nbflask

Serves pdf from underlying jupyter notebook via http.

Create good looking, dynamic reports from jupyter notebooks.

features

• executes notebook with given request parameters
• caches resulting pdf
• execute notebook locally
• develop notebook as usual

requirements

• pandoc
• latex

usage

• make install for installing requirements
• make install-kernel install custom jupyter kernel
• make to-pdf to create a pdf from notebook
• make to-md to convert notebook to markdown
• make from-md to create notebook from markdown
• make notebook to start notebook as jupyter notebook in browser
• make start for running flask webservice on http://localhost:5000
• link for opening in browser
• curl http://localhost:5000/pdf?polynome=1,2,3 -o out.pdf for download

examples

see examples/test.md

TODO

• dockerize
• nicer plots (resolution?, svg?)
• error handling
• health checks
• extend example to use an sqliteDB
• build package
• consider pinning requirements
• check vulnerabilities
• create actual command line interface, using argparse or click
• add license
• create kernel via Makefile