embedded-database

In Big Query there is a function array_concat_agg that aggregates array fields by concatenating the arrays. In Snowflake there is a flatten function that can unnest nested arrays into single array. I am looking for similar functionality in duckdb.

select flatten([[1, 2], [2, 3], [4, 5]] would return [1, 2, 2, 3, 4, 5]
I would also need a distinct option:
`select flatten(DISTINCT [[1, 2],

Right now, to get an acceptable experience in dtlv, one has to use rlwrap, i.e. launch it like this rlwrap dtlv. It would be nice to integrate with jline3 library to have a better experience natively.

https://github.com/risinglightdb/risinglight/blob/7b5bc7a1a258afd64a483990ca95f4c799482791/src/storage/secondary/version_manager.rs#L258-L263

Currently, users will need to manually call pin and unpin. If there are futures cancelled or somehow, it is possible that we fail to unpin by ourselves.

We should implement a struct called Version, and implement Drop for it. pin will return a

The persistence layer of CloverDB is abstracted by the StorageEngine interface. The default implementation persists data on disk only and makes use of the badger kv-store.
It would be useful to also add an alternative StorageEngine implementation running completely in-memory, arranging documents in a map.