Background
Querying Overture via DuckDB requires knowing the exact S3 release path.
labs.overturemaps.org/data/latest.ddb solves this with pre-built views,
but it's a separate artifact that must be regenerated each release.
Proposal
Add a duckdb field to the STAC catalog JSON with SQL that creates one view
per type pointing at the correct S3 paths:
curl https://stac.overturemaps.org/catalog.json | jq -r .duckdb | duckdb overture.db -c -
# then: SELECT * FROM place;
Background
Querying Overture via DuckDB requires knowing the exact S3 release path.
labs.overturemaps.org/data/latest.ddbsolves this with pre-built views,but it's a separate artifact that must be regenerated each release.
Proposal
Add a
duckdbfield to the STAC catalog JSON with SQL that creates one viewper type pointing at the correct S3 paths: