Give the power of Jupyter to your vim.
Note: This plugin is EXPERIMENTAL so far.
- Vim 8
+pythonor+python3
- Python 3 (as
python)jupyterjupyter_kernel_gatewaypsutil
:JupyterKernelConnect- Write some code within code fence.
- NOTE: Language must be specified in the upper fence.
- Press
<S-Return>(Shift+Return) in the code fence. - Yay!
- Example:
JupyterKernelConnect [address:port] [kernel][address:port](optional): Address and port to Jupyter Kernel Gateway (e.g.localhost:8888). If omitted, Jupyter Kernel Gateway will run in localhost with random port.[kernel](optional): Kernel name (e.g.python).
g:jupyterkernel#default_kernel: Default kernel. Defaults topython.
Say you have a powerful machine called remote-machine, and you want to calculate on it using jupyterkernel.vim.
You may do following steps:
- Login to
remote-machineand executejupyter kernelgateway- Requires
jupyter_kernel_gatewayinstalled. - You may need to specify
--portand--ip.
- Requires
- At your local machine, launch vim and execute
JupyterKernelConnect remote-machine:8888
+-----+
| Vim |
+--+--+
| TCP
+--+-----------------------------------------------+
| jupyterkernel_client.py (on machine vim running) |
+--+-----------------------------------------------+
| Websocket
+--+-----------------------------------+
| Jupyter Kernel Gateway (on any host) |
+--------------------------------------+
- Run
jupyterkernel_client.pymanually.- Can be interrupted by
Ctrl-C.
- Can be interrupted by
python autoload/jupyterkernel/jupyterkernel_client.py --vim_port 55555- Launch Vim
- Set
g:jupyterkernel#_client_portto the port- e.g.
let g:jupyterkernel#_client_port = 55555
- e.g.
- Execute command
JupyterKernelConnect. - Run some code.
- Log should show in the terminal which runs
jupyterkernel_client.py.
- Save to, and load from
.ipynb - Notify disconnect
- Explicit quitting, other than closing buffer
- Closing buffer is only a way to close session (i.e. kill kernel), so far.