Details
-
Improvement
-
Status: Open
-
Major
-
Resolution: Unresolved
-
0.7.0
-
None
Description
We'd like to have paragraph's code generated by a preceding paragraph.
For example, one of the use cases we have is when %pyspark generates Hive DDLs. (can't run those in Spark in some cases)
Any chance an output of a paragraph can be redirected to a following paragraph?
I was thinking something like this could be used
https://zeppelin.apache.org/docs/latest/rest-api/rest-notebook.html#create-a-new-paragraph
http://[zeppelin-server]:[zeppelin-port]/api/notebook/[notebookId]/paragraph
But not sure if there is a easy way to call Zeppelin API directly through "z" variable?
Something like z.addParagraph(...)
In most cases a paragraph generates a SQL code that can't be run directly as Spark SQL and has to be run by a different engine, for example, by Hive or by a JDBC backend. It's one of the use cases that we have, but I am sure there are many more use cases too.
Reply from moon in email distribution list
Although you can always create your function that call Zeppelin's rest API
to add paragraph, providing capability to add paragraph through 'z' (more precisely, from Interpreter) helps provide user more interactive usage of notebook i think.