I'm using org.locationtech.geomesa.spark.jts.util.RowGeoJSON to write out GeoJson from a DataFrame like so: ``` val geoJsonDF = ??? // First data frame column is a GeometryUDT geoJsonDF.toGeoJSON(0).write.text(outputUri.toString) ``` One of the other columns written to GeoJson properties is a "name" field that can contain double quotes, like so: ``` val exampleName = "Julie \"Jules\" Hoover Community Center" ``` When I use the `toGeoJSON` method above with strings that contain these double quotes, I get a line delimited geojson feature that looks like: ``` { "type": "Feature", "geometry": ..., "properties": { "name": "Julie "Jules" Hoover Community Center" }} ``` This is invalid JSON and causes issues when I attempt to stream the output geojson into other downstream tools such as Tippecanoe. I would expect something like: ``` { "type": "Feature", "geometry": ..., "properties": { "name": "Julie \"Jules\" Hoover Community Center" }} // or { "type": "Feature", "geometry": ..., "properties": { "name": "Julie 'Jules' Hoover Community Center" }} ``` |