Investigate using podoc in nbconvert

[willingc]

As a follow on to the notebook to markdown discussion on nbsphinx repo, it would be good to try using podoc for conversion of notebooks as well. @rossant has done a bunch of work the past couple of weeks and the project is really taking shape.

cc/ @fperez @bnaul @takluyver @mgeier

[ellisonbg]

I am excited to see where podoc goes - definitely has the potential to
really help in our conversion machinery.

On Tue, Apr 5, 2016 at 6:30 PM, Carol Willing [email protected]
wrote:

As a follow on to the notebook to markdown discussion on nbsphinx repo, it
would be good to try using podoc https://github.com/podoc/podoc for
conversion of notebooks as well. @rossant https://github.com/rossant
has done a bunch of work the past couple of weeks and the project is really
taking shape.

cc/ @fperez https://github.com/fperez @bnaul https://github.com/bnaul
@takluyver https://github.com/takluyver @mgeier
https://github.com/mgeier

—
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub
#283

Brian E. Granger
Associate Professor of Physics and Data Science
Cal Poly State University, San Luis Obispo
@ellisonbg on Twitter and GitHub
[email protected] and [email protected]

[rossant]

I would be glad if podoc could help in nbconvert. Also I would really like to see podoc as the basis of a "Jupyter-aware RMarkdown for Python". I think the community would really benefit from something like this.

As @willingc said, podoc is taking shape. It can already convert notebook <-> markdown <-> pandoc, enabling conversions with all formats supported by pandoc. podoc now works in real time in the Jupyter Notebook, as a custom ContentsManager. I'm currently trying the library on increasingly complex notebooks and Markdown documents in order to fix various edge cases. In a few weeks hopefully, podoc should be stable enough to be used in production.

In any case I would love to collaborate with the Jupyter team on the long run, and I'm really open to discussion on all aspects of the library. I will open soon a few specific issues that I would like to discuss.

[ellisonbg]

Right now the design of the Notebook-HTML conversion is decent - in that
case, jinja based templates make a good amount of sense. - I know lots of
folks who are customizing HTML output by writing custom jinja templates.
But for other output formats, such as rst, pdf, latex, etc. the
architecture of nbconvert isn't ideal.

On Wed, Apr 6, 2016 at 3:15 AM, Cyrille Rossant [email protected]
wrote:

I would be glad if podoc could help in nbconvert. Also I would really like
to see podoc as the basis of a "Jupyter-aware RMarkdown for Python". I
think the community would really benefit from something like this.

As @willingc https://github.com/willingc said, podoc is taking shape.
It can already convert notebook <-> markdown <-> pandoc, enabling
conversions with all formats supported by pandoc. podoc now works in real
time in the Jupyter Notebook, as a custom ContentsManager. I'm currently
trying the library on increasingly complex notebooks and Markdown documents
in order to fix various edge cases. In a few weeks hopefully, podoc should
be stable enough to be used in production.

In any case I would love to collaborate with the Jupyter team on the long
run, and I'm really open to discussion on all aspects of the library. I
will open soon a few specific issues that I would like to discuss.

—
You are receiving this because you were mentioned.
Reply to this email directly or view it on GitHub
#283 (comment)

Brian E. Granger
Associate Professor of Physics and Data Science
Cal Poly State University, San Luis Obispo
@ellisonbg on Twitter and GitHub
[email protected] and [email protected]

[mpacer]

Any update on this?

[willingc]

@michaelpacer I believe that @rossant put some of his development on pause for a bit. His books are still some of my favorites so when dev restarts we will want to stay in the loop by following the repo. ☀️

[willingc]

And @rossant, if you haven't yet worked with @michaelpacer, he's working with us on Jupyter at Berkeley as a post-doc. Super helpful and interested in converting docs.

[rossant]

I've been swamped with personal matters lately but I'll come back to this in the coming weeks

[mgeier]

If I'm not mistaken, podoc uses CommonMark-py, which would be a nice step towards using CommonMark in the Jupyter world, see jupyter/notebook#1371.

But what about the CommonMark extensions that Jupyter is using, namely tables (see readthedocs/commonmark.py#28) and math (see readthedocs/commonmark.py#49)?

[mgeier]

OK, I saw that @rossant has included a pre-processing step to "decorate" math. So what about tables?