scovich commented on issue #6528:
URL: https://github.com/apache/arrow-rs/issues/6528#issuecomment-3724891237

   > The logic would ensure that the new `NullBuffer` is a subset of the 
existing validity (i.e., we can turn valid slots into nulls, but we cannot turn 
null slots back into valid values). This satisfies the primary use case 
mentioned (combining existing nulls with a new filter) while guaranteeing we 
never typically unmask garbage or undefined data.
   
   I wonder if we should just define the method as unioning the new null buffer 
with the existing one? Trivial to document (just link `NullBuffer::union`), and 
has the same cost as detecting an error but without needing an error. I don't 
think anyone was complaining that `NullBuffer::union` was too slow... only that 
there's no easy way to apply the result back to the target array afterward?


-- 
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]

Reply via email to