[Connector]: Parquet

[bschne]

What do you use this database for?

The immediate need for this is that I have some custom tools running on personal data that output Parquet files. I could change that to CSV without too much hassle, but I think it might be generally interesting to have this option. I figure others will also have Parquet dumps of data they'd like to analyze, and it should be relatively simple.

Are there any existing node clients for this data source?

IIUC the existing CSV connector directly loads the CSV files into DuckDB for querying, and I think the same approach could be used here. See DuckDB docs on reading and writing Parquet files.

Are you interested in contributing this connector?

I'd like to contribute, and I'll open a PR

[archiewood]

You can actually use the existing DuckDB connector for this. Add the parquet files to the folder containing the DuckDB source information, and then in your source SQL you can write:

select * from 'my_file.parquet'