[GitHub] [iceberg] rymurr commented on pull request #2286: Add Arrow vectorized reader

[GitBox <gi...@apache.org>]

rymurr commented on pull request #2286:
URL: https://github.com/apache/iceberg/pull/2286#issuecomment-789712836


   Hey @mayursrivastava Ive gotten part of the way through this. Expect a more complete review soon. Early comment:
   
   I am unsure of exposing the `VectorSchemaRoot` directly. Its not an easy object to work with for non-experts and creates potential lifetime/ownership issues. I think a wrapper object similar to Sparks `ColumnBatch` may be more accessible for people who just want the speed of vectorised reads w/o the effort of directly reading Arrow data structures. Though I could see the desire to still expose an advanced user mode to consume the `VectorSchemaRoot` directly with arrow friendly tools.
   
   What do you think? What were the use cases you were thinking of for this?


----------------------------------------------------------------
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:
users@infra.apache.org



---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For additional commands, e-mail: issues-help@iceberg.apache.org