[
https://issues.apache.org/jira/browse/DRILL-4760?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15360738#comment-15360738
]
Paul Rogers commented on DRILL-4760:
------------------------------------
The last section is entitled, "Data Description Language Support" and has this
comment: "Implement custom storage plugins to create Parquet readers/writers
for formats such as Thrift."
I believe this may be factually wrong. Parquet provides adapters to write
Thrift objects to Parquet, and to convert a Thrift schema to a Parquet schema.
But, once the data is on disk, it is not obvious that the data "remembers" the
source of the schema. Drill's Parquet reader should be able to read data
written from any source schema. (More research is needed to verify.)
> Add technical description of Drill's Parquet support
> ----------------------------------------------------
>
> Key: DRILL-4760
> URL: https://issues.apache.org/jira/browse/DRILL-4760
> Project: Apache Drill
> Issue Type: Improvement
> Components: Documentation
> Affects Versions: 1.7.0
> Reporter: Paul Rogers
> Priority: Minor
>
> Drill documentation (http://drill.apache.org/docs/parquet-format/) provides a
> description of Drill's Parquet support. However, the material is mostly at
> the "marketing level", describing how wonderful the format is. Those who
> actually build systems would greatly benefit from improved technical content.
> This entry is a place to gather the required information so it is available
> to the Doc. team.
> Some of the proposed material might go into a new "Develper Info" section
> rather than clutter the existing user-oriented Parquet page.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
