Running in VSCode using jupyter

[JCzz]

Running in VSCode using jupyter

[Igorgro]

I have no problems doing this. After installing using instruction in README, the kernel just appeared in the vscode kernel selecton list

The only problem is that despite the gopls LSP is starting, there still no autocomplete , but the kernel iteslf working well

[joelschutz]

Following up on my previous comment:

1. The hack is the only way to work with notebooks on most of the languages servers currently. And both VSCode and Jupyter use it in Python.
2. LSP added support for notebook at version 3.17(latest). This mean that it could be possible to remove the hack.
3. Both gopls and vscode-go are mantained by Google, and notebook support is not a priority.

Maybe it would be possible to use this to create a middleware to handle notebooks.

[cosmos72]

It says "A Visual Studio Code extension that provides basic notebook support for language kernels that are supported in Jupyter Notebooks today" so yes, in theory it may work.

I think nobody tried gophernotes with it yet, so you're welcome to post here the results of any attempt you may want to do.

[Ale0x78]

I keep getting

The 'go' kernel is not available. Please pick another suitable kernel instead, or install that kernel. 

When inside VS-Code (no issues in JupyterLab though). Any chance anyone else has run into this issue?
I tried adding it into trusted kernal's with

"jupyter.kernels.trusted": ["~/Library/Jupyter/kernels/gophernotes"]

But still no luck.

[Ale0x78]

NVM looks like running a JupyterLab server via the terminal and then connecting to it works. Probably some kind of a config thing I am missing in VSCode.

[Nigel2392]

To add to @Igorgro, for me it is working great. Except:

1. There is no syntax highlighting.
2. The VSCode Go to Definition functionality does not work (Among others functionality.)

[joelschutz]

I'd been the entire day trying to find a way to do it. It's not a Kernel issue, go-kernel has the same issue.

The main problem is in the VSCode LSP protocol itself, it sends a path to the server to analyze. On the vscode-python they created a hack in the extension to concat a file before sending it the language server.

I don't think it's worth it to create something like it for Go, that's why in trying jupyter now.