Accessing notebook cell metadata via a BUILTIN plugin about jupyterlab-markup HOT 7 CLOSED

At present we use the existing JupyterLab scaffolding for Markdown rendering. This means that the Markdown renderer only sees the Markdown source, rather than the cell model. Because of this, it is not currently possible to access the cell through a plugin unless you do this yourself.

You could hack this in - the plugin mechanism is just a wrapper around the JupyterLab extension mechanism. You could write a more advanced extension that references the notebook tracker, and try to do some kind of lookup, but it would be fairly fragile.

from jupyterlab-markup.

In the long run, there is definitely scope for thinking about how cells should interact with Markdown rendering. There are already areas where this has interesting ramifications (https://github.com/agoose77/jupyterlab-imarkdown/)

from jupyterlab-markup.

... the Markdown renderer only sees the Markdown source, rather than the cell model.

Ah, ok... I think that's probably beyond my skillset atm. I did have a look at jupyterlab-imarkdown/ and will see if I can get get enough of a grip on it to strip it back to just giving me access to a cell / cell markdown, customising metadata on the basis of tag settings, and then seeing if I can just render that customised markdown.

from jupyterlab-markup.

Dragons beware! It's pretty gnarly code, mainly because JupyterLab's notebook rendering is not very customisable at the moment. jupyterlab-imarkdown overrides the machinery that returns a Markdown cell. It returns a special cell that listens for rendered markdown, and post-processes it to identify the special expression syntax. Meanwhile, there is a separate code path that evaluates the parsed cell expressions and replaces the placeholders in the HTML with the rendered results

from jupyterlab-markup.

Presumably the model.value.text is the original, raw markdown. When the markdown cell has rendered, can you also access the rendered output on the cell?

from jupyterlab-markup.

Unfortunately the cell content is not identical to the text given to the markdown renderer - we strip out the math expressions first.

When the markdown cell has rendered, can you also access the rendered output on the cell?

By default, no. That's what the imarkdown extension has to overload the cell class for.

from jupyterlab-markup.

The design of the notebook handler / renderer all seems completely messed up to me...

from jupyterlab-markup.