nbhtml - save & load notebooks as HTML

[Thomas Kluyver]

Hi all,

I've been working on a project to save notebooks to HTML and load them back without losing any data. It's still quite experimental, but enough is working that I'd like people to kick the tyres:
https://github.com/takluyver/nbhtml

The idea, from discussions at the end of last year, is that we could eventually combine this with a tool like ipymd, so when you save notebooks you'd get two complementary formats:

- 'Complete' HTML notebooks, including images and outputs, which can easily be shared with people who may not have the Notebook application installed.
- 'Clean' notebooks containing only the input (code and markdown) and metadata, convenient for version control (e.g. in git), making diffing and merging easy.

I'm trying to do HTML roundtrip without duplicating all of the notebook content, so where possible it uses the visible data in the HTML file to reconstruct the notebook. Where the necessary data is not visible, it hides JSON data in invisible elements.

Thomas