get geometries from Oracle Spatial. I know, this class is in SharpMap yet but it's very very slow so I develop the same methods with an XML reader (when WKB needs 4 minutes, XML needs only 8 seconds :ppp).
Nevertheless, I have a little problem: I need operate on Polygon and MultiPolygon (Intersection, Distance, Union, ...) but lot of methods are not implemented ! I would know if anybody knows when these methods will be developped ?
Thanks for answering !
Salut tout le monde ;)
(désolé pour mon pauvre anglais, je suis français...)
J'ai implémenté quelques méthodes pour récupérer des géométries d'Oracle Spatial. Je sais que cette classe existe déjà mais elle est très lente donc j'ai développé les même méthodes mais en utilisant un lecteur XML (quand WKB met 4 minutes, XML met seulement 8 secondes :ppp).
Cependant, j'ai un petit problème : j'ai besoin de réaliser des opérations sur des Polygon et des MultiPolygon (intersection, distance, union, ...) mais beaucoup de méthodes ne sont pas implémentée. Je voudrais savoir si quelqu'un sais quand est-ce que ces méthodes seront développées ?
Merci de vos réponses !
Oct 15, 2007 at 9:13 AM

Does anyone know if any progress has been made with the unimplemented methods?

I just tried the polygon.union() method but its not implemented according to the error message i get.

Oct 15, 2007 at 3:58 PM
SharpMap v2.0 Beta 2 will implement these, since it will use NTS to replace the current geometry object model.
Oct 15, 2007 at 11:55 PM
For what it is worth... We've been using NTS for Geometry functions fairly intensely for the last three weeks... things like autobounding, buffering complex linestrings, unioning polygons and multi-polygons. I am very impressed with its performance. I find low-level functions that bark, w/o error trapping on some of the junk geometry we have to throw at it (real time GPS data can be pretty ugly). We've added some high-level code to create a log file of exactly what the operation is that we are trying to peform, the source data, and the location of the bark. I'll make all of this available in a couple of weeks, and we can decide if we need more error handling at a low level, or what the behavior should be at the high level.

Anyway... kudo's to the NTS effort. Well done, Diego, et al.
Oct 16, 2007 at 10:12 PM

Magnum4610 wrote:
Anyway... kudo's to the NTS effort. Well done, Diego, et al.

Thanks a lot ;)