What is this?

This is a simple utility to extract nicely formated code from a Zeppelin notebook

How does it work

The utility extract the code for the given interpreters and tries to nicely format comments based on Zeppelin interpreter titles and md blocks.

Usage Cases

1. You want to extract all code blocks from all interpreters. The result may not be able to run as it is likely a mix of different languages but it is convenenient if you want to diff your changes. Use 'all' for extract option.
2. You want to extract the code blocks specific to a given language. The result may be able to run. Stiill, you typically have to deal with graphing outputs in a non interactive way when you move from a notebook to a script. You will have to handle this yourself.

Limitations

• You may have to customize the interpreters list for your target languages and for 'all'
• Change the default Zeppelin notebook directory if it does not match yours

Possible developments

Add this under Zeppelin menu with an interactive list of intepretersd to include

Examples

• zep2code all *.json -> produces a file for each .json notebook into a code subfolder, add them to git
• python ./zep2code.py mynotebook.json /tmp/test3.py python -> produces a .py file runable in interactive mode
• python ./zep2code.py 2E27K8X9M /tmp/test3.r r -> produces a .r file runable in interactive mode
• python ./zep2code.py 2E27K8X9M /tmp/test3.all all -> produces a likely not runable file useful for checking code differences

Author(s)

Initial idea from https://github.com/sat28/zeppelin_notebook_to_script/blob/master/zeppelin_to_script.py Then all improvements by G. Demaneuf.