kevinjqliu commented on issue #3063: URL: https://github.com/apache/iceberg-rust/issues/3063#issuecomment-5553987430
I looked into apache/avro-rs#654 and I don't think it actually blocks us. It only drops `logicalType` when parsing schema JSON. The only code we have that reads that attribute is `avro_schema_to_schema`, and nothing calls it outside its own tests. Writing is unaffected, so manifests come out the same as before. For the failing test, I made `avro/schema.rs` also recognize the map-as-array shape, i.e. an array whose item is a record named `k<keyId>_v<valueId>` with `key` and `value` fields. Java, PyIceberg and iceberg-go all write it that way so there's no ambiguity. The attribute check still runs first, so once 0.23 is out with the fix we can just drop the shape check. I've got the upgrade working locally with this plus the API changes. Want me to open a PR, or would it be easier if I send you the diff for your draft? -- 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]
