Closed asjohnston-asf closed 1 year ago
@asjohnston-asf Is there any reason why we don't just use the ESRI-shapefile format? That's what the shorelines file is, and I see that it's supported in the list that you linked. Is it just that the extracted portion of the shapefile is rendered in GeoJSON right now? Is it much harder to just keep it in the shapefile format?
That was my first thought, my only (very minor) hesitation is that format is multiple files, so we might have to tweak the temp file details to clean up the temporary sidecar files.
That makes sense. I guess we could test the bumped-up GeoJSON approach to see if it does the trick, and try out the shapefile approach if there are issues (or concerns about what it would take to find an appropriate object size limit).
I've confirmed the trufflehog failure is a false-positive related to story map URLs already merged to develop.
Here's a crude and quick fix to the
ERROR 1: GeoJSON object too complex, please see the OGR_GEOJSON_MAX_OBJ_SIZE environment option
errors we see when processing these scenes over the Aegean Sea:S1B_IW_SLC__1SDV_20201115T162313_20201115T162340_024278_02E29D_5C54
S1A_IW_SLC__1SDV_20201203T162353_20201203T162420_035524_042744_6D5C
The
OGR_GEOJSON_MAX_OBJ_SIZE
setting is specific to the GDAL GeoJSON driver, documentation is at https://gdal.org/drivers/vector/geojson.html#configuration-optionsWe're using GeoJSON for the temp file out of convenience, we might be able to avoid this issue by using any other GDAL-supported vector format.
Alternatively, if we stick with GeoJSON, it would be nice to set
OGR_GEOJSON_MAX_OBJ_SIZE
based on the size of the temporary geojson file instead of to an arbitrary big value.@jacquelynsmale any thoughts on those options? Would you be interested in taking a stab at implementing either? I've only checked that this code still passes the unit tests, I haven't yet verified it fixes the issue for the specific pair in question.