maartenbreddels commented on pull request #8755: URL: https://github.com/apache/arrow/pull/8755#issuecomment-733078145
Ok, I think that would solve the issue, because `Read(int64_t nbytes, void* out)` would then *not* call `.read` and hence not see the memoryview. I'm ok with that solution, but such a file object would not be compatible with e.g. Pandas. Also, as you mentioned, returning a memoryview from `.read` is not really standard. Do you think we should introduce another method (read_buffer) to return a memoryview/buffer instead? (Btw, to make this concrete, I have a memory mapped file which I expose using the File, which I do not want to copy at all) ---------------------------------------------------------------- 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. For queries about this service, please contact Infrastructure at: [email protected]
