cell magic syntax highlighting (Ruby)

[MatthewSteen]

I'm seeing this in Jupyter for 2.7. See this comment and the following one by @takluyver.

• jupyter_client-4.4.0
• jupyter_console-5.0.0
• jupyter_core-4.2.0

---

Jupyter - long cell magic applies Python highlighting to Ruby (expected)

---

Jupyter - short cell magic highlights all code as a string (unexpected)

[takluyver]

What version of the notebook package do you have? If you've got time, can you try installing from master (see setting up a development environment)?

I think it's this code that is going wrong.

[MatthewSteen]

I have notebook-4.2.3-py27_0.

[takluyver]

Can you try with the latest version, currently 4.3.1?

[MatthewSteen]

Same result with notebook-4.3.1-py27_1 after updating, shutting down my running notebooks, and restarting Jupyter.

[takluyver]

Thanks. In fact, I already tested on master but forgot about it earlier (sorry about that).

[Carreau]

I think it's this code that is going wrong.

Yes, I'm almost sure of what's wrong. It switches language in the middle of the first line instead of waiting to end of first line to switch, which screw up codemirror highlighting.

In general the auto-highlight of IPython should be folded into a real IPython codemirror mode. Though that's a non-negligible task. When I was familiar with CodeMirror internals and mode writing, I would have said that a clean IPython mode would take 1 to 2 good uninterrupted days of work. Plus about 1, to integrate into the current notebook JS.

So I'm going to guess this will likely end up being bumped to 5.1.

[takluyver]

Bumped for now; if anyone works out a fix, I'm happy to land it for 5.0, but I don't want to hold up the release for it.

[gnestor]

@Carreau Is there a quick-and-dirty fix for this that we can get into 5.1? Otherwise, let's mark as backlog...

[Carreau]

@Carreau Is there a quick-and-dirty fix for this that we can get into 5.1? Otherwise, let's mark as backlog...

No, not that I know of. We might also want to push that to jupyterlab