I found a solution.
Basically you define several paragraphs.
1. utility paragraphs that prepare data and user defined functions,
collapsed to be hidden from end-users.
2. business paragraphs that use #1, like "%sql select ... from
defined_table1"
2015-09-08 20:07 GMT+04:00 Eugene <blackorange...@gmail.com>:
> In spark-shell I can use :load statement to include common file parts
> stored at hard drive.
>
> Is there a include facility in Zeppelin?
> I see this in output:
> <console>:2: error: illegal start of definition :load
> /home/hadoop/spark-shell-helpers/load-from-s3-parallel.scala
>
> I want to enable simple users to start using Zeppelin, but heavy lifting
> programming should be hidden from them, leaving only actual code where they
> will type in SQL.
>
> For example:
>
> :load /home/hadoop/spark-shell-helpers/load-from-s3-parallel.scala
> :load /home/hadoop/spark-shell-helpers/utility-udfs.scala
> val businessMeaningfulDF = sqlContext.sql("SELECT ...")
>
> --
>
>
> Best regards,
> Eugene.
>
--
Best regards,
Eugene.
