alamb commented on PR #9160: URL: https://github.com/apache/arrow-datafusion/pull/9160#issuecomment-1935215131
> So my question is would you rather see a new compression type added for `BGZIP` which mostly mirrors GZIP, but adds stream decoding of multiple members and explicit errors when trying to compress to bgzip... what the PR starts to do... or call `multiple_members(true)` by default so that the gzip async and non-async decoding would both support multiple members? I personally think what is important is to handle reading from files compressed with normal `gzip` (which I think means single member) If this can be accomplished by allowing the gzip encoding to handle multiple members at the same time, that seems great to me. Returning an error when the underlying decompression library can handle the data seems like a poor user experience to me and overly pedantic Hope that helps -- 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]
