Hi Igor,

I think it's also good idea to change loading logic but it would be
affected all step while loading, thus you should be careful to change those
codes, especially side-effect. one good idea is to copy
interpreter-setting.json including r settings when user build it with
`-Pr`. It looks a bit weird but because zeppelin-zrinterpreter.jar doesn't
have to include interpreter-setting.json inside it, it's ok to have all
interpreter-setting.json in zeppelin-spark.jar. I'm willing to review your
new PR though. :-)

Regards,
Jongyoul

On Mon, Jan 23, 2017 at 7:26 PM, Igor Drozdov <igor_droz...@epam.com> wrote:

> Thank you for reply!
>
> I found an issue https://issues.apache.org/jira/browse/ZEPPELIN-1081. I
> think It's good idea not to include R interpreter with R profile disabled.
>
> Small changes to InterpreterFactory will allow us to load several
> interpreter-settings.jsons from different jar.
> Interpreter/zeppelin-spark.jar will contain settings with spark, sql, dep
> and pyspark interpreters and zeppelin-zrinterpreter will contain r and
> knitr interpreters. What do you think of it?
>
> Igor Drozdov
>
> -----Original Message-----
> From: Jongyoul Lee [mailto:jongy...@gmail.com]
> Sent: Friday, January 20, 2017 7:02 PM
> To: dev <dev@zeppelin.apache.org>
> Subject: Re: R interpreter
>
> For avoiding it, you can make interpreter-setting.json file in a spark
> directory which includes all of spark and r settings. I think it's
> realistic solution for now.
>
> On Fri, Jan 20, 2017 at 5:13 PM, Igor Drozdov <igor_droz...@epam.com>
> wrote:
>
> > After I build zeppelin with R profile I got zeppelin-zrinterpreter
> > binary in interpreters/spark folder
> >
> > -----Original Message-----
> > From: Jeff Zhang [mailto:zjf...@gmail.com]
> > Sent: Thursday, January 19, 2017 4:51 PM
> > To: dev@zeppelin.apache.org
> > Subject: Re: R interpreter
> >
> > That is SparkR, R interpreter is in the r folder under the root of
> > zeppelin.
> >
> >
> >
> > Igor Drozdov <igor_droz...@epam.com>于2017年1月19日周四 下午8:06写道:
> >
> > > Hello!
> > >
> > > I'm working on https://issues.apache.org/jira/browse/ZEPPELIN-921
> > > and I have some doubts about R Interpreter.
> > >
> > > Currently, R interpreter jar is in the same directory as Spark
> > > interpreter. InterpreterFactory searches for interpreters in spark
> > > dir, successfully loads interpreter-setting.json for spark and stops
> > > scanning this folder. So R interpreter is never loaded!
> > >
> > > Could anyone confirm that R interpreter does not work for now? Or it
> > > is my local misconfiguration?
> > >
> > > Thank you,
> > > Igor Drozdov
> > >
> >
>
>
>
> --
> 이종열, Jongyoul Lee, 李宗烈
> http://madeng.net
>



-- 
이종열, Jongyoul Lee, 李宗烈
http://madeng.net

Reply via email to