[
https://issues.apache.org/jira/browse/BAHIR-35?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15386642#comment-15386642
]
Christian Kadner commented on BAHIR-35:
---------------------------------------
essentially, yes
> Include Python code in the binary jars for use with "--packages ..."
> --------------------------------------------------------------------
>
> Key: BAHIR-35
> URL: https://issues.apache.org/jira/browse/BAHIR-35
> Project: Bahir
> Issue Type: Task
> Components: Build
> Affects Versions: 2.0.0
> Reporter: Christian Kadner
> Original Estimate: 8h
> Remaining Estimate: 8h
>
> Currently, to make use the PySpark code (i.e streaming-mqtt/python) a user
> will have to download the jar from Maven central or clone the code from
> GitHub and then have to find individual *.py files, create a zip and add that
> to the {{spark-submit}} command with the {{--py-files}} option, or, add them
> to the {{PYTHONPATH}} when running locally.
> If we include the Python code in the binary build (to the jar that gets
> uploaded to Maven central), then users need not do any acrobatics besides
> using the {{--packages ...}} option.
> An example where the Python code is part of the binary jar is the
> [GraphFrames|https://spark-packages.org/package/graphframes/graphframes]
> package.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
