YuweiXIAO

> Does this implementation of external table support handle partitioned Parquet datasets for example, when using wildcard paths or recursive directory patterns such as: > > `read_parquet('/path/to/data/**/*.parquet')` > > In...

> Thanks for the work on this! I also had something like this in mind, but I was thinking about using FOREIGN TABLES instead of table access methods for this....

> Thinking about it more, I do think FOREIGN TABLE is a better fit for this semantically. Because the CREATE TABLE command that you have now isn't actually creating the...

Hi @JelteF , the PR is ready for review: - Switched to the `FOREIGN TABLE` interface for table management. - A pre-created server (`ddb_foreign_server`) is added in the extension SQL...

Not reproducible on my side either. By the way, how do you run `scan_postgres_tables` in parallel within a single installcheck command? The tests might conflict due to the same table...

If we create indexes for those meta tables, the scan IO should be much better.

hi @JelteF , I am working on `copy from stdin` for `pg_mooncake`. Could you share any tech preference or direction for supporting `copy_from_stdin` in pg_duckdb? particularly if we're not planning...

Thanks for the insight! I am considering redirecting `COPY FROM STDIN` to local temp files on the server, and then triggering a DuckDB query (COPY FROM 'file://local') to load data...

For the TAM approach, we might need to implement TupleSlot cache to reduce the overhead of `pg_row->duckdb_chunk` conversion. And we will lose parallelism on the DuckDB side, right?

ah, supporting JSON/PARQUET format is a great idea! For the temp file approach, we could implement DuckDB loading in mini-batches and asynchronously. Specifically, the main thread keeps streaming data to...