Swapped lat long

[tfili]

@keyboardspecialist I had to do this for some japan data too so I hijacked your branch. We should try and detect this but for now instead of using a branch, I think we should just add an envvar to the released version so we can use it when we need to.

I thinks I fixed it for real. See https://gdal.org/en/latest/tutorials/osr_api_tut.html#crs-and-axis-order

We need to test his a bunch before we can merge. I'll create a asset-pipeline PR with it so we can run some tests.

Testing

• Run ./citygmlswayzetests in the asset-pipeline update-libcitygml branch - there was a test added to sets the env var
• Verify everything passes
• For extra validation you can run CITYGML_USE_SRS_ORDERING=1 ./citygmlswayzetests
• Verify there are failures with the Lat/Lon being swapped

[keyboardspecialist]

Awesome, glad to see a real fix.

Here's the issue I opened earlier this year https://github.com/CesiumGS/asset-pipeline/issues/1528

I was on the right track, but never tried that flag.

[tfili]

@shehzan10 you know more about CityGML. Is there a reason that the incorrect ordering was specified on purpose? Should it be an option?

[shehzan10]

@tfili We used the previous option based on the GDAL documentation. If you are confident that the new flag is the right one to use to always return [long, lat], then lets go with that.

It wasn't so much that the incorrect ordering was on purpose. The EPSG:6697 CRS is the only one I've seen that uses Lat, Long ordering, and we couldn't figure out a way until now to always make the function return long, lat regardless of input.

[tfili]

Looking into it a bit more, it seems like CityGML generally expects traditional GIS order. Looking at https://gdal.org/en/latest/drivers/vector/gml.html it is what GDAL defaults to when loading CityGML. I think the best option is to use an env var for us to swap it if needed for our tilesets. We may want to potentially expose it to customers at some point but I don't think there has been a need so far.

[tfili]

@keyboardspecialist This is ready for review.

[keyboardspecialist]

I'm getting a ton of test failures on my local linux build (release and debug)

[  FAILED  ] 24 tests, listed below:
[  FAILED  ] CityGMLSwayzeMainTests_GeoRES.CountModels
[  FAILED  ] CityGMLSwayzeMainTests_GeoRES.CountModelInstances
[  FAILED  ] CityGMLSwayzeMainTests_GeoRES.CountTextures
[  FAILED  ] CityGMLSwayzeMainTests_GeoRES.ReadProperties
[  FAILED  ] CityGMLSwayzeMainTests_GeoRES.CountBatchedAndInstanced
[  FAILED  ] CityGMLSwayzeMainTests_GeoRES.TotalPrimitivesAndTexels
[  FAILED  ] CityGMLSwayzeMainTests_GeoRES.BatchTablePropertiesAreSet
[  FAILED  ] CityGMLSwayzeMainTests_GeoRES_LOD2.ReadProperties
[  FAILED  ] CityGMLSwayzeMainTests_NYC12.CountModels
[  FAILED  ] CityGMLSwayzeMainTests_NYC12.CountModelInstances
[  FAILED  ] CityGMLSwayzeMainTests_NYC12.ReadProperties
[  FAILED  ] CityGMLSwayzeMainTests_NYC12.CountBatchedAndInstanced
[  FAILED  ] CityGMLSwayzeMainTests_NYC12.TotalPrimitivesAndTexels
[  FAILED  ] CityGMLSwayzeMainTests_NYCWTC.CountModels
[  FAILED  ] CityGMLSwayzeMainTests_NYCWTC.CountModelInstances
[  FAILED  ] CityGMLSwayzeMainTests_NYCWTC.CountBatchedAndInstanced
[  FAILED  ] CityGMLSwayzeMainTests_NYCWTC.TotalPrimitivesAndTexels
[  FAILED  ] CityGMLSwayzeMainTests_NYCWTC.GltfModelHasVertexColors
[  FAILED  ] CityGMLSwayzeMainTests_NYCWTCDisableColors.GltfModelDoesNotHaveColors
[  FAILED  ] CityGMLSwayzeMainTests_NYCWTCDisableVC.GltfModelDoesNotHaveVertexColors
[  FAILED  ] CityGMLSwayzeMainTests_GeoRESWithoutSRSWithSRSOption.CountModels
[  FAILED  ] CityGMLSwayzeMainTests_GeoRESWithoutSRSWithSRSOption.CountModelInstances
[  FAILED  ] CityGMLSwayzeMainTests_GeoRESWithoutSRSWithSRSOption.CountBatchedAndInstanced
[  FAILED  ] CityGMLSwayzeMainTests_GeoRES_Swapped.ReadProperties

[tfili]

@keyboardspecialist The tests seem to be passing here. Did you update submodules?

[keyboardspecialist]

@tfili For some reason my tests are failing when building through docker. Using my system's gcc works fine.

May be long overdue for a system clean.

[keyboardspecialist]

Looks good. Tests are passing for me now.