codewarnab commented on issue #5981: URL: https://github.com/apache/arrow-rs/issues/5981#issuecomment-6724410784
Hi, this issue is now more than 2 years old. I would like to work on it. Steps 1 and 2 are done: - #7940 removed the option that preserved dict IDs. - The deprecation started in 54.0.0, so that window has passed. Step 3 is still open. `Field` still has `dict_id`. The IPC reader still uses it in two places: 1. `arrow-ipc/src/reader.rs`: it reads `field.dict_id()` while decoding a record batch. 2. `arrow-ipc/src/reader.rs`: it looks up a field by dict ID while reading a dictionary batch. My plan: when the reader parses the schema, it builds the dict ID mapping once. The reader then walks that mapping alongside the fields, so it does not need `Field::dict_id`. After that, I can remove `dict_id` from `Field`. A PR for this would close the issue. Removing it is a breaking change, so it is a good fit for the next major release. Can I take this? @Jefffrey @alamb -- 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]
