stefan-grafberger commented on code in PR #635: URL: https://github.com/apache/parquet-format/pull/635#discussion_r4466488992
########## LogicalTypes.md: ########## @@ -743,14 +745,15 @@ are: equality comparison and not otherwise interpreted. `checksum` applies to the resolved bytes, except for `ETAG`, which is the -object-store eTag for the whole file referenced by `uri`. +object-store eTag for the whole file referenced by `uri`. If `inline` and `uri` are both +set, `checksum` describes the source at `uri` as it was when `inline` was copied from it. +Like the locator, it is then provenance only. ##### inline The referenced bytes stored inline in the value. If `inline` is set, it supplies the -bytes and any locator fields (`uri`, `offset`, `size`) that are set record where those -bytes came from. A reader may resolve the value from `inline` or from the locator, -whichever suits it. +bytes and any locator fields (`uri`, `offset`, `size`) that are set are provenance +only. Review Comment: Yeah, a bit like cache invalidation, but readers never invalidate anything. Refreshing would always be an explicit write that produces a new value. My understanding is that this cached-copy use case is what #603 was aiming for :) The wording in this PR is the wording from the initial spec PR (#585), so I thought that might be easiest to agree on, as it was already agreed on once. However, happy to change this if everyone agrees on a better wording! What about keeping the sentence from before this PR and only replacing the "either" part: "If inline is set, it supplies the bytes and any locator fields (uri, offset, size) that are set record where those bytes came from. Readers resolve the value from inline, not from the locator." I'd then also drop "provenance only" from the checksum sentence. -- 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] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
