Ryan Blue

@swapnilushinde, that use case sounds reasonable so let's add them back. Are there other commands that make sense to have a glob also? Someone is adding it to the schema...

@swapnilushinde thanks! I'll take a look soon-ish.

Looks fine to me.

@gszadovszky, `ParquetProperties` isn't considered part of the public API. We have needed to change it over time. See https://github.com/apache/parquet-mr/commit/6b605a4ea05b66e1a6bf843353abcb4834a4ced8#diff-84c5123768d3411558f1761aca7559a9 for an example.

None of the org.apache.parquet.column classes are public (see https://github.com/apache/parquet-mr/blob/master/pom.xml#L250). I know it is annoying to not have a public API, but I think it is much worse to slow development...

I don't agree with a requirement for full binary compatibility across the entire codebase because we lack a public API. We already have significant drag from maintaining compatibility in the...

@aeroevan, can you post a simple test that reproduces the problem? I'd like to make sure that this has a test so it doesn't become a regression. That will also...

@zivanfi, @gszadovszky, can you explain how this was fixed?

Sounds great, thank you!

@isnotinvain, is the solution to the empty union simply that thrift can't project a subset of the union's branches? I don't really like the idea of filtering rows that don't...