Details
-
Improvement
-
Status: Resolved
-
Major
-
Resolution: Fixed
-
None
-
None
-
None
Description
Currently, to select interpreter, user have two options.
- select interpreter by using '%[Name]' in the beginning of paragraph
- omit '%[Name]' to select default interpreter.
As more backend system integrated into Zeppelin, I propose new way of selecting interpreter.
- select interpreter by using '%[Group].[Name]' in the beginning of paragraph
- omit '[Name]' selects first available interpreter in the '[Group]'
- omit '[Group]' selects default '[Group]' of binded interpreter to specific note.
For example
Let's say there're four interpreters. 'spark' and 'sql' in interpreter group 'spark', 'ignite' and 'sql' in interpreter group 'ignite'.
Basic usage will be
'%spark.spark', '%spark.sql', '%ignite.ignite', '%ignite.sql'.
If we can omit name. then it'll be select first interpreter in the group. so can be used like
'%spark', '%spark.sql', '%ignite', '%ignite.sql'.
When notebook is configured group 'ignite' as a default interpreter, we can omit group like,
'%spark', '%spark.sql', '' (equals to '%ignite'), '%sql' (equals to '%ignite.sql')