andishgar commented on issue #49740: URL: https://github.com/apache/arrow/issues/49740#issuecomment-4699766095
@fenfeng9 A few notes: 1. `ArrowArray` can be produced in many ways, and using builders is just one of them. 2. The most notable feature of `BinaryView`/`StringView` is that they can contain multiple data buffers. These buffers may or may not be referenced by the views, and the presence of non-deduced data buffers is entirely optional. 3. > I also don't see the spec explicitly clarifying this, but I tend to think this should be considered an invalid state. Regarding the above comment, the Arrow specification is intended to serve as a reference for other implementations. If a null data buffer is to be considered an invalid state, I think it would be important to verify whether other reference implementations (Julia, Rust, Java, Go, Erlang, etc.) also treat it as invalid. Even if all reference implementations consider a null data buffer invalid, this behavior should be explicitly documented in the specification to avoid ambiguity for other implementations. Perhaps after @pitrou confirms whether a null data buffer can cause problems, I should open a separate issue to clarify this in the specification. -- 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]
