kracekumar / jut Goto Github PK

Textual looks interesting and makes jut look better for rendering image and pagination.

The user should be able to filter the input cells matching the text or containing the text.

I don't see a command-line flag for this? Is there a way to do this?

jut --input-file foo.ipynb --from 23 --to 30

The above command should display the cells from 23 to 30.

• https://github.com/derf/feh
• https://github.com/trashhalo/imgcat

I'm finding that even if I use less -R, when paginating jut's output I lose color highlights, not sure why... But given that your typical notebook document is long-ish, I would suggest adding pagination support less-Style, if it's not too hard...

Alternatively, figure out a way if possible to make it so coloring survives through less, which would save you from having to reimplement less's features (I'm sure if you add pagination, people will then want search, regexes, etc, and you'd end up having to reinvent less itself :)

Great tool, thanks! I just discovered it thanks to @davidbrochart's ping on an nbterm issue. Delighted to see more terminal-based tools evolving to help us work with notebooks when a full GUI isn't needed.

Quick suggestion: I think the most common use case is simply "show me this file", so I think it would be a nice usability improvement to allow jut filename.ipynb work automatically without further flags needed. This amounts to making -i the default behavior.

You could further auto-detect remote urls with a regex along the lines of qualifier:://path (I haven't toyed with that but I'm sure they exist) to allow the omission of --url too, but I see that as secondary.

In any case, thanks again!

(great thanks for your nice tool!)

It would be a handy enhancement for the CLI to accept a positional argument and guess whether it's a URL or FILE, an ac accordingly.

When one of the cells is an image, allow exporting the image to a file and print the file path as output.

Support all v4 versions parsing specified in nbformat - https://github.com/jupyter/nbformat/tree/master/nbformat/v4

also, let user specify the version from the command line argument

Accept type of cells to display from the python notebook.

Reference: https://github.com/jupyter/nbformat/blob/master/nbformat/v4/nbformat.v4.0.schema.json#L276

As of now, the only code output is displayed. The code error is ignored from rendering.

Given a notebook or a URL, allow the user to edit the cell and save the cell alone.

Selecting and copying blocks within the reported output includes the | symbol, which makes it tedious to strip out these symbols to get the valid embedded Python code. The visual display is great, but sometimes it makes working with the output in the terminal tedious.