Closed jonded94 closed 1 day ago
BTW using the Row
API is likely to be quite slow to do processing with parquet. If you care about performance I would recommend looking at the other APIs that are available such as the arrow reader
If you care about performance I would recommend looking at the other APIs that are available such as the arrow reader
Yes, I implemented the data manipulation functionality that I need also already for whole RecordBatches & Arrays, but for a specific use case I also want to modify single rows. I'm aware that this is very much less efficient than it could be.
Thank you for your quick replies and collaboration!
Since you did a release only a few days ago, one can't expect a new release for another month, right?
Since you did a release only a few days ago, one can't expect a new release for another month, right?
That is correct. The next scheduled release is
Which issue does this PR close?
Closes https://github.com/apache/arrow-rs/issues/6761 Closes https://github.com/apache/arrow-rs/pull/6762
Rationale for this change
parquet::record::make_row
was the only method to createRow
instances and was private. With a class method::new
, there is a new way of creatingRow
instances that is publicly exposed.What changes are included in this PR?
parquet::record::make_row
parquet::record::Row::new
method.