extrakto for tmux

Output completions - you can complete commands that require you to retype text that is already on the screen. This works everywhere, even in remote ssh sessions.

You can fuzzy find your text instead of selecting it by hand:

press tmux prefix + tab to start extrakto
fuzzy find the text/path/url/line
use custom filters (ctrl + f)
press
- tab to insert it to the current pane,
- enter to copy it to the clipboard,
see other features in HELP (press ctrl-h in extrakto)

Use it for paths, URLs, options from a man page, git hashes, docker container names, ...

Requirements

tmux - popups require 3.2, otherwise extrakto will open in a split window.
fzf
Python 3.6+
Bash (tested with 5.0+, on macOS please brew install bash first) see the Wiki for Bash 4 support

Supported clipboards:

Linux Xorg (xclip) and Wayland (wl-copy)
macOS (pbcopy)
WSL (aka "Bash on Windows")
bring your own, see the Wiki for examples (like termux)

Installation with Tmux Plugin Manager

Add the plugin to the list of TPM plugins in .tmux.conf:

set -g @plugin 'laktak/extrakto'

Hit prefix + I to fetch the plugin and source it.

You should now have all extrakto key bindings defined.

Manual Installation

Clone the repo:

$ git clone https://github.com/laktak/extrakto ~/clone/path

Add this line to the bottom of .tmux.conf:

run-shell ~/clone/path/extrakto.tmux

Reload the tmux environment:

# type this in terminal
$ tmux source-file ~/.tmux.conf

You should now have all extrakto key bindings defined.

Wiki

Add or look for special tips in our wiki.

Options

To set any of these options write on your ~/.tmux.conf file:

set -g <option> "<value>"

Where <option> and <value> are one of the specified here below

Common Options

Option	Default	Description
`@extrakto_grab_area`	`window full`	Whether you want extrakto to grab data from the `recent` area, the `full` pane, all current window's (`window recent`) areas or all current window's (`window full`) panes. You can also set this option to any number you want (or number preceded by "window ", e.g. "window 500"), this allows you to grab a smaller amount of data from the pane(s) than the pane's limit. For instance, you may have a really big limit for tmux history but using the same limit may end up on having slow performance on Extrakto.
`@extrakto_filter_order`	`word all line`	Filter modes order. The first listed mode will be the default when opening extrakto. You may use `word`, `line`, `path`, `url`, `quote`, `s-quote` or any of your own filters separated by a space. `all` applies all filters at the same time.

Keys

Option	Default	Description
`@extrakto_key`	`tab`	The key binding to start. If you have any special requirements (like a custom key table) set this to 'none' and define a binding in your `.tmux.conf`. See `extrakto.tmux` for a sample.
`@extrakto_copy_key`	`enter`	Key to copy selection to clipboard.
`@extrakto_insert_key`	`tab`	Key to insert selection.
`@extrakto_filter_key`	`ctrl-f`	Key to toggle filter mode.
`@extrakto_grab_key`	`ctrl-g`	Key to toggle grab mode.
`@extrakto_edit_key`	`ctrl-e`	Key to run the editor.
`@extrakto_open_key`	`ctrl-o`	Key to run the open command.

All but @extrakto_key are controlled by fzf and must follow its conventions.

Advanced Options

Option	Default	Description
`@extrakto_clip_tool_run`	`bg`	Set this to `tmux_osc52` to enable remote clipboard support or `fg`/`bg` to have your clipboard tool run in a foreground/background shell.
`@extrakto_clip_tool`	`auto`	Set this to whatever clipboard tool you would like extrakto to use to copy data into your clipboard. `auto` will try to choose the correct clipboard for your platform.
`@extrakto_editor`		This defaults to `$EDITOR` if not set.
`@extrakto_fzf_layout`	`default`	Control the fzf layout which is "bottom-up" by default. If you prefer "top-down" layout instead set this to `reverse`. In fact, this value is passed to the fzf `--layout` parameter. Possible values are: `default`, `reverse` and `reverse-list`
`@extrakto_fzf_tool`	`fzf`	Set this to path of fzf if it can't be found in your `PATH`.
`@extrakto_fzf_unset_default_opts`	`true`	Unsets custom FZF_DEFAULT_OPTS as it can potentially cause problems in extrakto operations
`@extrakto_open_tool`	`auto`	Set this to path of your own tool or `auto` to use your platforms open implementation.
`@extrakto_popup_position`	`C`	Set position of the tmux popup window. Possible values are in the `display-popup` entry in `man tmux`. Set this to `x,y` to set the x and y positions to `x` and `y` respectively.
`@extrakto_popup_size`	`90%`	Set width and height of the tmux popup window. Set this to `w,h` to set the width to `w` and height to `h`.
`@extrakto_split_direction`	`a`	Whether the tmux split will be `a`uto, `p`opup, `v`ertical or `h`orizontal
`@extrakto_split_size`	`7`	The size of the tmux split (for vertical/horizontal)

Examples

set -g @extrakto_split_size "15"
set -g @extrakto_clip_tool "xsel --input --clipboard" # works better for nvim
set -g @extrakto_copy_key "tab"      # use tab to copy to clipboard
set -g @extrakto_insert_key "enter"  # use enter to insert selection
set -g @extrakto_fzf_unset_default_opts "false"  # keep our custom FZF_DEFAULT_OPTS

Custom Filters

You can define your own filters by creating a file in ~/.config/extrakto/extrakto.conf:

[quote]
regex: ("[^"\n\r]+")

To override an existing filter copy it to your file first.

If you want to remove one of the alternate filters you can set it to None:

[quote]
alt2: None

See extrakto.conf for syntax and predefined filters.

CLI

You can also use extrakto as a standalone tool to extract tokens from text.

Installation

For now simply clone the repository and link to the tool somewhere in your path:

git clone https://github.com/laktak/extrakto
cd extrakto
# assuming you `export PATH=$PATH:~/.local/bin` in your `.bashrc`:
ln -s $PWD/extrakto.py ~/.local/bin/extrakto

Requires Python 3.6+.

CLI Usage

usage: extrakto.py [-h] [--name] [-w] [-l] [--all] [-a ADD] [-p] [-u] [--alt] [-r] [-m MIN_LENGTH] [--warn-empty]

Extracts tokens from plaintext.

optional arguments:
  -h, --help            show this help message and exit
  --name                prefix filter name in the output
  -w, --words           extract "word" tokens
  -l, --lines           extract lines
  --all                 extract using all filters defined in extrakto.conf
  -a ADD, --add ADD     add custom filter
  -p, --paths           short for -a=path
  -u, --urls            short for -a=url
  --alt                 return alternate variants for each match (e.g. https://example.com and example.com)
  -r, --reverse         reverse output
  -m MIN_LENGTH, --min-length MIN_LENGTH
                        minimum token length
  --warn-empty          warn if result is empty

Contributions

Thanks go to all contributors for their ideas and PRs!

If you make a PR, please keep it small so that it's easier to test and review. Try to create one PR per feature/bug.

Please run black if you change any python code and run shfmt if you change any bash files.

laktak / extrakto