How to test for Esri ArcPro Arc

Hi ​@juliarozema I tried multiple ways of splitting, reading and rebuilding the line but I see no arc. Linestring on read and on out. In the example you provided (thanks for providing!!) I could not see an arc. But this still fails or SDE? In the process of being written to SDE is there a writer that is defining these as Arc?
 

Thank you, ​@crutledge  for taking a look at this.
Here are further details.

Within our GIS data, we are able to have ARCs, but the data has to be loaded into another system and it doesn’t accept ARCs.  So, we use FME to check the data and let us know if an ARC snuck into our data.  The secondary system says that the segment is an ARC, but maybe it is actually classified as something else within the SDE?

Using the sample feature from my first post, when I go into ArcGIS Pro and select “Edit Vertices” I see the node-handle bar things that are used to adjust an Arc (see image). 

For road segments without an Arc, usually I only see vertices.

Might this sample Road Segment that seems like an Arc, be classified as something other than an Arc?

And, how can we use FME to test for it?

Thanks again for your thoughts and help

The Road Segment has a spline geometry, which is another type of curve. FME does not support spline geometry, so the reader is automatically stroking it to linestring that closely follows the spline path. This is generally sufficient for working with the data, but there is no way to distinguish a stroked spline from a regular polyline.

here is the FME geometry model that helps understand the fme supported geometries

https://docs.safe.com/fme/html/FME-Form-Documentation/FME-ReadersWriters/!FME_Geometry/FME_Geometry_Model.htm

Thank you, to each of you, for looking into this for me.

Nutshell:
With FME transformers, there is no way to test for a spline as this is not currently a geometry supported by FME (but maybe it will be?).

I can look into ArcPy functions to see if there is a way to isolate a spline and, if so, then use a Python Caller to do this within FME.

​@juliarozema Yeah. I am at a loss here. I think this should go in the suggestion box. My only other suggestion is to test a combination of the following and see if chopping, geometry coercing etc can ‘get rid’ of that phantom geometry type.
Can you get a report of what fails and then re-process those pieces. Not elegant but….
 

If I’m understanding correctly, you want FME to identify Arcs but there’s no requirement for FME to actually process the data? The SDE data goes straight to destination but fails if it contains Arcs/Splines, so you need to identify and rectify at source?

In ArcPy you can access SHAPE@JSON which will give you the json representation that will include the word curve

{"curvePaths":[[[94903.242399999872,5908395.4097000007],{"b":[[94942.239400000311,5908341.5438000001],[94918.435656980277,5908385.1962123485],[94942.307510492625,5908376.6501305643]]},[94943.542000000365,5908337.5573999994]]],"spatialReference":{"wkid":102185,"latestWkid":3401}}

​@ebygomm Yes!  This is exactly the situation.  Just need to identify the Arcs/Splines.

Might you be wiling to share an example of what is needed within the PythonCaller to access the SHAPE@JSON?

Or is there a transformer that can also extract this info?

Thank you for your help.

The geometry filter can test for arc geometries type (though not sure how it will go with the splines)
https://docs.safe.com/fme/html/FME-Form-Documentation/FME-Transformers/Transformers/geometryfilter.htm

You can then use an arc stroker or geometry coercer to turn the arcs into lines
https://docs.safe.com/fme/html/FME-Form-Documentation/FME-Transformers/Transformers/arcstroker.htm

https://docs.safe.com/fme/html/FME-Form-Documentation/FME-Transformers/Transformers/geometrycoercer.htm

I think I have an example of how to identify the objectids of features with splines using a python caller, I will see if I can locate it.

import fme
import fmeobjects
import arcpy


class FeatureProcessor(object):

    def __init__(self):
        pass

    def input(self, feature: fmeobjects.FMEFeature):
        arcpy.env.workspace = FME_MacroValues['SourceDataset_GEODATABASE_FILE']
        fc = 'ARCS'
        with arcpy.da.SearchCursor(fc, ["OID@", "SHAPE@JSON"],feature.getAttribute('datequery')) as cursor:
            cursor.reset()
            for row in cursor:            
                if "curve" in  cursor[1]:
                    feature.setAttribute("OBJECTID",cursor[0])
                    feature.setAttribute("ContainsCurves","True")
                    self.pyoutput(feature)




    def close(self):

        pass

This produces a feature with the objectid for any feature that contains curves. The datequery attribute is used to filter only recently edited features. I can’t say how performant it is, it was only used to check a handful of recently edited features on each run