Open BlakeDraper opened 7 years ago
I am not very familiar with the setup of the geojson codec, but I looked at it a bit and it seems that a bbox is calculated behind the scenes for each feature via the .net FeatureBase in Jeremy's WiMServices. I'm thinking it is up to the client app to determine the full extent based on some kind of spatial loop over the feature's bbox property. Is this a feasible solution? If not, I'd have to spend some more time digging in to what Jeremy has going on in the inherited WiMservices project to find a solution, if one is possible.
The best solution is just to drop the bbox property from the output - it's not required, and a point feature should not have a bbox anyway. I am not using the bbox property in any WiM app or trying to, but having it in there and being invalid makes the whole geoJSON output invalid, and that may be why AGOL is rejecting it. Even if it isn't the cause of the AGOL failure, we should fix it anyway so we are returning valid geoJSON.
the bbox property is being returned from @jknewson WiMServices Feature class. Jeremy, should this be stripped out of WiMServices, or should I try to remove this property in the loop that builds my FeatureCollection in the STNGeoJsonDotNetCodec?
Ensure refactored services are using valid geojson
Marty noticed that the geoJSON endpoint is failing a geoJSON linter. He discovered this:
"bbox":[-96.08236,29.91331],
bounding box should not be a point, should be 4 points.Example endpoint that has this problem: https://stn.wim.usgs.gov/STNServices/Events/180/HWMs.geojson