Specification for determining file path when inserting data?

[mickeyreiss]

The specification does not seem to include details about the file path that is chosen for each parquet file.

• How is the file path determined for each insert?
• Are there hive-style path parameters that are included for partitioned tables?
• Is there a way to configure DuckLake, so that the parquet files for each table are stored in a dedicated subfolder? (This would make it easier to read a DuckLake stored in GCS directly in BigQuery¹.)
• Will there be a way to configure DuckLake, so that the parquet files are stored in a dedicated folder per ingestion date or hour (similar to BigQuery's ingestion time partitioning²)?

Footnotes

1. https://cloud.google.com/bigquery/docs/hive-partitioned-queries#supported_data_layouts ↩

2. https://cloud.google.com/bigquery/docs/partitioned-tables#ingestion_time ↩

[Mytherin]

I've written up some docs on this here - https://ducklake.select/docs/stable/duckdb/usage/paths