A program for interactive navigating through weighted hierarchical structured data and analyzing them. (weighted means that each item have its own value - number)
This is a nice tool for analyzing and researching datasets, and simplicity of input format makes it easy to write your own scripts that produces data you want to analyze fast, right in shell (like something in files, git meta data (number of commits per each file), etc. For more see examples).
File space usage:
Git commits per file, linux kernel 6.8 repo:
- interactive navigation
- items sorted by their recursive weight-sums
- accepts any data from stdin, so you can easily make your owns (see
examples)
- input format can be altered (see Configuration section)
- interface:
- preview window (can be toggled)
- human-readable big numbers formatting (can be toggled)
... | nctok
Where ...
should produce output in format:
<number> <path/to/item>
<number> <path/to/another_item>
For concrete usecases see examples.
j
/k
- go down/up in listl
/h
- go deep down/up (inside/out)g
- go to top (very first item)G
- go to bottom (latest item)w
- toggle preview windowa
- toggle human-readable numbers formattingq
- quit
$ nctok --help
...
Options:
--reverse Invert sort order
Parser:
--number-delimiter <NUMBER_DELIMITER>
Delimiter between items number value and path [default: " "]
--path-separator <PATH_SEPARATOR>
Separator in items path [default: /]
Interface:
--preview
Show preview window for entry under cursor (can also toggle with 'w' key)
--no-human-readable
Disable formatting big numbers in human-readable (can also toggle with 'a' key)
With cargo (from crates.io)
cargo install nctok
cargo install --path .
- ncdu - (ncurses) disk usage analyzer. This is main source of inspiration. I took exactly their idea and made unified interface for analyzing any data with the same (and a bit better) interface.
- vifm - terminal file manager. I took preview window from it.
- tokei - lines-of-code counter. Well, at very first I wanted tokei+ncdu experience, so this project was born.