kn — nvgt/fldrs/qckly

cargo install kn

Then follow the configuration instructions.

Features

• Abbreviations
• Wildcards
• Multiple dots
• --exclude-old-pwd

Abbreviations

You can use kn just like you'd use cd. The difference is that you can also navigate with abbreviations instead of full dir names. For example, instead of foo/bar you can type fo/ba.

.
├── foo
│  └── bar
├── bar
├── photosxxxxxxxxxxx2021
└── photosxxxxxxxxxxx2020

kn foo/bar          # Use `kn` just like `cd`...
kn fo/ba            # ...or navigate with abbreviations! No asterisks required.
kn pho2021          # Type only the significant parts of the dir name. You can skip the middle part.

Wildcards

You can also use wildcards - to avoid typing a dir name altogether i.e. kn -/ba to go to foo/bar. Note that kn f-/b- will not match foo/bar. In this case - functions as a literal character.

kn -/bar            # Wildcards can be used to skip a dir name altogether (changes dir to ./foo/bar/).

Multiple dots

kn splits the arg into two parts, a prefix and a sequence of abbreviations. The prefix may contain components like c:/, /, ~/, ., .. and it is treated as a literal path. It may also contain components with more than two dots, which are interpreted like this:

kn ..               # Go to parent dir (as usual).
kn ...              # Go to grandparent dir (same as ../..).
kn ....             # Go to grand-grandparent dir (same as ../../..).

kn ........         # Type as many dots as you want!
kn .../....         # This works as well.

kn .../..../abbr    # You can put abbreviations after the prefix.

If any of the mentioned components occurs in the path after an abbreviation, it is treated as an abbreviation.

kn ./../foo/bar/../baz
#  ^---^                 prefix
#       ^------------^   abbreviations

. and the first .. mean current dir and parent dir, while the second .. is treated as an abbreviation, that is, it will match a dir name containing two dots.

--exclude-old-pwd

This flag excludes your previous location from the search. You don't have to type it when using kn, just set it in your shell script (notice the underscore in _kn):

_kn init --shell fish --exclude-old-pwd

It's useful when two paths match your abbreviation and you enter the wrong one:

my-files/
$ kn d

my-files/dir-1/
$ kn -

my-files/
$ kn d # just press arrow up twice

my-files/dir-2/
$ # success!

In order for kn to exclude the previous location there must be at least one other match and the provided arg must not be a literal path (that is, it must be an abbreviation).

Installation

Make sure to configure your shell after the installation.

From crates.io

cargo install kn

From source

1. git clone https://github.com/micouy/kn.git

2. cd kn

3. Put the binary in a folder that is in PATH:

   cargo build -Z unstable-options --out-dir DIR_IN_PATH --release

   Or just build it and copy the binary to that dir:

   cargo build --release

   cp target/release/_kn DIR_IN_PATH

From the release page

Download a binary of the latest release for your OS and move it to a directory which is in your $PATH. You may need to change the binary's permissions by running chmod +x _kn.

If there are any problems with the pre-compiled binaries, file an issue.

Configuring your shell

Then add this line to the config of your shell (notice the underscore in _kn):

• fish (usually ~/.config/fish/config.fish):

  _kn init --shell fish | source

• bash (usually ~/.bashrc):

  eval "$(_kn init --shell bash)"

• zsh (usually ~/.zshrc):

  eval "$(_kn init --shell zsh)"

You may also want to enable the --exclude-old-pwd flag. To be able to use kn, reload your config or launch a new shell instance.

Help wanted

In this project I have entered a lot of areas I have little knowledge about. Contributions and criticism are very welcome. Here are some things you can do:

• Check the correctness of scripts in init/.
• Add scripts and installation instructions for shells other than fish, bash and zsh.
• Review Github Actions workflows in .github/workflows/.

The algorithm

kn doesn't track frecency or any other statistics. It searches the disk for paths matching the abbreviation. If it finds multiple matching paths, it orders them in such a way:

1. Compare each component against the corresponding component of the abbreviation. The components of the path may or may not match the abbreviation. If a component matches the abbreviation, there are three possible results:

   • Complete if the corresponding components are equal.
   • Prefix if the abbreviation's component is a prefix of the path's component.
   • Subsequence(coefficient) if the abbreviation's component is a subsequence of the path's component. The coefficient is the Powierża coefficient of these strings.

   Retain only these paths in which all of the components match.

2. Order the paths in reverse lexicographical order (compare the results from right to left). Complete then Prefix then Subsequence. Order paths with Subsequence result in ascending order of their coefficient's.

3. Order paths with the same results with alphanumeric_sort::compare_os_str.