You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@parquet.apache.org by Lars Volker <lv...@cloudera.com> on 2017/05/05 12:50:32 UTC
Can we add a special ColumnOrder for testing?
Hi All,
PR #46 <https://github.com/apache/parquet-format/pull/46> introduced
ColumnOrder with the limitation that a reader should ignore stats for a
column if the corresponding ColumnOrder in FileMetaData contains an unknown
value. In order to test this logic, it would be helpful to have a special
value InvalidOrder or UnsupportedOrder that would never be supported by a
reader. I assume this may be helpful to test other implementations, too.
Cheers, Lars
Re: Can we add a special ColumnOrder for testing?
Posted by Lars Volker <lv...@cloudera.com>.
Here it is: https://github.com/apache/parquet-format/pull/53
Let me know if we should have a separate JIRA for the Makefile path fix.
On Fri, May 5, 2017 at 6:40 PM, Ryan Blue <rb...@netflix.com.invalid> wrote:
> That sounds like a good idea to me. Want to submit a PR?
>
> On Fri, May 5, 2017 at 5:50 AM, Lars Volker <lv...@cloudera.com> wrote:
>
> > Hi All,
> >
> > PR #46 <https://github.com/apache/parquet-format/pull/46> introduced
> > ColumnOrder with the limitation that a reader should ignore stats for a
> > column if the corresponding ColumnOrder in FileMetaData contains an
> unknown
> > value. In order to test this logic, it would be helpful to have a special
> > value InvalidOrder or UnsupportedOrder that would never be supported by a
> > reader. I assume this may be helpful to test other implementations, too.
> >
> > Cheers, Lars
> >
>
>
>
> --
> Ryan Blue
> Software Engineer
> Netflix
>
Re: Can we add a special ColumnOrder for testing?
Posted by Ryan Blue <rb...@netflix.com.INVALID>.
That sounds like a good idea to me. Want to submit a PR?
On Fri, May 5, 2017 at 5:50 AM, Lars Volker <lv...@cloudera.com> wrote:
> Hi All,
>
> PR #46 <https://github.com/apache/parquet-format/pull/46> introduced
> ColumnOrder with the limitation that a reader should ignore stats for a
> column if the corresponding ColumnOrder in FileMetaData contains an unknown
> value. In order to test this logic, it would be helpful to have a special
> value InvalidOrder or UnsupportedOrder that would never be supported by a
> reader. I assume this may be helpful to test other implementations, too.
>
> Cheers, Lars
>
--
Ryan Blue
Software Engineer
Netflix