jorisvandenbossche commented on PR #702: URL: https://github.com/apache/arrow-adbc/pull/702#issuecomment-1561532300
> Is your idea to create separate capsules for every object or just one capsule to expose the API? We do the latter in pandas to better share functions across extensions: Yeah, that's a bit a different kind of use case for capsules. The idea here is not to expose some module's C API, but indeed the individual objects (result sets). For context, see also the descriptions in https://github.com/apache/arrow/issues/34031 and https://github.com/apache/arrow/issues/35531. The end goal would be to be able to do something like `cursor.fetch_arrow_stream()` which returns an object that has a `__arrow_c_stream__` method which returns a capsule wrapping a C Stream struct pointer, and so which can be consumed by any library that can consume the C interface (duckdb, datafusion, etc) without relying on pyarrow. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
