* -> ipynb converter

[jdfreder]

Migrated from https://github.com/ipython/nbconvert/issues/96

---

Meta issues to track the other format we would like to convert from

• rst
• markdown
• knitr ?
• .py
• other propriatary format?

[bollwyvl]

More as a lark than anything else, I started a general implementation of nbconvert importers, with an initial markdown importer:
https://github.com/bollwyvl/ipython/tree/markdown-importer

It borrows most of its implementation language from the exporters.

This one handles:

• markdown(!)
• indented code blocks
• fenced code blocks (with language hint, which it optimistically turns into a magic)
• doctests in either of the above

There are some strange things happening:

ipython nbconvert --to notebook --from markdown notebook1.md

You get, as one might expect, notebook1.md.ipynb and notebook1.md.nbconvert.ipynb.
If you run it again, you start getting notebook1.md.ipynb.nbconvert.ipynb. Haven't figured that out yet.

Interested in feedback, will PR if there is interest.

[minrk]

I think we probably should add the notion of a Reader or Importer (whatever we end up calling it) to nbconvert. I think we should not add this functionality before splitting nbconvert into its own standalone project.

[damianavila]

It is funny that nbconvert start standalone and will be standalone again... jeje... regarding the issue itself, I think that the Reader/Importer is a need for the future... some random thought for the future... I think we need to keep the symmetry in the exportation/importation cycle...