SF geometry tables are currently presented as vectors of vectors. However, this loses important type information (polygons and multilinestrings have the same structure), which could be picked up from the R data. Wrapping these in Julia geometry is trivial with GeoInterface.jl wrapper geometries. These also allow the CRS to be stored along with the object.
Would there be interest in doing this, and where could I begin? All of the attributes necessary are available when convert = false, so it's only a matter of processing on the Julia end.
SF geometry tables are currently presented as vectors of vectors. However, this loses important type information (polygons and multilinestrings have the same structure), which could be picked up from the R data. Wrapping these in Julia geometry is trivial with
GeoInterface.jlwrapper geometries. These also allow the CRS to be stored along with the object.Would there be interest in doing this, and where could I begin? All of the attributes necessary are available when
convert = false, so it's only a matter of processing on the Julia end.