Skip to main content

I am trying to write some CAD data to an ArcGIS Enterprise Feature Service.

I noticed there has been an update in the latest version but unfortunately we are using 2025.0 at present and wanted to see if there is a work-around to an issue related to true-curves.

 

I am getting the following error message: 

ArcGISFeatureServiceWriter: Feature type 'Levels' has 'Use True Curves' enabled, but the feature service does not support true curves
Python Exception <FormatException>: FMEException: 1: ARCGISFEATURESERVICE: Feature type 'Levels' has 'Use True Curves' enabled, but the feature service does not support true curves
SAFE.ESRI-AGOL.ARCGISFEATURESERVICE writer: An error has occurred. Check the logfile above for details

 

The writer has ‘Use True Curves’ set to ‘No’ but I still get the error message.

 

 

Is there are way to remove ‘true curves’ or write the data to our ArcGIS Enterprise Portal to by-pass the issue?

It’s a bit of a long shot, but you might want to try the ArcStroker transformer.


 

To prevent the issue when loading the data to AGOL or Portal, I use a geometry filter to isolate the true curves (geometry type = Arc), then I use an ArcStroker to turn the true curves into a series of interpolated line segments. Then I combine the 2 data streams using a junction.

This screenshot is from FME v 2021.2, but I can’t imagine the workflow would be much different in v 2025.0

 

 


Thanks ​@karol_works and ​@hvukasinovic I have added the arcstroker and still get the same error. I will keep testing and see what I can figure out but I appreciate you taking the time to respond.


Hi ​@deanhowell if you filter for the curves with the GeometryFilter or maybe the GeometryPartExtractor (specifying "a child of this part has a Geometry Type of IFMEArc." in the XQuery), do non-arcs write fine to your feature service? Does this give a new error?

apart from stroking the arcs, you could maybe put the data through the GeometryValidator to see if they get repaired at all. I saw in another thread that a PythonCaller could help as well, but I’m not sure myself. 

If these suggestions don’t help and you’re still stuck, could you provide some sample data (e.g. a few features with the curves in question) for us to test?