euporie

About

Euporie is a text-based user interface for running and editing Jupyter notebooks.

Install

Euporie is on pypi, so can be installed using pip:

pip install euporie

If you are using Windows, you may wish to install some optional python dependencies to render images and HTML tables:

pip install euporie[html-mtable,images-timg]

Screenshots

Usage

To open a notebook file, pass the file name as a command line parameter:

euporie ~/my-notebook.ipynb

Features

• Execute notebooks in the terminal
• Autocompletion in code cells
• Rich output support, including:
  • Markdown
  • Tables
  • Images
• Open multiple notebooks side-by-side
• Good performance with large notebook files

Image Support

Euporie will attempt to render images in the best possible way it can. The following methods are supported:

• Sixel

  If supported by your terminal, euporie can show graphical images in cell outputs This requires one of the following:

  imagemagik timg teimpy

• Kitty's terminal image protocol

  If your terminal supports kitty's terminal graphics protocol, euporie will use it to render images

• Ansi art

  This requires one of the following:

  timg catimg icat timg tiv viu img2unicode jp2a img2txt

The kitty & sixel image rendering methods will fall back to ansi art images when rendering images in partially obscured cells, to prevent clipped images destroying the user interface.

For SVG support, cairosvg or imagemagik are required.

HTML Support

Euporie will attempt to render HTML outputs. This requires one of the following:

w3m elinks lynx links mtable

Note: only HTML tables will be displayed if mtable is used

If none of these commands are found in your $PATH, the plain text representation will be used.

Key Bindings

Command	Key Binding
Quit	Ctrl + q
Open notebook	Ctrl + o
New notebook	Ctrl + n
Close notebook	Ctrl + w
Select next cell	Up, k
Select previous cell	Down, j
Page up (move up 5 cells)	Pgup
Page down (move down 5 cells)	PgDown
Scroll up	[
Scroll down	]
Enter edit mode	Enter
Exit edit mode	Esc*, Esc, Esc
Edit cell in $EDITOR	e
Run cell	Ctrl + Enter**, Ctrl + Space, F20
Run cell and select next cell	Shift + Enter**, F21
Insert cell above selected cell	a
Insert cell below selected cell	b
Toggle line numbers	l
Copy cell	c
Cut cell	x
Paste cell	v
Delete cell	dd

* There is a slight delay detecting an escape key-event. To exit edit mode quickly, double-press the escape key.

** These entries require your terminal to support CSI-u mode. If your terminal does not support this, it may be possible to work around this by remapping the keys in your terminal emulator - see below).

When in edit mode, emacs style key-bindings apply.

Key Remapping

By default, VT100 terminal emulators do not distinguish between Enter, Ctrl + Enter & Shift + Enter. In order to work around this, it is possible to re-map these key bindings so they produce the escape code of another key. To replicate the Ctrl + Enter & Shift + Enter of Jupyter, you will need to remap the following shortcuts in your terminal:

Key Combination	Output
Ctrl + Enter	Ctrl + F20
Shift + Enter	F21

xterm

Add the following to your ~/.Xresources

*.vt100.translations: #override \n\
    Ctrl <Key>Return: string("\033\[19;6~") \n\
    Shift <Key>Return: string("\033\[20;2~") \n\

konsole

In the menu, navigate to:

Settings -> Edit Current Profile -> Keyboard -> Edit

Change the existing entry for Return+Shift to Return+Shift+Ctrl (or whatever you prefer), then add the following entries:

Key combination	Output
Return+Ctrl	\E\[19;6~
Return+Shift	\E\[20;2~

Roadmap

• Add a configuration file to expose configurable settings
• Add ability to dump formatted notebooks
• Add command line argument parsing
• Render outputs asynchronously in a separate thread
• Upstream markdown tables in rich
• Cell attachments
• LaTeX
• Widgets

Related projects

• https://github.com/davidbrochart/nbterm

  An alternative effort sponsored by QuantStack

• https://github.com/chentau/nbtui

  A cli Jupyter notebook viewer with support for kitty's terminal graphics protocol

• https://github.com/mosiman/jupytui

  A cli Jupyter notebook viewer

• https://kracekumar.com/post/jut/

  Another cli Jupyter notebook viewer