jorisvandenbossche commented on pull request #12137: URL: https://github.com/apache/arrow/pull/12137#issuecomment-1011995754
I would expect that for non-matching resolution in the input, we cast to the most detailed resolution. But, this doesn't necessarily need to have kernels for all combinations of resolutions. I think the casting step before calling the kernel should take care of that (like `add(int64, int32) -> int64` is not a registered kernel (only `add(int64, int64)` is), but still works). The same is true for `subtract(timestamp, timestamp) -> duration`, which is for now only implemented for matching resolutions, and also the automatic casting doesn't seem to work yet for timestamps. -- 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]
