Closed robyngit closed 1 year ago
I am going to start working on this. I think we should restructure things a little so that: 1) If there is a deduplication method set in the config, then duplicates are always identified during the staging process 2) The deduplicate_at config option will be interpreted as the step(s) at which the duplicates should be removed (staging, raster, and/or 3dtiles)
This feature is working in the feature-9-flag-duplicates
branch. I've tested with both the neighbor and footprints method, but we will need to make updates to other parts of the workflow before we can implement this:
The new deduplication strategy is now merged into the main branch of all relevant packages. Briefly, with the new strategy:
staging_duplicated
) is added that identifies polygons as duplicates or notdeduplicate_at
config option is interpreted to mean the step(s) at which polygons flagged as duplicates should be removed
For the IWP dataset, we decided to keep geometries that are classified as "duplicates" during processing in the version of the geopackage tiles that we archive. We will, however, mark duplicate geometries with a "duplicates" property. We therefore need to update the deduplication method to optionally keep duplicated geometries, but set a new "duplicated" property to True.
Duplicates should still be removed during the creation of the web products (3d tiles, PNG web tiles)